Log

Append-only. One entry per operation: date, operation, source, pages created / updated.

2026-08-30 — setup#

• Created folder structure, CLAUDE.md rulebook, README.md, raw/sources.md.
• Adapted the LLM-wiki guide (natural20.com) for chiptune: added techniques/ as a first-class page type, kind for entities, slug filenames, per-article ingestion for multi-article sources, triage and scout operations.

2026-08-30 — ingest: SID-Wizard 1.8 User Manual (raw/sid-wizard/SID-Wizard-1.8-UserManual.txt)#

• Summary: s-sid-wizard-manual.
• Created entities: sid-wizard (with full cheat sheet), sid, commodore-64, hermit, goattracker, vice, sid-format.
• Created concepts: tracker, instrument-tables, adsr-envelope, sid-player-routine, orderlist.
• Created techniques: hard-restart, filter-programming, wavetable-programming, ring-modulation-and-sync, multispeed; SID-Wizard sections in chord-arpeggio, vibrato, pitch-slide-and-portamento, pulse-width-modulation, shuffle-funktempo, detune, fake-echo, octave-bass, instrument-design.
• Skipped: nothing; the whole manual is in scope. Key/menu minutiae only partially reproduced (cheat sheet keeps the most-used keys).

2026-08-30 — ingest: 0 3 7 lab (raw/037-lab/037-lab-techniques-and-theory.md, raw/037-lab/README.md)#

• Summary: s-037-lab.
• Created entities: 037-lab, openmpt, impulse-tracker (stubs for the last two).
• Created concepts: semitone-math, scales-and-modes, diatonic-chords, chord-inversions, chord-progressions, melody-writing, rhythm-and-groove, arrangement.
• Created techniques: chord-arpeggio, vibrato, pitch-slide-and-portamento, pulse-width-modulation, detune, fake-echo, octave-bass, sidechain-pump, channel-interleaving, instrument-design.
• Open question for the owner: what "DUET" (the DUET SW clipboard target) is.

2026-08-30 — triage: TraxWeekly archive (raw/traxweekly/, 128 files, 5.6 MB)#

• Downloaded all issues from https://resources.openmpt.org/traxweekly/ plus the archive index text; summary s-traxweekly-archive, entity traxweekly.
• Generated inbox/traxweekly-toc.md: 909 articles, 89 flagged ★ by title keywords. Nothing ingested yet — the owner picks articles.
• Note: TW-ART1..7.TXT are ASCII-art logo files, not articles.

2026-08-30 — lint#

• 45 pages, 0 broken links apart from the 11 intentional "wanted" red links in the index, 0 orphans, all frontmatter valid, 28 (unverified) markers across 23 pages awaiting sources (mostly LSDJ/IT commands and SID hardware details).

2026-08-30 — ingest: Little Sound Dj v9.2.6 manual (raw/lsdj/LSDj_9_2_6.pdf)#

• Downloaded the PDF, extracted text with a PDFKit script (raw/lsdj/LSDj_9_2_6.txt). Summary: s-lsdj-manual.
• Created entities: lsdj (screens, instruments, tables, grooves, synth, full command table, workflow, live mode), game-boy, game-boy-apu, johan-kotlinski. Created concept: sync. Created techniques: retrigger, randomization-and-probability, stereo-panning.
• Rippled ### LSDj sections into chord-arpeggio, vibrato, pitch-slide-and-portamento, pulse-width-modulation, shuffle-funktempo, fake-echo, detune, instrument-design, wavetable-programming, sidechain-pump, instrument-tables, orderlist, tracker, arrangement, rhythm-and-groove, adsr-envelope; replaced the earlier (unverified) LSDJ notes with sourced ones.
• Skipped: appendix B allophone list and appendix C SRAM memory map (reference material, in raw), sample-kit history blurbs beyond the kit names.

2026-08-30 — ingest (basics only): DUET README (raw/duet/README.md)#

• Summary: s-duet-readme. Created entity duet. Updated 037-lab (DUET question resolved), sid-wizard (1.97 / .swq exist), hermit, impulse-tracker (ITTECH.TXT), openmpt (clipboard shared with Schism/DUET).
• Per the owner: DUET's docs are not ingested for now.

2026-08-30 — ingest: TraxWeekly practical articles (34 articles from 24 issues)#

• Articles: Modal and Chord Theory (#1), Beginning the Tracking Experience (#3), The Death of the Chiptune / Chiptunes Aren't Dead! (#37–38), Chord Theory? / Chord What? ×2 (#39–40), Panning / Panning? Hello? (#55–56), Tracking Hints (#59), Motherboard Sampling (#74), Introduction to Musical Theory vols 1–5 (#74, 76, 79, 81, 86), IT Percussion Tips (#75), Realism in MODs (#76), A Chip Off the Block I–II + Chiptune Seminar 3–4 (#77, 78, 80, 81), General Tracking Tips (#77), Impulse Tracker Tip of the Week #1–8 (#81–89) + More Impulse Tracker Tips (#83), A Little Panning (#106; #107 is a reprint).
• Skipped: "Bass: A Perspective" (#5) — a recruitment ad for a group called Bass; the Modsquad's parts 3–4 except their chord list; ads and closing sections.
• Summaries (14, series and debates grouped): s-tw001-modal-and-chord-theory, s-tw003-beginning-the-tracking-experience, s-tw037-death-of-the-chiptune-debate, s-tw039-chord-theory-debate, s-tw055-panning-debate, s-tw059-tracking-hints, s-tw074-motherboard-sampling, s-tw074-intro-to-musical-theory, s-tw075-it-percussion-tips, s-tw076-realism-in-mods, s-tw077-chip-off-the-block, s-tw077-general-tracking-tips, s-tw081-it-tip-of-the-week, s-tw106-a-little-panning.
• Created: concept chiptune, tracking-workflow; techniques chip-samples, new-note-actions, sample-offset, instrument-envelopes; entities jeffrey-lim, pinion, greg-heo, zinc, trixter, basehead, maelcum, necros, psibelius, leviathan, the-modsquad, kosmic, hornet, 20-minute-chip-compo, scream-tracker-3, fasttracker-2, protracker.
• Rewritten: impulse-tracker (author, features, command cheat sheet), traxweekly (staff, distribution, ingested list).
• Rippled: vibrato, pitch-slide-and-portamento, chord-arpeggio, stereo-panning, fake-echo, randomization-and-probability, tracker, melody-writing, arrangement, rhythm-and-groove, diatonic-chords, chord-inversions, scales-and-modes, semitone-math, chord-progressions, instrument-design, detune, adsr-envelope. Several earlier (unverified) IT/S3M command notes are now sourced.
• Next candidates found while reading: Catspaw's frames-and-tempo articles (#58–59), "Perfect Samples: The Piano" (#84), "Wrecking Samples with Impulse Tracker" and "Hexadecimal 101" (#86), "Loops and Extended Samples" (Zapper, #25), "Impulse Tracker v1.01" (#42), Necros's "What it's like to be a PC Musician" (#5), and a possible vol. 6 of Greg Heo's series.

2026-08-30 — ingest: TraxWeekly candidate articles (8 articles) + Impulse Tracker docs#

• TraxWeekly: What it's like to be a PC Musician (#5, Necros), Impulse Tracker v1.01 (#42, Pulse), The Finishing Touches parts 1–2 (#58–59, Catspaw), Perfect Samples: The Piano (#84, Sklathill), Hexadecimal 101 and Wrecking Samples with Impulse Tracker (#86), Loops and Extended Samples (TW.025, The Zapper). No vol. 6 of Greg Heo's theory series exists; instead the TOC shows the Modsquad's "Music Theory: A Modern Approach" (#87–97) and Zinc's "Layman's Music Theory" (#107) as further candidates. Summaries: s-tw005-what-its-like-to-be-a-pc-musician, s-tw042-impulse-tracker-v1-01, s-tw058-the-finishing-touches, s-tw084-perfect-samples-the-piano, s-tw086-hexadecimal-101, s-tw086-wrecking-samples-with-impulse-tracker, s-tw025-loops-and-extended-samples.
• Impulse Tracker docs (raw/impulse-tracker/): IT.TXT manual and HINTS.TXT ingested in full (s-it-manual, s-it-hints); ITTECH.TXT at the semantic level only — volume/slide/pan formulas, instrument fields, channel allocation (s-ittech); byte layouts stay in raw per the owner's note that it is a technical reference.
• Created: it-format, catspaw, the-zapper; techniques song-init-and-looping, volume-slides, sample-looping, multisampling. Rewritten: impulse-tracker as a full reference (screens, complete effect table, editing keys, samples, instruments, song variables, community knowledge).
• Rippled into necros (real name now sourced), jeffrey-lim, psibelius, tracker, retrigger, arrangement, chiptune, scream-tracker-3, fasttracker-2, hornet, traxweekly, chip-samples, stereo-panning, fake-echo, detune, vibrato, pitch-slide-and-portamento, chord-arpeggio, randomization-and-probability, instrument-envelopes, new-note-actions, sample-offset, tracking-workflow, instrument-design, volume-slides.
• Remaining TraxWeekly candidates: Basehead's tracking tips (#15, cited by Catspaw), Catspaw's "Real Music" (#53) and interview (#57), "The Importance of Music Theory" (#7), Modsquad "Music Theory: A Modern Approach" (#87–97), Zinc "Layman's Music Theory" (#107), Necros's reviews.

2026-08-30 — query: clean, punchy SID bass without clicks#

• Consulted: hard-restart, adsr-envelope, instrument-tables, wavetable-programming, pulse-width-modulation, filter-programming, octave-bass, instrument-design, ring-modulation-and-sync, sid; grep of raw/ for "click"/"bass" (TraxWeekly only covers sample clicks — not relevant to the SID).
• Answered from the wiki: hard-restart settings (ADHR, timer, type, frame-1 waveform), gate-off / legato in patterns, envelope + short pitch-drop table + 25 % pulse or filtered waveform for punch. General-knowledge additions (typical ADHR/timer values, ADSR-bug explanation, test-bit phase reset) marked unverified.
• Offered to save the synthesis as a technique page (sid-bass-design); not created yet.

2026-08-30 — scout: SID sound design#

• Wrote inbox/scout-sid-sound-design-2026-08-30.md: 10 ranked candidates (GoatTracker 2.72 readme, WitchMaster's Creating Chip Tunes with SID-Wizard, Chordian's SID Factory II instruments tutorial, two Lemon64 ADSR/hard-restart threads, Linus Walleij's SID player notes, C64 Programmer's Reference Guide ch. 4, chipmusic.org "C64 Music for Dummies", Newman's Driving the SID chip, a Lemon64 resource thread, Furnace's C64 docs) plus 7 lower-priority sightings. Nothing ingested.
• Noted a contradiction to check on ingest: Walleij says hard restart matters only on the 6581; the Lemon64 threads describe the envelope lockup generically.
• chipmusic.org blocks non-browser fetchers (curl with a browser UA works); sidmusic.org was unreachable.

2026-08-30 — ingest: GoatTracker v2.72 readme (raw/goattracker/goattracker-2.72-readme.txt)#

• Fetched from the leafo/goattracker2 GitHub mirror; Latin-1 original plus a UTF-8 copy. Summary: s-goattracker-readme.
• Created: lasse-oorni. Rewrote goattracker from a stub into a full reference (keys, commands, instrument parameters, four tables, tick schedule, hard restart, packer options).
• Added ### GoatTracker sections to hard-restart, wavetable-programming, pulse-width-modulation, filter-programming, chord-arpeggio, vibrato, pitch-slide-and-portamento, multispeed, shuffle-funktempo, octave-bass, instrument-design; GoatTracker sections in instrument-tables, orderlist, sid-player-routine, adsr-envelope; waveform bits and routing in sid, ring-modulation-and-sync; notes in tracker, vice, sid-format, tracking-workflow, sid-wizard.
• Skipped: the byte-level .SNG/.INS/SFX file-format sections and the version history beyond the value-changing notes.

2026-08-30 — ingest: Creating Chiptunes with SID-Wizard, 2nd edition (raw/sid-wizard/Creating-Chiptunes-with-SID-Wizard-2nd-edition.pdf)#

• The owner's local copy (WitchMaster & Hermit, 2014-02-11, for SID-Wizard 1.6); text extracted with the new tools/pdf-to-text.swift. Summary: s-witchmaster-creating-chiptunes-with-sid-wizard.
• Created: witchmaster, bob-yannes. Updated: sid-wizard (book section: version history, BPM formula, player numbers), hermit, instrument-design (Bazz / Snare / Arpeggio recipes with every value), adsr-envelope (timing table), sid (datasheet waveform descriptions, 6581/8580, test bit), hard-restart (common ADHR values, staccato type), wavetable-programming (WF/ARP semantics, absolute-note values, note-off rows), pulse-width-modulation, filter-programming (combined bands, cutoff range, filter-switch override), instrument-tables, chord-arpeggio, vibrato (four types), pitch-slide-and-portamento (pattern tutorial), multispeed (tables-only design, Rubicon 2×), shuffle-funktempo (funktempo defined, BPM), orderlist (1.6 value ranges), sid-player-routine, sid-format (SID-Maker outputs, SWMconvert), sync (MIDI table), commodore-64 (frame rates, transfer hardware, noise fixes), vice (setup), tracking-workflow, ring-modulation-and-sync.
• Skipped: the key-reference appendices (already in the 1.8 manual summary), the hex conversion table, the BPM tables beyond the formulas.
• Noted: the book prints the $84 filter-switch override as "channel 4" (evidently channel 3); the 1.6 orderlist range $B0–$FD = tempo differs from 1.8's separator range.

2026-08-30 — ingest: Chordian, Composing in SID Factory II part 4 (raw/chordian/…part-4-instruments.txt)#

• Summary: s-chordian-sf2-instruments. Created: chordian, sid-factory-ii. Updated: hard-restart (two ticks, 0F 00), adsr-envelope (examples, note-off behaviour), sid (combined-waveform audibility table, noise lock-up, chip dates), ring-modulation-and-sync (test-bit semantics), wavetable-programming (SF2 snare, offsets vs effects), instrument-design, pulse-width-modulation, instrument-tables.
• Parts 1–3 of the series were not fetched (sequencing/UI only per the search results).

2026-08-30 — ingest: Lemon64 threads on the ADSR bug (raw/lemon64/)#

• Two threads (2008 t=26968, 2015 p=659158), post text extracted with author and timestamp. Summary: s-lemon64-hard-restart-threads.
• Updated: hard-restart ("Why it works" now sourced: Lasse's lock-up explanation, two-frame rule, encore's three strategies, iAN CooG's retrigger sequence), adsr-envelope, sid, ring-modulation-and-sync, lasse-oorni. Removed the corresponding (unverified) items on hard-restart, adsr-envelope, sid.
• Open contradiction for lint: Linus Walleij's page (scouted, not ingested) claims hard restart matters only on the 6581; neither ingested thread makes that restriction.

2026-08-30 — ingest: Newman, "Driving the SID chip" (raw/gamejournal/newman-2017-driving-the-sid-chip.txt)#

• G|A|M|E journal 6/2017, CC BY-NC 4.0. Summary: s-newman-driving-the-sid-chip.
• Created: rob-hubbard, martin-galway, commando, monty-on-the-run, sanxion, hvsc (fills a wanted link), and bob-yannes (shared with the book). Updated: wavetable-programming ("Why it works" + Heard in), channel-interleaving (multiplexing section), chord-arpeggio, fake-echo (Baldwin's single-voice echo), pulse-width-modulation (Galway, Huelsbeck), filter-programming (Daglish, Beach-Head), sid (history, DC-offset samples), sid-player-routine (the classic drivers), chiptune, tracking-workflow, octave-bass, pitch-slide-and-portamento, instrument-design, commodore-64, ring-modulation-and-sync.
• Skipped: the theoretical framing (affordance theory) and the literature review beyond quotations that carry facts.

2026-08-30 — plan: next ingests#

• Wrote inbox/plan-next-ingests-2026-08-30.md (temporary): Game Boy depth (Pan Docs + an LSDj scout) → PC-tracker references (OpenMPT manual, ST3/FT2/ProTracker docs, Schism) → TraxWeekly interviews and remaining ★ → cheap C64 confirmations (Programmer's Reference, HVSC format doc, Walleij) → optional NES/Amiga breadth. Committed state at that point: d6ab26f.

2026-08-30 — scout: LSDj sound design#

• Wrote inbox/scout-lsdj-sound-design-2026-08-30.md: 10 ranked candidates — Defense Mechanism's Intense Tech series (22 LSDj articles, CC BY-NC-SA 4.0: tables, kicks, noise, ADSR, wave synth, grooves, Z, a Kotlinski interview), Infu's Getting started with LSDj (CC BY-NC 4.0), chipmusic.org threads (LSDJ FAQ, Advanced Tricks, Complete Drum Kit on the Noise Channel, Learning Tables, Chord Cheat Sheet), two Kotlinski interviews (2012, 2014/Devise), Sabrepulse's getting-started tutorial, older LSDj manuals for version history — plus lower-priority sightings. Nothing ingested; the owner picks.
• wiki.littlesounddj.com (named in the manual) no longer answers; its Fandom mirror blocks fetchers; chipmusic.org needs a browser user agent (curl works).

2026-08-30 — ingest: Pan Docs, audio section (raw/pandocs/Audio.md, Audio_Registers.md, Audio_details.md)#

• Copied verbatim from github.com/gbdev/pandocs (src/, master, CC0); provenance in raw/pandocs/README.md. One summary per page: s-pandocs-audio-overview, s-pandocs-audio-registers, s-pandocs-audio-details. Created: pan-docs.
• Rewrote game-boy-apu (channel table, register map, timing and formulas, signal path and pops, quirk ↔ LSDj table). Updated game-boy (models and their sound), lsdj (hardware behind the parameters), pulse-width-modulation (four duties, 25 % = 75 %, phase), wavetable-programming (wave RAM facts), adsr-envelope (the hardware ramp, zombie mode), retrigger (what a trigger does), stereo-panning (NR51/NR50, pops), pitch-slide-and-portamento (period formula, resolution, sweep overflow), instrument-design (noise LFSR, sweep, kill clicks), chiptune ("8-bit").
• Resolved (unverified): the duties / wave RAM / LFSR / envelope item on game-boy-apu and the duty item on pulse-width-modulation; the model item on game-boy re-marked (years still unverified). Links between LSDj commands and registers are marked "(inference)". randomization-and-probability left alone — the noise facts belong on instrument-design and game-boy-apu.
• Skipped: the mdBook diagrams ({{#include}}), the HPF C code beyond its constants, the emulator-test details (Prehistorik Man, Duck Tales) beyond a mention.

2026-08-30 — ingest: Intense Tech with Defense Mech, six articles (raw/intense-tech/)#

• Fetched the Markdeep pages of articles 09 (tables), 10–11 (kicks, parts 1–2), 13 (Kotlinski interview), 17 (noise), 18 (ADSR) from defensemech.com (CC BY-NC-SA 4.0); the .md URLs fall back to the site's front page, the .md.html pages carry the text. Summaries: s-intense-tech-09-lets-table-this-discussion, s-intense-tech-10-kicks-part-1, s-intense-tech-11-kicks-part-2, s-intense-tech-13-kotlinski-interview, s-intense-tech-17-the-joys-of-noise, s-intense-tech-18-adsr-makes-life-easier.
• Created: defense-mechanism, intense-tech. Rewrote johan-kotlinski (background, scene history, live mode, milestone versions, roadmap; the "rolemodel" alias and the 2000 date are now sourced).
• Updated lsdj (noise shape cheat sheet, version history table), instrument-design (Defense Mechanism's drum recipe table), wavetable-programming (kick/snare/kick-to-bass tables, Kyoto kick), instrument-tables (volume-column loop, A nesting, A20, even/odd-row quirk), adsr-envelope (8.1.0 ADSR, 8.8.0 software volume, speed chart), pitch-slide-and-portamento (PITCH modes and their history; S/P/C/transpose on noise shapes), retrigger (noise retrig pulses, R00/R01 versions, ADSR tremolo), randomization-and-probability (Z crashes, probability sequencing), chord-arpeggio, shuffle-funktempo (kick-to-bass groove), stereo-panning (even/odd table rows), volume-slides (new LSDj section), chiptune (Kotlinski on 1990s "chip music"), tracking-workflow (finishing songs), game-boy-apu (noise shape ↔ NR43: the LSDj byte is evidently the complement of the register — inference from the two tables; the 1-in-256 mute figure).
• Naming note: the noise instrument's FREE/STABLE setting is called S CMD, then S MODE in the articles (LSDj 8) and PITCH FREE/SAFE in the 9.2.6 manual; treated as the same setting (inference). "Automate" (articles) = TABLE STEP (manual), likewise by inference.
• Skipped: the video-only examples, the Patreon/donation footers, the Spanish translations, embedded SoundCloud/Bandcamp players; save files not fetched.

2026-08-30 — ingest: Intense Tech 01, 02, 06, 07; Infu's LSDj course; eight chipmusic.org threads#

• Intense Tech (raw/intense-tech/): the two wave-synth deep dives (2018) and the two groove articles (2019). Summaries: s-intense-tech-01-wave-synth-deep-dive-part-1, s-intense-tech-02-wave-synth-deep-dive-part-2, s-intense-tech-06-groove-and-tick-tricks-part-1, s-intense-tech-07-groove-and-tick-tricks-part-2. Article 07's 3/4 groove example is a figure missing from the text.
• Infu, Getting started with LSDj (raw/infu/, CC BY-NC 4.0, single page plus a captions-only mini-guide): s-infu-getting-started-with-lsdj; created infu.
• chipmusic.org (raw/chipmusic/, fetched with a browser UA, posts extracted with author/date): LSDJ FAQ (85 posts), Advanced Tricks, Noise snare + Complete Drum Kit on Noise Channel, Chord Cheat Sheet + Learning Tables, and two Kotlinski interviews (2012 school paper; 2014 Devise Magazine). Summaries: s-chipmusic-lsdj-faq, s-chipmusic-lsdj-advanced-tricks, s-chipmusic-noise-drum-threads, s-chipmusic-tables-and-chords-threads, s-chipmusic-kotlinski-interviews; created chipmusic-org.
• Rippled: wavetable-programming (the wave synth in depth: harmonics, sine recipe, PHASE/DIST/LIMIT/VSHIFT), shuffle-funktempo (ticks, Hypnogram's groove numbering, per-channel swing, triplets, table grooves), fake-echo (groove delay), chord-arpeggio (LazierGunz's tables, G01 2/2, variation), instrument-tables (think in ticks, H loops, Infu's snare table), instrument-design (community recipe table: Infu, Mrwimmer, metatronaut, Victory Road, SketchMan3, Kotlinski's bass duty, VFF cymbal), channel-interleaving (new LSDj section: never dedicate a channel), pitch-slide-and-portamento (L on the target note), vibrato (VFF = FM), volume-slides and adsr-envelope (pre-8.8.0 clicks explained, Infu's ENV), stereo-panning (panning clicks, stereo snare trick), sync (half-clocked Game Boys, speed knob, MIDI keyboard quantisation, DMG-07 wiring), tracking-workflow (LSDj habits), song-init-and-looping (silent chains), retrigger, rhythm-and-groove, orderlist, pulse-width-modulation (wave-channel PWM, Kotlinski's duty), game-boy, chiptune, johan-kotlinski (2012/2014 facts: Hack n' Trade, C compiler, Lo-Bat, DMG preference), lsdj (version rows 3.9.2, 3.9.e, 8.5.1/9.2.A), intense-tech, defense-mechanism.
• Skipped: the FAQ's hardware-repair and USB-transfer posts, YouTube links without transcripts, memes; the LSDj wiki links quoted in the threads point to the dead wiki.littlesounddj.com.
• Page candidates noted (3+ mentions, no page yet): Hypnogram (groove numbering, Zenarchy, the delay trick), Zef (3/4 grooves, saves, videos), SketchMan3, Dire Hit, danimal cannon, Toriena, Nanoloop, Musicline Editor. Not created — only one-line facts each so far.

2026-08-30 — ingest: Step 2, PC-tracker references (OpenMPT manual, ST3, FT2, ProTracker, Schism, tracker-history)#

• Raw: raw/openmpt-manual/ (14 wiki chapters as wikitext, CC BY-SA 3.0), raw/scream-tracker-3/ (ST3.DOC from textfiles.com, TECH.DOC from GitHub, CP437 + UTF-8), raw/fasttracker-2/ (the 2.08 manual's OCR text from archive.org; FT2.DOC itself not found), raw/protracker/ (2.3A ReadMe/history, 2.0A effect list and file format, version table, the 2.3d help file edited by echolevel), raw/tracker-history/ (six lineage files), raw/schism-tracker/ (GitHub wiki pages, README, configuration doc). OpenMPT's Manual:_Main_page does not exist; the chapter list came from the MediaWiki API.
• Summaries (13): s-openmpt-manual-effect-reference, s-openmpt-manual-module-formats, s-openmpt-manual-compatible-playback, s-openmpt-manual-zxx-macros, s-openmpt-manual-about-openmpt, s-openmpt-manual-instruments, s-openmpt-manual-samples, s-openmpt-manual-song-properties (with Tempo Swing), s-st3-manual, s-st3-tech-doc, s-ft2-manual, s-protracker-23d-help, s-protracker-23-readme-and-docs, s-tracker-history-notes, s-schism-tracker-wiki.
• Rewritten from stubs: openmpt, scream-tracker-3, fasttracker-2, protracker (all with cheat sheets). Created: schism-tracker, s3m-format, mod-format, xm-format (three wanted pages filled), sami-tammilehto, future-crew, triton, lars-hamre, peter-hanning, olivier-lapicque, storlek, technique opl-fm-instruments, concept effect-command-crosswalk.
• Rippled: chord-arpeggio (MOD/XM 0xy now sourced), stereo-panning, sample-offset, sample-looping, instrument-envelopes, volume-slides, retrigger, vibrato, pitch-slide-and-portamento, new-note-actions (FT2/OpenMPT item sourced), tracker, chip-samples, song-init-and-looping, multisampling, detune (finetune per format sourced), shuffle-funktempo (speed-alternation item sourced; OpenMPT Tempo Swing), fake-echo, filter-programming (new IT/OpenMPT section), randomization-and-probability, chiptune, orderlist, instrument-design, impulse-tracker and it-format (2.14/2.14p5 dates sourced, filter/MMX/source-release items narrowed), jeffrey-lim.
• Resolved (unverified): ST3 author/date/channels, FT2 authors/date/features, ProTracker lineage, OpenMPT lineage/Wine/plugins/compatible playback, MOD/XM 0xy, FT2 no NNAs, permanent detune, IT speed alternation. Still open: the Paula chip itself (amiga-paula stays wanted — the PT docs only imply the hardware), IT's filter version and 2014 source release.
• Skipped: OpenMPT's Patterns/General/Basics/Setup/Note Properties chapters (UI; raw kept) and its ~75 other pages (plugins, dialogs); Schism's Player-abuse-tests (a test catalogue, raw kept); the FT2 manual's CD-dumper and DPMI chapters; ST3's setup and info-screen details beyond the keys.
• Caveat recorded on the summaries: the FT2 manual is OCR text with errors; TECH.DOC is summarised semantically, byte tables stay in raw.

2026-08-30 — ingest: Step 3, TraxWeekly interviews and the remaining ★ articles (five themed batches)#

• Raw: unchanged (raw/traxweekly/); 75 articles extracted for reading with tools/extract-traxweekly-article.py (its targets list now holds the Step-3 set; sub-headed interviews and a few oddly headed issues were cut by explicit line ranges).
• Interviews (16 summaries): s-tw001-interview-basehead, s-tw001-interview-phoenix, s-tw003-interview-five-musicians, s-tw003-interview-chuck-biscuits, s-tw009-interview-maelcum, s-tw010-interview-jugi, s-tw014-interview-pinion, s-tw018-interview-basehead, s-tw027-interview-basehead-lotus-position, s-tw027-interview-zapper, s-tw032-interview-zinc, s-tw041-interview-zinc, s-tw057-interview-catspaw, s-tw080-trax-mass-interview, s-tw105-newbie-interviews, s-tw109-interview-skaven. Created skaven, purple-motion, five-musicians, epinicion, phoenix, jugi; rewrote basehead, necros, kosmic; extended maelcum, pinion, the-zapper, zinc, catspaw, psibelius, leviathan, future-crew.
• Tracking tips / IT (13): s-tw009-samples-lameness-preconceptions, s-tw012-advanced-tracking-tips, s-tw015-advanced-tracking-tips-ii, s-tw018-advanced-tracking-tips-iii, s-tw021-tracking-tips, s-tw032-not-so-advanced-tracking-tips, s-tw033-4channel-lives-on, s-tw044-impulse-tracker-trivia, s-tw094-stereo-samples-in-impulse-tracker, s-tw096-studio-tracking-tips, s-tw098-impulse-tracker-investment, s-tw098-impulse-tracker-midi, s-tw100-tracking-on-impulse. Created lead-articulation, midi-in-trackers; rippled stereo-panning (ST3 Xxx/S8x values, the two Xxx ranges, stereo sample pairs), pitch-slide-and-portamento, vibrato, volume-slides, fake-echo, shuffle-funktempo, effect-command-crosswalk, tracking-workflow, orderlist, channel-interleaving, new-note-actions, impulse-tracker, jeffrey-lim, fasttracker-2, instrument-design, tracker, scream-tracker-3.
• Theory / composing (6): s-tw007-the-importance-of-music-theory, s-tw046-serial-composition, s-tw050-aesthetics-of-composition, s-tw072-composing-whole-modules, s-tw087-music-theory-a-modern-approach (the Modsquad's seven parts #87–#97 are satire — one combined summary quarantines the jokes and keeps the drum-programming and form ideas), s-tw107-laymans-music-theory. Created serial-composition; rippled semitone-math, scales-and-modes, diatonic-chords, chord-progressions, melody-writing, rhythm-and-groove, arrangement, the-modsquad.
• Samples / mixing / realism (5): s-tw024-sample-ripping-debate (#24, #25, #64, #65; TW.025 is a re-issue of #25), s-tw066-sound-recording, s-tw070-realism-debate (#70–#73, continued by the already ingested #76), s-tw118-reverb-and-compression, s-tw119-reverb-and-audio-gear. Created sample-ripping, realism-in-tracked-music, reverb-and-compression, recording-samples; rippled multisampling, sample-looping, sample-offset.
• Scene / history / business (12): s-tw011-how-to-enhance-your-music-collection, s-tw032-20-minute-chip-compo (#26–#33, #50), s-tw032-tracker-survey (#32–#34), s-tw044-hornet-ratings (#43–#45), s-tw051-how-to-listen-to-tracked-music, s-tw064-potted-tracking-history, s-tw065-tracking-as-an-olympic-sport (#65–#66), s-tw099-why-tracking-is-the-perfect-artform, s-tw114-live-tracking, s-tw114-commercial-tracking-debate (#113–#118), s-tw115-tracking-semantics, s-tw118-tracking-the-format. Created tracker-scene-history, compos, commercial-tracking, live-tracking; rewrote 20-minute-chip-compo; extended hornet, traxweekly, chiptune.
• Resolved (unverified): Necros = Andrew "Andy" Sega, Five Musicians co-founder (#3, #73 signature, #80); Basehead's FM membership confirmed — the old line-up guess "Zodiak and Wave" was wrong (FM 1995: Necros, Mellow-D, Big Jim, Basehead, Purple Motion); Basehead's "real name Dan Nicholson" was wrong — three sources print Dan Grandpre; Kosmic founded fall 1992 (not 1991); Hornet's ftp.cdrom.com home sourced. Still open: Necros's later career, Maelcum's surname, Kosmic's end, Hornet's closing year, IT's filter version / 2014 source release.
• Contradictions recorded on pages: Pulse's "IT213 will be the very final public release" (May 1997) vs 2.14 (September 1997) on impulse-tracker/jeffrey-lim; Xxx pan centre X40 (DMP era) vs X80 (IT) on stereo-panning; Behemoth vs brn on sampling (Nyquist) on realism-in-tracked-music; Future Assassin's "13 years of theory" for Necros vs his own "five years informal" (#7 vs #3) on necros. TOC: the header of #28 carries a wrong date (28 October 1995 between 21 September and 6 October).
• Skipped (marked in the TOC): #12 4 Impulse 1995 (party announcement), #14 5 Trax Tutorial Update (announcement), #14 7 Bass Productions and #110 7 Blitz Beat Productions (group member lists), #24 5 Food for Tracking (a food recipe). Read but not summarised: Lord Pegasus's flame appended to the Pinion interview, Dilvish's "The Power of Headphones" (#118 2).
• Cross-batch ripples: each batch wrote inbox/tw-batch-<a–e>-notes.md for pages it did not own; three integration passes applied them (people/scene/theory/sampling pages; the technique and tool pages; then batch B's notes) — 33 + 13 + 24 page updates, incl. traxweekly (interviewers by era, debates, columnists), tracking-workflow ("How the 1995 scene started a song"), scales-and-modes (Hartrell's step formulas), chord-inversions (Necros's voicing table), protracker, mod-format; the notes files were then deleted. Also created the-grey-note (song) and gravis-ultrasound (platform; 18 sourced facts), and s-tw024-in-review-asm95 (Necros's ASM95 review column: Skaven's "Catch that Goblin" won multichannel, Zapper's "Distance" disqualified for extended samples).
• Not applied from the notes: Zinc's Demotape Directory item (no summary supports it), a Music Contest event page (Sturm's remark lives on realism-in-tracked-music), the Modsquad ring-modulator aside.
• Page candidates (3+ mentions, no page): Mellow-D, Big Jim, Acidfrog (interview in #44), Aahz, Radiance, Force Ten, Complex, Second Reality, Remedy, AmusiC (Sotiris Varotsis), Behemoth (David Menkes), Snowman (Christopher G. Mann), Popcorn (Chris Campbell), Wave, NOISE, Deck Blasters, Fencer, Perisoft (interview in #87), Cool Edit / Sound Forge.
• TOC: 123 rows now link a summary; every ★ row is ingested or marked skipped (backlog 0, was 63). tools/triage-traxweekly.py re-run; links preserved. Open for later: the remaining interviews (Leinad, Nemesis & Stalker, TheHacker, Ryan Cramer, Blaze Runner, Big Jim, Claim, SV, Populus, AmusiC, Lord Pegasus, Quarex, Jase, the Faces-in-the-Crowd series, Loki, Blackwolf, Zalt, Mick Rippon, Shawn, Balrog, Pariah, Acidfrog, VadimVS, Daedalus, Ganja Man, Saxy, Quantam Porcupine, Scirocco, Injekted and Subhuman, Clef, Rage, Vegas, Ryan, Liam the Lemming, U4IA, Perisoft, Cosmic/RR, Mikmak, Khyron, Mental Floss) and the reviews.

2026-08-30 — ingest: Step 4, cheap C64 confirmations (PRG ch. 4, 6581 datasheet, HVSC SID format, Walleij, chipmusic tutorial, SID-Wizard 1.6 charts)#

• Raw: raw/c64-prg/ (PRG chapter 4 from devili.iki.fi; page 188 and Appendix O are not transcribed there), raw/mos-6581-datasheet/ (the datasheet PDF from 6502.org + PDFKit text — it is what Appendix O reprints, so it stands in for it), raw/hvsc/SID_file_format.txt, raw/walleij/ (sidmidi.html verbatim + a text extract), raw/chipmusic/c64-music-for-dummies-t8104.txt (40 posts, browser UA), raw/sid-wizard/Charts-and-Tables-for-SID-Wizard-1.6.{pdf,txt} (owner's local copy; the document's byline is Mikael Norrgård, not Hermit — manifest corrected).
• Summaries (6): s-c64-prg-chapter-4-sound, s-mos-6581-datasheet, s-hvsc-sid-file-format, s-walleij-sid-player-routine, s-chipmusic-c64-music-for-dummies, s-sid-wizard-charts-and-tables.
• Created: linus-walleij (person), c64-programmers-reference-guide (publication).
• Rippled: sid (datasheet register map with formulas, readback/pots/ext-in section, model differences), adsr-envelope (the timing table now cited to the datasheet, the PRG and the charts, with the 1 MHz-clock caveat; sustain semantics; "gate at any point"; Appendix B and PRG presets), ring-modulation-and-sync (datasheet chain, BASIC examples, the PRG "bit 3" slip), filter-programming (register section, Walleij's macro), pulse-width-modulation (register section, 1.0 tutorial numbers, pulse-width vibrato), hard-restart (Walleij's recipe, 1.0 HRTY, the 8580 contradiction), sid-player-routine (Walleij's pseudo-code, PRG/datasheet note tables), vibrato (oscillator-3 vibrato, Walleij's delayed LFO), sync (the MIDI catch-22), chord-arpeggio (0 3 7 from the player side), sid-format (PSID/RSID header table), hvsc (DOCUMENTS folder and authors), commodore-64 (POKE model, clock dependence, chip per model), vice (engines, key equivalents), tracking-workflow (the beginner's first hour), instrument-design (1.0 beginner set, datasheet/PRG presets), instrument-tables (charts' refinements), sid-wizard (1.0 facts, 1.6→1.8 effect table, extra keys, SDI layout).
• Resolved (unverified): sid-format's PSID/RSID description. Still open: VICE name/x64sc, HVSC Songlengths, the Drean C64, PWM sample-cycling, chord-arpeggio's fusion inference.
• Contradictions recorded on pages: Walleij's "hard restart only matters on the 6581 (or so I am told)" vs the Lemon64 threads, the SID-Wizard book, HVSC's generic wording and every tracker's practice — wiki position: apply it on both revisions until someone tests (on hard-restart); the PRG's "bit 3" for ring modulation and "logical ANDing" for sync vs the datasheet (on ring-modulation-and-sync, c64-programmers-reference-guide); Walleij's "the numbers in the C64 hardware reference manual are simply WRONG" (table unstated); the 1.6 charts vs the 1.8 manual ($04 ≥ $F0 vs $09 for the WF-ARP jump, $09 listed twice, $0F shift vs set, fast-forward 4× vs 6×, filter band $A missing from the charts) on sid-wizard / filter-programming; the PRG's frequency constant .06097 vs the datasheet's 0.0596 explained as NTSC vs 1 MHz clock (marked derived).
• Skipped: the datasheet's electrical/timing/pin tables and its 96-row note table, the PRG program listings verbatim, the charts' BPM and hex tables (all in raw); the tutorial thread's unanswered SD2IEC/SwinSID questions. Raw caveats: page-footer text appended to three chipmusic posts; the Walleij text extract lost a few pseudo-code lines (the HTML sibling is complete).
• Page candidates: Dag Lem (reSID), Michael Kleps, ant1, 4mat — single mentions each.

2026-08-30 — ingest: the remaining Intense Tech articles (03, 04, 05, 08, 12, 14, 15, 16, 19, 20, 21) and Sabrepulse's Getting started with LSDJ#

• Raw: 12 new files in raw/intense-tech/ — the 11 unread articles plus 22-wave-tech.md, an unlisted, unfinished draft ("Posted 2021", an outline of bullet points only) kept for completeness and not summarised; the series therefore has 21 published articles, not the 22 the scout counted. raw/sabrepulse/sabrepulse-getting-started-with-lsdj.txt — Adventure Kid's mirror of the tutorial (sabrepulse.com is offline; undated, pre-8.1 ENVELOPE field, no licence stated). Both recorded in raw/sources.md. Sabrepulse's tutorial had been an open scout candidate (#7), never ingested.
• Summaries (12): s-intense-tech-03-dont-sleep-on-z, s-intense-tech-04-liblsdj, s-intense-tech-05-wave-cruncher-instrument-library, s-intense-tech-08-dotcnts-wave-cruncher, s-intense-tech-12-scoping-out-new-features, s-intense-tech-14-lets-appreciate-version-8, s-intense-tech-15-lets-mix-it-up-and-down, s-intense-tech-16-cartridge-family, s-intense-tech-19-new-noise-and-910-news, s-intense-tech-20-lsdpatch, s-intense-tech-21-whats-new-in-lsdj-92, s-sabrepulse-getting-started-with-lsdj.
• Created (6): hypnogram, 4ntler, sabrepulse (people); liblsdj, lsdj-wave-cruncher, lsdpatch (tools, each with a cheat sheet — the viewer now shows 12 cheat sheets).
• Rippled (28 pages): lsdj (files/tools/cartridges/emulators section, the 9.1.0 ordered-noise layout with the old→new conversion rule read from the converter's table, command rows for B, C, P, R, V, Z, version history 7.7.4 → 9.2.H, screen-map note), intense-tech (all 21 rows filled, the draft noted), defense-mechanism (GitHub handles, psgcabal libraries, more tracks); three parallel passes — A: randomization-and-probability, vibrato, pulse-width-modulation, stereo-panning, chord-arpeggio, pitch-slide-and-portamento; B: adsr-envelope, instrument-tables, retrigger, volume-slides, arrangement, instrument-design (Sabrepulse's beginner instrument set, the Joys-of-Noise kick converted to 9.1.0 note names), channel-interleaving (kick/bass sharing PU1); C: wavetable-programming (version 7–9.2 wave changes, LIMIT as volume, the crunch workflow), game-boy (flash cartridges, emulators), game-boy-apu (silky wave, wave-RAM refresh bug vs LSDPatch blending, 9.2 software envelopes), johan-kotlinski, tracking-workflow (first-song path, save-file backups), chip-samples (crunched wavetables and kits); then from the passes' notes: detune (FINETUNE versions), shuffle-funktempo and rhythm-and-groove (Z on G since version 8).
• Resolved: johan-kotlinski's (unverified) "distributed as a ROM image" → confirmed (articles 16, 20; content-page markers 21 → 20 — lint's total of 29 includes this log's own mentions); lsdj's "PITCH FREE/SAFE … by inference" → sourced (9.1.0 removed FREE/STABLE, 9.1.3 added PITCH FREE/SAFE, article 21); instrument-tables' "Automate (STEP, by inference)" → sourced (version 8 renaming, article 14).
• Contradictions recorded on pages: Z with H/G/D — the version-6 manual "not with Hop, Groove, Delay" vs article 03's added note "works as of 8.1.9" vs article 14 and the 9.2.6 manual "every command except H" (randomization-and-probability, lsdj version history); noise P speed — 9.2.6 manual "applies S every tick" vs article 19's 9.1.0 scale P01 = S01 every 4 ticks, P04 every tick (pitch-slide-and-portamento, lsdj); Infu's "LIMIT above 0F rarely helps" vs Defense Mechanism's deliberate LIMIT overdrive (wavetable-programming); noise mutes inside one loop mode (9.1.3 rationale) have no hardware explanation in Pan Docs, whose LFSR lock-up covers only long→short (game-boy-apu).
• Skipped: article 22 (draft outline); the noise converter's 256-entry JavaScript table beyond the rule and anchor values; cartridge prices/shop links beyond the summary; the tutorial's comment-form boilerplate. Note: article 16's struck-through EZ Flash Jr instructions are summarised as withdrawn.
• Page candidates (mentioned, no page): DOTCNT/Ada (two articles; described on lsdj-wave-cruncher), Pain Perdu, Lemondrop, insideGadgets and Krikzz (on game-boy), Adventure Kid (mirror host), the emulators BGB / SameBoy / Gambatte / Emulicious (on game-boy and lsdj).
• Inbox: scout-lsdj-sound-design-2026-08-30.md status updated (only candidate 9, the old manuals, remains); plan-next-ingests-2026-08-30.md step 1 status updated; earlier today the SID scout's stale status line (#5, #6, #7 were already ingested in step 4) was corrected. tools/lint.py: 276 pages, 0 broken links, 0 orphans, all indexed; tools/build-site.py rebuilt site/. Not committed.

2026-08-30 — ingest: the SID scout's leftovers (Lemon64 ×2, Furnace docs, chipmusic.org ×2, uCApps, commodore-64.eu, Remix64, Hugi #38, sidmusic.org ×7)#

• Raw: raw/lemon64/ (+2 threads: "How does one learn SID sound design these days?" 2020, "SID-Wizard instrument recommendations?" 2021), raw/chipmusic/ (+2: "typical C64/SID sounds" 2015, "sid music hints'n'tips" 2013), raw/furnace/ (7-systems/c64.md, 4-instrument/c64.md with commit dates in README.md), raw/ucapps/ (Klose's Wavetable Sounds Tutorial #1), raw/commodore-64-eu/ (enthusiast overview, secondary), raw/remix64/ (Neil Carr's Galway interview, 2001-03-28), raw/hugi/ (Magic's Jeroen Tel interview, Hugi #38, early 2014), raw/sidmusic/ (Andreas Varga's SID Homepage: the August 1996 Bob Yannes e-mail interview, the Commodore Zone / 64'er / Happy Computer interview pages for Tel, Galway and Hubbard, "The creation of the SID chip", Alstrup's triangle and noise waveform examinations). sidmusic.org's root redirects to a SIDPLAY page but its sid/ pages serve; sidtech1–3 (the Appendix O transcription = the datasheet already in raw) and patent.html were not saved. All recorded in raw/sources.md. New helper: tools/extract-forum-posts.py (phpBB/PunBB post extraction with author and date; generic article mode).
• Summaries (15): s-lemon64-learning-sid-sound-design, s-lemon64-sid-wizard-instrument-recommendations, s-furnace-c64-docs, s-chipmusic-typical-sid-sounds, s-chipmusic-sid-music-hints-n-tips, s-ucapps-wavetable-sounds-tutorial-1, s-commodore-64-eu-sid-chip, s-sidmusic-alstrup-waveform-examinations, s-sidmusic-yannes-interview, s-sidmusic-creation-of-the-sid-chip, s-sidmusic-hubbard-interviews, s-sidmusic-galway-interviews, s-sidmusic-jeroen-tel-interviews, s-remix64-galway-interview, s-hugi38-interview-jeroen-tel.
• Created (10): furnace (tool, cheat sheet), siddump (tool), asger-alstrup, jeroen-tel, charles-deenen, andreas-varga (people), maniacs-of-noise (group), commodore-zone, sid-homepage (publications), helikopter-jagd (song).
• Rippled (two ingest agents, then an integration pass over each other's notes): sid (the designer's account of every block, Alstrup's $D41B measurements and the 23-bit noise LFSR, 4mat's "metal noise" control-byte table, a Contradictions section), adsr-envelope ("how the silicon does it", the 30 ms gate delay, Monty's 08F8, Galway's "school band effect"), hard-restart (Furnace's envelope reset, what Yannes does and does not explain, "before hard restart existed"), filter-programming, pulse-width-modulation, ring-modulation-and-sync, chord-arpeggio (### Furnace sections; the first PWM sweeps), wavetable-programming (the Auf Wiedersehen Monty bass drum frame by frame with a GoatTracker translation, marked derived), instrument-design ("SID recipes from the forums", the composers' views), instrument-tables, chip-samples (the digi from Yannes/Galway/Hubbard/Tel), multispeed, randomization-and-probability, channel-interleaving, fake-echo, arrangement, tracking-workflow (learning SID sound design, dissecting with siddump, how the C64 composers worked), sid-player-routine (Hubbard/Galway/Tel toolchains), sid-wizard, goattracker, vice (3.1 → 3.5), lasse-oorni, hvsc, chiptune, commodore-64, rob-hubbard, martin-galway, bob-yannes (rewritten), commando, monty-on-the-run, sanxion, c64-programmers-reference-guide, chordian, chipmusic-org, sid-format, detune, pitch-slide-and-portamento, hermit, witchmaster.
• Resolved: hvsc's sidplay claim; sid's history facts (1981 start, Ensoniq) are now primary-sourced. Still open: hard restart on the 8580 (no new source tests it), the Drean C64, VICE's name, the multispeed 4×–8× tip; new: "Hal Charpentier" as Galway spelt it (martin-galway), siddump's authorship (siddump).
• Contradictions recorded on pages: AND-ed waveforms — Rindeblad's column calls it a documentation error, Yannes and the datasheet say it is how combining works (sid, bob-yannes); "32 voices" — Newman's paraphrase vs Yannes's own words (Ensoniq's later figure); SID design year 1981 vs 1982; 8580 year 1986 vs 1987; OSC3 readback "never used" (Furnace, Yannes) vs the PRG's examples and Alstrup's measurements; the test bit's delayed noise reset (Alstrup: $2000–$8000 cycles later) vs the datasheet's "also reset"; Furnace's ring-modulation wording vs the datasheet's triangle-only condition; Hubbard's assembler (his own account) vs Newman's monitor.
• Skipped: the TR-808 kick (an image), MeneerJansen's pastebin notes, the SID patent and Appendix O duplicates, MIDIbox-specific CC numbers and firmware, commodore-64.eu's shop/FAQ blocks, the Hugi video list, the interviews' small talk (favourite babes, food, TV, the Dave Collier rumour), Alstrup's full assembler listings.
• Page candidates: auf-wiedersehen-monty, sidviz, sidplay, lemon64 (publication), midibox-sid, electronic-transfer (Gallefoss); songs Crazy Comets, Thing on a Spring, Arkanoid, Parallax, Times of Lore, Turbo Outrun; people David Whittaker, Ben Daglish, Tim Follin, Fred Gray, Barry Leitch, DRAX / Laxity / Mitch & Dane.

2026-08-30 — ingest: the 56 non-★ TraxWeekly interviews (five batches A–E, then three integration passes)#

• Raw unchanged (raw/traxweekly/); tools/extract-traxweekly-article.py's targets list extended by the 55 interview rows; 17 sub-headed pieces cut by line range (#4, #5, #8, #11, #15 SV, #22, #29, #30, #31, #32, #34, #54 7, #90); #7's extract regenerated (the tool had caught only the section header).
• Summaries (56): s-tw002-interview-leinad, s-tw004-interview-nemesis, s-tw004-interview-stalker, s-tw005-interview-thehacker, s-tw006-naid-opinion-interviews, s-tw007-interview-ryan-cramer, s-tw008-interview-blaze-runner, s-tw011-interview-perisoft, s-tw011-interview-beaner, s-tw013-interview-big-jim, s-tw015-interview-claim, s-tw015-interview-sv (a joke piece; note only), s-tw016-interview-populus, s-tw017-interview-amusic, s-tw018-interview-lord-pegasus, s-tw019-interview-quarex (facts only from a 7,000-word log), s-tw022-interview-jase, s-tw026-faces-in-the-crowd-intro (the series' charter, nobody interviewed), s-tw027-faces-in-the-crowd-master-of-darkness, s-tw028-faces-in-the-crowd-mesonyx, s-tw029-faces-in-the-crowd-smeghead, s-tw029-faces-in-the-crowd-esper-division, s-tw030-interview-ender, s-tw030-interview-paganus, s-tw031-interview-luv-kohli, s-tw031-interview-vicious, s-tw031-interview-fred, s-tw032-fitc-complaints (an article, not an interview), s-tw034-interview-deus-ex, s-tw035-interview-loki, s-tw040-interview-blackwolf, s-tw040-interview-zalt, s-tw041-interview-mick-rippon, s-tw041-interview-shawnm, s-tw042-interview-balrog, s-tw043-interview-pariah, s-tw044-interview-acidfrog, s-tw050-interview-vadimvs, s-tw052-interview-daedalus, s-tw052-interview-ganja-man, s-tw054-interview-saxy, s-tw054-interview-quantam-porcupine, s-tw054-interview-scirocco, s-tw054-interview-injekted-and-subhuman, s-tw055-interview-clef, s-tw060-interview-rage, s-tw060-interview-fisherman, s-tw062-interview-vegas, s-tw063-interview-ryan-sprott, s-tw066-interview-liam-the-lemming, s-tw067-interview-u4ia, s-tw087-interview-perisoft, s-tw090-interview-cosmic, s-tw103-interview-mikmak, s-tw111-interview-khyron, s-tw119-interview-mental-floss.
• Created (66 entity pages): 59 people — leinad, nemesis, stalker, thehacker, island-of-reil, mosaic, ryan-cramer, blaze-runner, perisoft, beaner, big-jim, claim, populus, amusic, lord-pegasus, quarex, jase, kal-zakath, master-of-darkness, mesonyx, smeghead, ender, paganus, luv-kohli, vicious, fred-ink, deus-ex, loki, blackwolf, zalt, mick-rippon, shawnm, balrog, pariah, acidfrog, vadimvs, daedalus, ganja-man, saxy, quantam-porcupine, scirocco, injekted, subhuman, clef, rage, fisherman, vegas, ryan-sprott, liam-the-lemming, u4ia, cosmic, mikmak, khyron, mental-floss, and from the integration pass popcorn, atlantic, dennisc, mellow-d, behemoth (each built from 3+ sourced facts across the summaries); 7 groups — defiance, esper-division, ultrabeat, noise-group, lok, renaissance (the red link from eight pages, filled), aim-higher. The wanted big-jim, acidfrog, amusic, perisoft and popcorn pages now exist.
• Integration (the batches wrote notes for pages they did not own; three agents applied them over disjoint page sets, plus the SID pass for tracking-workflow/chiptune/commodore-64): scene and concept pages — traxweekly (founding by Populus and Neurosis, Popcorn's dated rise, a full account of "Faces in the Crowd": charter, crew and roles, the repeated interview numbers, the 1996–98 interviewers), hornet, kosmic, epinicion, five-musicians, future-crew, triton, the-grey-note, 20-minute-chip-compo, tracker-scene-history (new Regions section), compos (NAID 1995/1996, Music Contest 2–4, Oz96, The Trackering), commercial-tracking, live-tracking, realism-in-tracked-music, sample-ripping, recording-samples; people pages — 38 updated (psibelius, zinc and maelcum as interviewers; necros, basehead, pinion, leviathan, phoenix, skaven, purple-motion, catspaw, the-zapper, jugi, jeffrey-lim, olivier-lapicque, and the new pages that received sibling notes); tool and technique pages — 28 updated (scream-tracker-3, fasttracker-2, impulse-tracker, protracker, openmpt with "What the 1995–98 interviews say" sections and the pre-ST3 tools Composer 669, MultiTracker/MMEDIT, ModEdit, KingMod, Whacker, Tetra Compositor, X-Tracker; gravis-ultrasound memory ceilings and "GUS ACE cannot record"; the formats; midi-in-trackers, stereo-panning, volume-slides, lead-articulation, retrigger (Vegas's three tips), sample-looping (loop/riff etiquette), chord-progressions, arrangement, sync, serial-composition, sample-offset, rhythm-and-groove, new-note-actions, instrument-envelopes, channel-interleaving, chip-samples, instrument-design); tracking-workflow gained "How the 1995–98 interviewees worked".
• Corrections applied: basehead — "Necros defended his NAID compo entry" → Basehead's entry; s-tw032-interview-zinc — "DennisC's tenth interview" → FITC's series number (the same number is printed on Luv Kohli's #31 interview).
• Contradictions recorded on pages: who ran TraxWeekly in spring 1995 (Populus → Popcorn → Psibelius, dated); Radiance's founders (Zinc: Jake Blues and Acidfrog; Loki: Zinc and imager); Bass Productions' head (Psibelius "coordinated" vs AmusiC's "Sirrus has the final word"); Rage's "beat Lizardking" at Assembly '95 (multichannel) vs Lizardking 3rd in the four-channel compo; Vegas's NAID '96 entry with two titles; Kosmic "only 13 members" in late 1994; Defiance's end (Perisoft "died soon after NAID '95" vs the June 1995 revival); Perisoft's "120 channels" vs IT's 64 (+256 virtual); the 1996 claim that NNAs degrade GUS quality; ST 3.0 "buggy" vs 3.20/3.21 crashes; three Ryans (Cramer #7, Hunt = Pinion #14, Sprott #63); Ender (#30) ≠ the "Ender" in Pariah's #43 log; Floss (20mc column) ≠ Mental Floss; "Quantam" (printed) vs "Quantum" (his own spelling).
• TOC: 184 rows now link a summary or are marked skipped (was 123); every interview row is done. Notes in the summary column: #15 8 SV note-only; #26 4 series introduction; #28 1's date should read 28 September 1995 (the header prints 28 October); #34 2 Song Reviews skipped (reviews are out of scope for this pass); #52 3's interviewer is Zinc, not Daedalus. Identifications: "Shawn" (#41) = shawnm, author of #46 Serial Composition; Ganja Man wrote #64's Potted Tracking History; Faces in the Crowd #60 = fisherman; #63's Ryan = Ryan Sprott of Grey; #29's DennisC header carries Master of Darkness's e-mail (a slip in the original).
• Skipped: IRC banter, greeting lists, printed addresses and phone numbers; the three reviews in #34 and the review boxes inside the #30/#31 extracts; the Zinc interview and Atlantic's tips at the ends of the #32 extract (already ingested); not ingested but noted as related: "Khyron Unleashed" (#30 4), Mental Floss's letters (#66 2, #103 6), Necros's In Review columns — these and the reviews remain open.
• Not applied from the notes: batch C's raw-check remark on #6's signature (not a wiki claim); Kal Zakath's "finland or sweden" line (ambiguous); NAID 1996 jury rules per Mr. Khan.
• Page candidates (3+ mentions, no page): C.C.Catch (~8, slug ambiguous), Kleitus (three interviews), LizardKing (7), Zodiak (6–7), Sirrus (4), Chuck Biscuits (7 files, summary only), NAID (event, 14+), Music Contest MC2–MC5 (event, 10+), ModEdit and MultiTracker/MTM (tools, 5+ each), Mazurka, VSL, Chill Productions, Radiance, Neophyte, The Sound Barrier BBS; songs Ascent of the Cloud Eagle, Unreal, Spartacus.
• Housekeeping: wiki/index.md +130 lines (all new pages and summaries; inbox statuses); inbox/scout-sid-sound-design-2026-08-30.md and inbox/plan-next-ingests-2026-08-30.md statuses updated (steps 1–4 complete; only optional step 5 remains — the plan file can go); the seven notes files deleted; tools/lint.py: 423 pages, 0 broken links, 0 orphans, all indexed, 42 (unverified) markers across 33 content pages — 43 counting this log entry (up from 29/21: the new people pages carry ## To verify items — Bass vs Sirrus, Acidfrog's "fred", the unidentified Floss, Mach One, DMF = X-Tracker's format, libmikmod, siddump's author); tools/build-site.py rebuilt site/. Not committed.

2026-08-30 — scout: C64 legends (Chris Hülsbeck, Matt Gray, Jeroen Tel)#

• Wrote inbox/scout-c64-legends-2026-08-30.md: 15 ranked candidates in three groups plus cross-cutting references and lower-priority sightings. Nothing ingested; the owner picks.
• Hülsbeck: the 64'er 10/1986 Soundmonitor article ("Musik wie noch nie", Listing des Monats) survives as an iDOC etext (cbm.6502.org) and an archive.org scan — it documents the Musicmaster driver, the track/step table, bars at $B000, and the 24-byte sound definitions; VGMPF's timeline (Musicmaster 1985 → Soundmonitor summer 1986 → The Final Musicplayer July 1987 → C64 TFMX summer 1988, never released); Remix64 (2001) and EmuWiki (2011) interviews.
• Matt Gray: SIDin #2 (2002-11-21, Stefano Tognon) contains a 30-page commented reverse-engineering of the Driller routine — the "documented driver"; Gray's own Dominator subtune 6 source ("PLAYER V4.2 (C)1987", CC BY-NC 4.0, released 2014-11-26 with 64tass) is uncommented but complete; the c64.com interview (2014-11-28) explains his sequencer approach and vibrato.
• Jeroen Tel: Charles Deenen's "MUSICFILE V01-07-1988" driver exists as source on GitHub (realdmx/c64_6581_sid_players, Cybernoid II with Deenen's routine labels) and on screen in Tel's two 2014 "behind the notes" videos; no prose documentation found — the Deenen interviews (Domination #11, Designing Sound 2010) and VGMPF describe how the group worked in it.
• Blocked or gone on 2026-08-30: Forum64 (Cloudflare), Exotica (browser check), namelessalgorithm Soundmonitor page (404), SIDin's original site (410; mirrored at commodore.software), justsolve (refused), Wayback Machine (offline). Downloads used for verification stayed in the session scratchpad, not in raw/.

2026-08-30 — ingest: C64 legends, first batch (Hülsbeck / Matt Gray / Maniacs of Noise)#

• Ingested the first batch of inbox/scout-c64-legends-2026-08-30.md: #1 the 64'er 10/1986 Soundmonitor article (iDOC etext), #2 VGMPF's Hülsbeck, Soundmonitor, TFMX Editor and Giana Sisters pages, #3 Remix64 (2001) and #4 Retrogaming Planet/EmuWiki (2011) Hülsbeck interviews, #6 SIDin #2's "Matt Gray's Driller music routine", #7 Gray's Dominator subtune 6 source (PLAYER V4.2, CC BY-NC) with the Remix64 competition article and chipmusic.org thread, #8 the C64.COM Gray interview, VGMPF's Matt Gray page, #11 the realdmx GitHub collection (Cybernoid II in MUSICFILE V01-07-1988, the MoN SFX player, the Driller listing in ACME), #13 VGMPF's MoN/Tel/Deenen pages + the C64-Wiki roster, #14 the Domination #11 (1998) and Designing Sound (2010) Deenen interviews, and #12 the two "behind the notes" videos as metadata with the Lemon64 thread.
• Raw: 13 new collections/files under raw/64er/, raw/vgmpf/, raw/remix64/ (2 files added), raw/emuwiki/, raw/sidin/ (PDF + lossy text extraction), raw/matt-gray/, raw/c64com/, raw/github-realdmx/, raw/recollection/, raw/designingsound/, raw/c64-wiki/, raw/lemon64/ and raw/chipmusic/ (1 file each added); all recorded in raw/sources.md. Web pages saved as extracted text; the videos are not saved.
• Summaries (12): s-64er-soundmonitor-article, s-vgmpf-chris-huelsbeck, s-vgmpf-soundmonitor-tfmx, s-remix64-huelsbeck-interview, s-emuwiki-huelsbeck-interview, s-sidin02-matt-gray-driller, s-matt-gray-dominator-source, s-c64com-matt-gray-interview, s-vgmpf-matt-gray, s-realdmx-sid-players, s-vgmpf-maniacs-of-noise, s-charles-deenen-interviews, s-cybernoid-behind-the-notes.
• New pages (21): people chris-huelsbeck, matt-gray, reyn-ouwehand; tools soundmonitor (cheat sheet: keys, 24 sound registers, recipes), tfmx, matt-gray-player (pattern bytes, instrument tables, PLEX, drum table), musicfile (MoN driver: constants, routines, instrument rows, drum/arp/filter/pulse tables), turbo-assembler; songs shades, great-giana-sisters, driller, last-ninja-2, dominator, cybernoid, cybernoid-ii; publications 64er, sidin; groups rainbow-arts, system-3, hewson, codemasters.
• Updated (20): jeroen-tel (MoN driver and Hewson years), maniacs-of-noise (founding, driver, business, dated roster, the 1990 split), charles-deenen (own words 1998/2010, driver authorship), rob-hubbard, martin-galway; sid-player-routine (comparison table of the three documented drivers + what their authors said), tracker (Soundmonitor as the pre-tracker), tracker-scene-history (C64 prehistory 1985–88), orderlist, instrument-tables; techniques chip-samples, channel-interleaving, hard-restart, instrument-design, chord-arpeggio, vibrato, wavetable-programming, filter-programming, pulse-width-modulation, pitch-slide-and-portamento; entity sid.
• Contradictions/unverified left for lint: VGMPF's Deenen page says he "used Jeroen Tel's sound driver" while the MoN page and Deenen himself say the driver was Deenen's (noted on charles-deenen); VGMPF gives Tel's birthplace as Eindhoven vs Helmond in the Hugi biography and a Probe start of January 1988 vs the 1991 move in the Commodore Zone interview (noted on jeroen-tel); readings of the uncommented MUSICFILE tables (starttabel/startlen as first-frame waveform, sequence bytes ≥ $80 as commands) and of the undocumented $AF/$DF/$EF/$97/$E7 bytes in Gray's V4.2 bars are marked unverified; the forum claim that Future Composer derives from a ripped MoN player is marked unverified.
• Index: 21 pages added under Tools/People/Groups/Publications/Songs and a new summaries group "C64 legends — Hülsbeck, Matt Gray, Maniacs of Noise". Not committed.

2026-08-30 — ingest: C64 legends, second batch (the rest of the scout)#

• Ingested the remaining candidates of inbox/scout-c64-legends-2026-08-30.md: #5 SIDwave's The Brief History of SID (Vandalism News #64, 2015), #9 the four shorter Matt Gray interviews (Vandalism News #63 2014, RVG 2016, Retro Gamesmaster, Arcade Attack 2020/21), #10 Remix64's Last Ninja 2 vs Ninja Remix, the cross-cutting SIDId list (cfg + nfo) and the Soundmonitor lineage (tracker-history + C64-Wiki.de + de.wikipedia), and from "also seen" the CSDb "Huelsbeck TFMX?" thread, the G.A.N.G. interview, Chordian's SID Musicians, Wikipedia's Hülsbeck article and the CSDb release pages for Soundmonitor V1.0 and Shades. Skipped: #15 The Retro Hour EP496 (audio only, no transcript), the Forum64 and Exotica pages (fetch-blocked), the Kickstarter updates (403) and the two Lemon64 Deenen threads (no technical content).
• Raw: 17 files in raw/recollection/ (+2), raw/rvg/, raw/retrogamesmaster/, raw/arcadeattack/, raw/remix64/ (+1), raw/csdb/, raw/audiogang/, raw/chordian/ (+1), raw/sidid/, raw/tracker-history/ (+1), raw/c64-wiki/ (+1), raw/wikipedia/; recorded in raw/sources.md.
• Summaries (13): s-recollection-brief-history-of-sid, s-vandalism63-matt-gray-interview, s-rvg-matt-gray-interview, s-retrogamesmaster-matt-gray-interview, s-arcadeattack-matt-gray-interview, s-remix64-ninja-remix, s-csdb-huelsbeck-tfmx-thread, s-csdb-soundmonitor-and-shades, s-audiogang-huelsbeck-interview, s-chordian-sid-musicians, s-sidid-player-list, s-soundmonitor-lineage-wikis, s-wikipedia-chris-huelsbeck.
• New pages (6): tools rockmonitor (split out of the soundmonitor aliases), future-composer, sidid (cheat sheet mapping wiki pages to SIDId names); people laxity, johannes-bjerregaard; group compunet. charles-deenen gained the aliases TMC / TMC/Scoop.
• Updated (32): chris-huelsbeck, soundmonitor, tfmx, shades, matt-gray, matt-gray-player, last-ninja-2, driller, dominator, reyn-ouwehand, maniacs-of-noise, musicfile, charles-deenen, jeroen-tel, rob-hubbard, martin-galway, system-3, hewson, codemasters, chordian; tracker-scene-history (the 1983–92 editor timeline), sid-player-routine (SIDwave's ripper taxonomy, JCH on Hubbard's freelance block, "no room for two drivers"), tracker, tracking-workflow (composing in an assembler); chip-samples (Galway's real-time vs Hülsbeck's NMI streaming, The Banana, Digitronix, later 8-bit methods), multispeed (Winterberg, Crowther, PRI/PVCF, SIDwave's 6×, Tel's 8×), chord-arpeggio (Galway "invented the chord, aka arpeggio … wibble"; Gray's plex + PWM sweep), pulse-width-modulation (JCH on Galway's Rambo hiscore pulsating), hard-restart (JCH as inventor, GRG's soft restart), channel-interleaving, wavetable-programming, instrument-design.
• Notes for lint: the Brief History is a memoir — its "firsts" (Galway invented the arpeggio, JCH invented hard restart, TMC wrote "the first MON player" as a Hubbard clone) are single-witness claims and are attributed as such; the C64 TFMX editor is dated 1987 by a CSDb poster and 1988 by VGMPF/SIDId; Digitronix is 1987 in the lineage file and 1988 in SIDId; Wikipedia's "TFMX co-created with Peter Thierolf" is marked unverified; de.wikipedia's "origin of the tracker" claim stays unverified. The addsec helper failed silently on the first run (zsh word-splitting) and was re-run — all sections verified present.
• Index: 6 pages and 13 summaries added; the inbox line and the scout file record the batch as complete. Not yet committed at the time of writing.

2026-08-30 — plan: C64 legends follow-up (for a fresh session)#

• Wrote inbox/plan-c64-legends-followup-2026-08-30.md (temporary): state, the full ingest pipeline as used in this series, and five steps — (1) the rest of realdmx/c64_6581_sid_players (Hubbard's Monty on the Run player, Galway, Whittaker, Dunn, Kimmel, Bulka/FAME, Bjerregaard, Ouwehand, Deenen's test tunes) → hubbard-player, galway-player pages and more rows on sid-player-routine's driver table; (2) SIDin issues 1 and 3–15 with a triage TOC first (and issue 2's untouched Lazy Jones chapter); (3) first-hand interviews from the VGMPF link lists (c64.com Deenen 2010, Recollection Tel 1994, zakalwe.fi 1996 texts, gamesart.de 2015, archive.org scans if OCR is readable; d64 diskmags skipped); (4) a lint pass with the known contradictions and the code-verifiable (unverified) readings listed; (5) owner-supplied leftovers (Forum64 saves, podcast notes, video notes).
• Moved the session's helpers into the repo: tools/html-to-text.py (HTML → text for raw/) and tools/add-section.py (ripple: insert a section before a heading, add sources to frontmatter and the Sources line, bump updated:); CLAUDE.md's map now lists all tools.

2026-08-30 — ingest: the rest of realdmx/c64_6581_sid_players (plan step 1)#

• Source: the 18 .asm files of https://github.com/realdmx/c64_6581_sid_players not copied on the first pass, saved verbatim to raw/github-realdmx/ (raw/sources.md row "GitHub realdmx … (the rest)"); Gray_Matt_Dominator.asm skipped (duplicates raw/matt-gray/dom6-public.asm). Provenance per header: McSweeney's Monty disassembly, Tognon's SIDin #4 Arkanoid, dmx87's reverse-engineering (Rambo loader, Panther, Mutants — the last unsigned), the rest original Turbo Assembler / DASM sources (Bjerregaard, Deenen, Ouwehand, Kimmel, Bulka, Prijt, Hughes/Dunn). Read in full: Hubbard ×2 and the Rambo loader by the maintainer; the other 15 files by seven reader agents producing line-referenced notes, every quoted value then re-checked against the raw files by grep before it entered the wiki. Sizes for DT88 / Loader 5 were measured by re-assembling the sources with 64tass in the scratchpad; other sizes are label-address or instruction-count estimates and are marked so.
• Summaries (9): s-realdmx-hubbard-player, s-realdmx-galway-rambo-player, s-realdmx-galway-arkanoid-player, s-realdmx-ocean-dunn-hughes-driver, s-realdmx-bjerregaard-player, s-realdmx-ouwehand-mon-tunes, s-realdmx-deenen-test-tunes, s-realdmx-fredgray-kimmel-whittaker, s-realdmx-audial-arts-fame-players.
• New pages (9): tools hubbard-player, galway-player, ocean-music-driver, bjerregaard-player, whittaker-player (each with a cheat-sheet table); people david-whittaker, jonathan-dunn, jeroen-kimmel, fred-gray. No page for Adam Bulka (FAME) or François Prijt (Audial Arts) — two sources each; their drivers are in the summary and the comparison table.
• Updated (34): sid-player-routine (second comparison table, ten drivers), musicfile (the two (unverified) readings replaced by code-verified ones; a section on the other builds and the Scoop player), rob-hubbard, monty-on-the-run, commando, martin-galway, johannes-bjerregaard, reyn-ouwehand, charles-deenen, jeroen-tel, maniacs-of-noise, sidin, sidid (rows for every new driver), cybernoid-ii, tracker-scene-history; techniques/concepts hard-restart (six more ancestors incl. FAME's dated 1988 hard restart), orderlist (pattern lists vs voice programs), instrument-tables (a table of ten layouts), instrument-design, vibrato, pulse-width-modulation, wavetable-programming, chord-arpeggio, pitch-slide-and-portamento, filter-programming, chip-samples (Arkanoid's "samples" are synthesized), multispeed (Arkanoid's two CIA rates), adsr-envelope, channel-interleaving, arrangement, fake-echo, tracking-workflow, detune; s-sidid-player-list extended with the batch's SIDId names and nfo notes.
• Settled from code: MUSICFILE's starttabel/startlen (a start waveform from the note's second frame; $81 also forces pitch $faxx for one frame) and the sequence bytes ≥ $80 (transpose up by byte & $1f, plus repeat $40–$5f and instrument offset $60–$7f). Corrected assumption: Whittaker's Panther driver is not the Lazy Jones routine of SIDin #2.
• Left for lint / later steps (recorded on the pages as attributed uncertainty): who wrote the Dutch Breeze player (Deenen or Ouwehand); Armada's seven song tables vs dmx87's six-song PSID; whether Bjerregaard's Myth is System 3's game; the "STORM LORD MUSIC (C)MON." BASIC-stub decoding (an inference); the Commando file's PBU/PBD constants named opposite to the code; the identity of "Paulie" (SIDId's link to pauliehughes.com suggests Hughes himself); who disassembled Mutants. s-realdmx-sid-players (first pass) still describes the collection as of the three files read then.
• Index: 9 pages + 9 summaries added; the Inbox line marks plan step 1 done; the plan file carries a status block. Lint: 494 pages, 0 broken links, 0 orphans, none missing from the index; 65 (unverified) markers across 50 pages (up from 54/43 — new To-verify items on the summaries above). Site rebuilt. Not committed at the time of writing.

2026-08-30 — triage + ingest: SIDin issues 1 and 3–15 (plan step 2, nine parallel work packages)#

• Raw: the 14 missing issues fetched from commodore.software (downloads/send/625-sidin-magazine/<id>-sidin-issue-<n>, ids 12349 and 12351–12363) as raw/sidin/sidinNN.pdf + PDFKit text (tools/pdf-to-text.swift); issue 1's prose font has no text mapping, so tools/pdf-ocr.swift (new; macOS Vision OCR, page by page, two-column aware) produced sidin01-ocr.txt; sidin02-pdfkit.txt re-extracted the already-present #2 cleanly. One row in raw/sources.md with all dates (#7's printed "9 January 2004" is a typo for 2005).
• Triage: inbox/sidin-toc.md — 75 articles over 15 issues: 26 ★ (player analyses, tool walkthroughs, technical), 15 interviews (issue 2's Cadaver interview had never been ingested), 3 ripping walkthroughs (ingested briefly for sid-format / hvsc), the JSIDPlay2 review skipped, the editorials and news folded into one digest summary.
• Work split into nine packages run as parallel agents (A: #1 + #2 leftovers · B: #3 + #5 · C: #4 · D: #6 + #7 · E: #8 + #9 · F: #10 + #11 · G: #12 + #13 · H: #14 + #15 · N: editorials + news). Each wrote its summaries and new pages and a ripple script; the maintainer ran the scripts serially (165 add-section.py calls on 66 existing pages), merged index and TOC lines, and rewrote sidin. (A session rate limit interrupted all nine during their reading phase; they were resumed with context intact — nothing had been written.) The package entries follow.
• Corrections applied to existing text before the ripple, from the package reports: s-sidin02-matt-gray-driller (the Lazy Jones chapter is now summarised), s-realdmx-galway-arkanoid-player (SIDin #4's date settled to 2003-06-21; the $D418 mask is $DD18 in PSample1), [[chris-abbott]] links on reyn-ouwehand / maniacs-of-noise; matt-gray-player cheat sheet and s-matt-gray-dominator-source: in PLAYER V4.2 $FD ci is the chord-plex command, not a duration, and durations are the whole range $70–$F8 (byte − $70) — verified in raw/matt-gray/dom6-public.asm (AGAIN3 CMP #$FD → PLEXSETUP, VIBDELAY CMP #$F9, NOTEDUR CMP #$70 / SBC #$70), which also settles the plan's "unverified $AF/$DF/$EF/$EE/$97/$E7 bar bytes" (durations 63/111/127/126/39/119). New person page martin-walker (six mentions). Aliases SID Factory, SIDFactory, SID-Factory added to sid-factory-ii.

2026-08-30 — ingest: SIDin #1 (Carrafiello, Iseq, Dig Dug) and #2 (Cadaver, Lazy Jones) — work package A#

• Sources: raw/sidin/sidin01.pdf (prose from sidin01-ocr.txt, code from sidin01.txt — the PDF's own text layer is unreadable) and raw/sidin/sidin02.pdf (sidin02-pdfkit.txt); five rows of inbox/sidin-toc.md: the Carrafiello interview (pp. 8–10), "Iseq music composition program" (pp. 11–29), "Ripping a game: DIG DUG" (pp. 30–52), the Cadaver interview (#2 pp. 8–9) and "David Wittaker's Lazy Jones music routine" (#2 pp. 10–20). Each article read in full; the Dig Dug and Lazy Jones listings decoded byte by byte from the PDFKit text.
• Summaries (5): s-sidin01-carrafiello-interview, s-sidin01-iseq, s-sidin01-dig-dug-rip, s-sidin02-cadaver-interview, s-sidin02-whittaker-lazy-jones.
• New pages (3): person luca-carrafiello; tool iseq (cheat sheet: keys, CIA tempo values, instrument fields, tune notation, Judd's recipes; the Commando cover's instruments); song dig-dug (the eight subtunes as data, the scheduler and setTrack, the 22-entry instruction set as a table).
• Updated (17): whittaker-player (a "The Lazy Jones routine (1984)" section: 20 × $80-byte pieces, 4 bytes per step, AD $66, $20/$21 toggle, tune 1 = 10+11, the cracked-version track, PSID details, 1–4 raster lines), david-whittaker, lasse-oorni (the 2002 interview: Oulu, 1998 start, the player lineage and its reasons, favourites), goattracker (the 2002 state: pattern memory cost, "missing step-programming"), sid-format (the hand-written PSID headers of both rips, the two-build source, the ripper's checks), hvsc (Sid Hunt, rips from cracked images, Plus/4-composed tunes, Agemixer's cover), commando (Judd's Iseq cover instrument by instrument), tracker-scene-history (the 2002 editor argument and lineages), vibrato and wavetable-programming (Iseq's frq+ vibrato and three-call drum), sid-player-routine (Dig Dug's instruction-stream scheduler, Lazy Jones, Iseq's timer values), octave-bass (Lazy Jones and Dig Dug octave pairs, the [+0 +12] arpeggio), future-composer (the Plus/4 conversion), last-ninja-2, jonathan-dunn, sanxion, multispeed (Carrafiello's compo-category argument).
• Contradictions: none between the source and whittaker-player / david-whittaker ("raw two-voice frequency lists with equal note lengths and no effects" holds); s-sidin02-matt-gray-driller's article: field still says the Lazy Jones chapter is "not summarised" — to update. Unverified items kept on the summaries: the datasheet reading of Lazy Jones's $D417 = $03 / $D418 = $1F / $D416 = $FF (a wide-open low-pass on voices 1–2), the graphics-like wrong block, Tognon's German note names; Dig Dug's subtune-to-game-situation mapping and its size/rastertime (not stated); the column assignment of Iseq's three "scream" instruments and the partly legible key list. Dig Dug's composer is unknown in the source (author field <?>).

2026-08-30 — ingest: SIDin #3 and #5, work package B (Schmid interview, pattern searching 1–2, ripping BASIC tunes, Hülsbeck interview, HVMEC)#

• Raw unchanged (raw/sidin/sidin03.txt pp. 8–43, raw/sidin/sidin05.txt pp. 13–33). Articles read in full; the note/duration tables and the Testcard BASIC listing of the ripping article and the un-annotated part of the Future Composer disassembly were not summarised.
• Summaries (6): s-sidin03-schmid-interview, s-sidin03-pattern-searching-1, s-sidin03-ripping-basic-tune, s-sidin05-huelsbeck-interview, s-sidin05-pattern-searching-2, s-sidin05-hvmec.
• Created (3): stephan-schmid (person, Steppe: HVSC crew, Demodungeon), hvmec (publication: Tognon's engines collection, editor vs tracker by explicit note length), player-identification (concept: signatures with wildcarded addresses, the Lostack database format Player { Name / ID / Message / OrderPos / SecPos }, quoted signatures, the road to SIDId).
• Rippled (18): sidid (the 2003–04 predecessor), sid-player-routine (signature rules, OrderPos/SecPos, Colin's 7-byte-per-voice player), sid-format (BASIC-driven tunes: SID POKEs → init, an IRQ player for TI timing, the hand-written PSID header), chris-huelsbeck (his own count: third player → Soundmonitor → TFM → samples → TFMX; Rockmonitor "a hack"; disks on CD-ROM; 2004 favourites), soundmonitor, tfmx, rockmonitor, tracker (editor vs tracker), orderlist (track/sector positions, tune length from end marks), hvsc (the crew at work in 2003, /VARIOUS, 450 engines in 5 seconds, FC 1.0's 1000+ tunes), vice, sid (6581 vs 8580 per Schmid and Hülsbeck), commodore-64, future-composer (its database signature and end marks), channel-interleaving (a 1991 Soundmonitor interleave), matt-gray-player and galway-player (fingerprints), chordian (JCH NewPlayer v1–v3 case study; a 2003 user of the editor).
• Contradictions noted on pages: Hülsbeck's order of revisions (TFM, then the sample build) vs VGMPF's timeline (sampler 1986–87 before TFM July 1987); the Lostack SecPos=$0924,$0924,$0926 repeat for FC 1.0 (probably a typo for $0925). Unverified: "Warren P."; Yodelking = Peter Sandén; the FC track bytes' AND #$1F/AND #$3F as MoN-style transpose/repeat; HVMEC 0.4/0.9 dates (news digest); "TSID" is Tognon's Time Sid Manager, not the identification project — not added as an alias.
• TOC: 6 rows link a summary. Not committed.

2026-08-30 — ingest: SIDin #4 (WP-C: Chris Abbott interview, the Telengard RSID rip, Galway's Arkanoid prose)#

• Source: raw/sidin/sidin04.txt (SIDin issue 4, version 1.00, 2003-06-21), pp. 9–11, 12–22, 23–31 and 98. The Arkanoid listing (pp. 32–97) is the code already in the wiki as raw/github-realdmx/Galway_Martin_Arkanoid.asm (s-realdmx-galway-arkanoid-player); only Tognon's prose was summarised, with the listing consulted to check values (flag tests and #$81 / and #$02, the $D418 mask values, the per-tune minimum durations, the per-voice opcode handlers).
• Summaries (3): s-sidin04-abbott-interview, s-sidin04-ripping-basic-rsid, s-sidin04-galway-arkanoid.
• New pages (2): person chris-abbott (C64Audio, Back in Time, BIT Live, the 2003 canon, his appearances elsewhere in the wiki), song arkanoid (facts from the two Arkanoid summaries).
• Updated (12): galway-player (Tognon's explanations: per-voice opcode sets, the 34-byte running instrument copy, flag semantics, SFX = instrument + pitch, the hidden sample synth, zero-page map, PSID vs RSID), martin-galway (Tognon's verdict; Abbott on the Galway Project and Galway's C128D and source disks), sid-format (RSID for BASIC-timed music; why Arkanoid needs RSID), hvsc, rob-hubbard, fred-gray, david-whittaker (Abbott's 2003 canon); techniques filter-programming (16-byte filter table layout), pulse-width-modulation (Arkanoid's two-phase pulse program and its flags), vibrato (why the middle phase is doubled), pitch-slide-and-portamento (the one-phase portamento), chip-samples (repeat value × delay table, the masked $D418, PSID vs RSID).
• Settled: SIDin #4's date (2003-06-21) for the (unverified) note on s-realdmx-galway-arkanoid-player. Recorded under To verify on s-sidin04-galway-arkanoid: Tognon's "bit 1" wording vs the listing's delay-phase slide; $1C = 0 (table) vs $FF (prose) for "no hard restart"; SETCI listed for three voices but implemented for two; "13 values" for the 14-byte SETFI; "6 octaves" vs $00–$5E; the excerpt's delayTabP1-1,y; tune1 = $09 vs the realdmx copy's $04.
• Skipped: the News section (another package); wavetable-programming, multispeed, hard-restart, chord-arpeggio (the prose adds nothing beyond the code summary); the listing pp. 32–97 (already summarised).

2026-08-30 — ingest: SIDin #6 pp. 11–29 and #7 pp. 17–52 (WP-D of the SIDin batch)#

• Source: raw/sidin/sidin06.txt (issue 6, 2004-05-08) and raw/sidin/sidin07.txt (issue 7, 2005-01-09; the PDF's "2004" is a typo). Six articles read in full: the Jan Harries (Rambones) interview, the Asterion Sid Tracker v1.0 review, "Inside DigiOrganizer" (editor manual + reverse-engineered NMI sample player), the Dustbin (Stefano Palmonari) interview, "Inside Modules" (Ivan Del Duca's 1988 game driver, full commented listing) and "xxlarge" — which turned out to be Ninja/theDream's cover of Jeff's "X large" in under 256 bytes, printed with its source and analysed as the model for the Tiny SID Competition (2005-01-15 to 2005-04-21, 256/512-byte classes). The two issues' News sections belong to another package.
• Summaries (6): s-sidin06-harries-interview, s-sidin06-asterion-sid-tracker, s-sidin06-digiorganizer, s-sidin07-dustbin-interview, s-sidin07-inside-modules, s-sidin07-xxlarge.
• New pages (5): people jan-harries (aliases Rambones, nmioaon), dustbin (Stefano Palmonari); tools asterion-sid-tracker (cheat sheet: sound-table fields, macro rules, track bytes, the one-byte effect list), digiorganizer (cheat sheet: keys, track/pattern/definition formats, sox + S64 conversion, the 4-voice harness, NMI internals); song modules (the engine as tables: 4-byte track entries, note/duration bytes, the $80–$9B jump-table commands, 9×16 instrument tables with sample rows, effect engine, memory map).
• Updated via ripple (25 sections): hvsc (a founder's history: PlaySID → Addition → Ripp Off → NemeSIDs → HVSC 1996 with 1364 rips; 10,000 vs 25,000), sid-player-routine (a ripper's method and harness, the Modules driver, the 256-byte player, a fourth voice from a second player), chip-samples (DigiOrganizer's format, rates, NMI, cohabitation rules, sample library), instrument-tables, chord-arpeggio, vibrato, pulse-width-modulation, filter-programming, pitch-slide-and-portamento, hard-restart, randomization-and-probability (osc3/env3 register copy), orderlist, wavetable-programming (the Gallefoss snare), multispeed, fake-echo, adsr-envelope, tracker-scene-history (the 2004–05 native editors: JCH, SDI, Cybertracker, Odintracker, Sadotracker, Asterion), tiny-sid-compo (WP-E page: the announcement and rules), rockmonitor, future-composer, soundmonitor, sanxion, cybernoid, chordian, jeroen-tel.
• Left for lint (marked (unverified) on the pages): readings of the Modules listing that the article's prose does not state ($80 disables the voice and, once all three are off, the player; $86 n 0 clears the voice's $D417 bit; $9B $05 is a tied note; the LFO is a triangle of period limit × delay; durations = 480/n ticks; the play stub runs the engine twice per frame; "Modules" vs the PSID's "Modulus"); Asterion's transpose zero point ($C0 "middle" vs $8C read as +$0C); DigiOrganizer's derived NMI rate (about 6.2 kHz at speed $A0) and digi step rate; xxlarge's three-voice canon, note-tracked cutoff and the STA $D418-5,y write; the garbled file names of Harries's best/worst tunes; Sadotracker's table semantics in Dustbin's snare.
• Proposed for the driver comparison table on sid-player-routine: a Modules row and an xxlarge row (in the WP report). Index lines and TOC rows delivered to the maintainer. Not committed.

2026-08-30 — ingest: SIDin #8 and #9 (WP-E: Eeben, Tiny Sid Compo, Nata, Catweasel Mk4)#

• Source: raw/sidin/sidin08.txt pp. 18–20, 21–46, 47–53 and raw/sidin/sidin09.txt pp. 12–14, 15–51, 52–60 (Stefano Tognon's SIDin, issues of 2005-07-30 and 2006-01-14). Read in full; the #8 Catweasel review, triaged as skipped, was summarised after all because the #9 follow-up depends on its hardware facts.
• Summaries (6): s-sidin08-eeben-interview (John Player; Polly Tracker's timer-polled 4-channel mixing through $D418, the CIA timer-B guard, SID voice 3 as clock), s-sidin08-tiny-sid-512b (15BB, Crue Gurl, Twone Five, Empty — pattern formats, instrument bytes, restart schemes, the 2005 result tables from the issue's news page), s-sidin08-catweasel-mk4 (the card, sockets, jumpers, hardsid driver/firmware, sidplay2), s-sidin09-nata-interview (GoatTracker 2.2 with 8580 emulation + reSID interpolation, MMC64, VICE vs CCS64, 8580 verdict with named tunes), s-sidin09-tiny-sid-256b (ten entries: zero-page players, Frantic's 3-byte instruments and jsr $b5ff sweep, Laxity's Repeat Me, screen-RAM and $D41B sequencing, a BASIC entry, Splatform256), s-sidin09-catweasel-6581 (driver timing bug, the hardware buffer and the three conditions for a SID card, tune-by-tune tests, Nulone's filter finding via sid2midi and a $D417 patch, two 6581s compared).
• New pages (4): people aleksi-eeben, nata (Stephan Parth); event tiny-sid-compo (rules, dates, full 2005 results, trick-per-entry table; Tiny Sid 2 announced); tool catweasel (facts + Linux cheat sheet).
• Updated by ripple (12): sid-player-routine (new section "Minimal players (Tiny SID compo)" with an 8-entry table and the size tricks), chip-samples (Polly Tracker's method; the card's buffer for sample tunes), chord-arpeggio (Empty's 5-byte arpeggio instrument, Crue Gurl's chord voice, the LN2 two-octave arp dump), adsr-envelope (AD-only / SR-only / single-value envelopes and the explicit restarts), hard-restart (the four tiny-player restart variants; robustness against timing jitter), sid (the Catweasel tests: what a SID needs besides registers, two 6581s, filtered voices vanishing; Nata's and Eeben's chip verdicts), commodore-64 (SID-in-a-PC card; MMC64 as test rig), laxity (TPlayer 01.g0 and TinyPlayer 02.g0), goattracker (Nata's settings; GT as sketchpad for Random Ninja; MK3/MK4 support), vice (hardsid support 1.17/1.18; Nata vs CCS64), hvsc (two composer remarks), last-ninja-2 (Random Ninja; the card test and intro dump).
• Discrepancies recorded on the summaries: Imperial March's nibble order (prose vs code — the pages follow the code); "Tapio Vitanen" vs "Viitanen"; Crue Gurl's "raster $80" vs lda #80; 15BB's "$A030"/"$A000" pulse limits vs $07/$0A in the code. Unverified: Jens = Jens Schönfeld; the 17 D4 → 17 E4 patch mechanism; Repeat Me's adc spdcnt octave wobble; freakyDNA's $15 control byte. Page candidates for lint: 4mat (4+ mentions), Polly Tracker / John Player (tool pages once a second source exists), Agemixer, Frantic, Steve Judd, Simon White, Individual Computers. "nata" on arrangement.md is "sonata", not the person.
• Cross-links: tiny-sid-compo cites the #7 announcement (s-sidin07-xxlarge; end date 21 April 2005 there vs 8 May 2005 in the #8 result page — both recorded) and the news digest for the Tiny Sid 2 results; aleksi-eeben and catweasel take one bullet each from s-sidin-news-digest (tool versions; MK3/MK4/MK4plus features).
• Index: 4 pages and 6 summaries (lines in the WP-E index-lines.md); TOC rows in toc.md. Not committed.

2026-08-30 — ingest: SIDin #10–#11, WP-F (Bayliss, Tiny Sid 2 parts 1–2, Goattracker 2, Laxity, SID Factory vs Ninjatracker)#

• Source: raw/sidin/sidin10.txt (2006-07-16) pp. 13–15, 16–35, 36–41 and raw/sidin/sidin11.txt (2007-02-14) pp. 16–18, 19–29, 30–39; the News sections of both issues were left to the news digest.
• Summaries (6): s-sidin10-bayliss-interview, s-sidin10-tiny-sid-2-part-1, s-sidin10-goattracker-2, s-sidin11-laxity-interview, s-sidin11-tiny-sid-2-part-2, s-sidin11-sid-factory-vs-ninjatracker.
• New pages (2): person richard-bayliss; tool ninjatracker (V2 cheat sheet: track values, pattern row, command, wave/pulse/filter tables).
• Updated (23): tiny-sid-compo (the 2006 compo: eight entries, six analysed with authors and sizes), sid-player-routine (a six-row table of the Tiny Sid 2 players), chip-samples (Bloomquist's 512-byte nibble-packed digi drums), hard-restart (Frantic's two-byte restart row), chord-arpeggio (Odland's one-byte chords, Frantic's JUMP arpeggio, RM's frequency-add arpeggio), orderlist (relative offsets, dropped end markers, transpose lists, no orderlist), instrument-tables (register-stream rows, 2-byte instruments, SID Factory's driver-defined rows), tracking-workflow (MIDI → Rosegarden → GT2; learning from example tunes), multispeed (SID Factory's v5 driver), filter-programming (ENV3 → cutoff in Resolution; SID Factory and Ninjatracker filter tables; GT2's persistent filter program), fake-echo (delayed-copy voices in X-Large 5 and Miniature Rhapsody), arrangement (timer-driven recombination; MIDI reduction), goattracker (the 2006 state and wishlist, Bayliss's verdict, the Ninjatracker converter), sid-factory-ii (a "SID Factory (2007)" section: origin, screen, drivers v5.0x/v6.0x with instrument bytes, tables and commands), laxity (the 2007 interview: ten drivers, SID Factory's design principle, tastes), lasse-oorni (Ninjatracker; the GT2 review), maniacs-of-noise, future-composer, rob-hubbard, martin-galway, fred-gray, johannes-bjerregaard, david-whittaker (one-line mentions from the interviews and the RM cover).
• Recorded as open in the summaries' To-verify lists: Tiny Sid 2 placings are stated in neither part (part 3 is in #12); Laxity's "Tiny Sid … last year (2005)" is ambiguous; Resolution's duration-flag prose disagrees with the listing; "MK4" read as HardSID MK4; Bayliss's birth year by arithmetic. SID Factory (2007) was documented on sid-factory-ii rather than as its own page; its aliases still need SID Factory.

2026-08-30 — ingest: SIDin #12–13, package WP-G (Hermit interview, Tiny Sid 2 part 3, Trivia Arcade rip, TSM & Freedom, JITT64 ×2)#

• Source: raw/sidin/sidin12.txt pp. 15–32 and raw/sidin/sidin13.txt pp. 16–50 (PDFKit text of the SIDin PDFs already in raw/sidin/; issue 12 dated 2008-05-24, issue 13 2010-06-28). Read in full; the JITT64 screenshots (track example, instrument-example tables) and the two "Inside JITT64" block diagrams are not in the text, so the example instruments are described without values. The News sections of both issues belong to another package.
• Summaries (6): s-sidin12-hermit-interview, s-sidin12-tiny-sid-2-part-3, s-sidin12-ripping-trivia-arcade, s-sidin13-tsm-freedom-interview, s-sidin13-jitt64-tracker, s-sidin13-inside-jitt64.
• New pages (4): tool jitt64 (cheat sheet: track commands, the 16 pattern commands with the arithmetic read from the source, hard-restart fields and test-bit recipes, the D/N instrument table, player structure); people stefano-tognon (ice00, editor of SIDin, JITT64 author — the aliases "Stefano Tognon"/"ice00" still sit on sidin.md), tsm (Carmine Migliaccio), freedom (real name not given).
• Updated via ripple (19): hermit (the 2008 3SID tracker, its motives and inspirations), tiny-sid-compo (Plaster and Back To Basics), sid-player-routine (two sub-kilobyte players; JITT64's shadow-first IRQ and one-frame latency; the Trivia Arcade three-entry-point convention), sid-format (a ripping workflow ending in an RSID), vice (monitor commands for ripping; Hermit's and Freedom's views on emulation), hard-restart (JITT64's HR record, timing and write order; Plaster's 0F01 write one tick early), instrument-tables (JITT64's unshared tables with a delay/repeat field), chord-arpeggio, vibrato, pitch-slide-and-portamento, pulse-width-modulation, filter-programming (JITT64 command arithmetic; Freedom's 8580 filter-type switch; the tiny players' sweeps), multispeed (Trivia Arcade's four raster lines; JITT64 "up to 4X"; Hermit's double framespeed), wavetable-programming (Plaster's $09/$81 starts; JITT64's noise drums), instrument-design (JITT64's example bank), goattracker (why TSM and Freedom chose it; the JITT64 GT2 import), jeroen-tel, matt-gray, cybernoid-ii (reception bullets).
• Proposed for the driver comparison on sid-player-routine: a JITT64 row (in the package report). Contradictions with existing pages: none found; hermit.md's To-verify item (SID emulation work, scene tunes) is not confirmed by the 2008 interview, which does add his real name, age, job and the 3SID tracker.
• Left unverified (on the summaries): that the 2008 3SID tracker is SID-Wizard's ancestor; $0326 as the CHROUT vector behind the "autostart"; PORT_MULT = 2; Emil's identity; whether the C64 Pushover was released; the rastertime of the JITT64 player (no figure in the source).

2026-08-30 — ingest: SIDin #14 and #15 (WP-H — Scarzix, the Dominator and Serpent Demo players, Hunter's Moon, the Matt Gray interview)#

• Source: raw/sidin/sidin14.txt pp. 16–74 and raw/sidin/sidin15.txt pp. 9–41 (PDFKit text of the commodore.software mirrors, already in raw/sources.md). Read in full; the Dominator listing (pp. 41–56) was skimmed to verify the command parser, the fade-out, the hat and the drum-table reader; the Serpent Demo listing was read for the sample routine, the bank tables and the pattern data. The News sections and the #15 JSIDPLAY2 review were left to the digest / skipped.
• Summaries (5): s-sidin14-scarzix-interview, s-sidin14-dominator-player, s-sidin14-hunters-moon, s-sidin15-matt-gray-interview, s-sidin15-serpent-demo-player.
• New pages (3): person scarzix (Carsten Berggreen); tool cheesecutter (one-row cheat sheet — only SHIFT+F2 is given in the source); song hunters-moon (Thalamus 1987: Gray's music, Martin Walker's sound-effects engine with its 13 tables). No page for the Serpent Demo (the text gives no year or group) — its material sits on the summary and on matt-gray-player.
• Updated (17): matt-gray-player (V4.2 as Tognon reads it: $70–$F8 durations, $F9 vibrato delay, $FD ci plex, 16-byte instrument layout, drum-table set/subtract rule, hat, fade-out, A4 = 424 Hz; the Serpent Demo build's sample voice), matt-gray (the #15 interview), dominator (build steps and the empty template; the competition's outcome), driller (the Driller → Serpent → Dominator order), chip-samples (the 4-bit NMI sample voice from the code), chord-arpeggio, pulse-width-modulation, vibrato, instrument-tables, orderlist, filter-programming, ring-modulation-and-sync (Hunter's Moon's ring-modulated plonk), randomization-and-probability ($D41B pitch jitter in a 1987 game), chordian (Scarzix on JCH's editor), future-composer, rockmonitor, compos (the 2015 Ambient SID compo), sanxion.
• Settled from Tognon's commented source: on s-matt-gray-dominator-source and matt-gray-player the V4.2 $FD byte is the plex (arpeggio) command, not the note duration, and the "uncommented" bytes $AF/$DF/$EF/$EE/$97/$E7 are durations (kk − $70); durations are the whole range $70–$F8, not a $70–$7F nibble. Recorded, not yet applied to those two pages (maintainer's call).
• Contradictions noted: Tognon's pattern table gives $70..$F9 as durations while his own listing decodes $F9 nn as the vibrato delay (VIBDELAY); his Serpent example calls $FA,$03 "index 3 of bank 1" while his table makes the high nibble the bank (0); he writes "65tass" once. Unverified readings marked on the summaries: the fade-out length (≈ 30 s computed), the bank-1 loop-point reading of BANKHI1, the nibble rate (≈ 3–4 kHz computed), "younger" = earlier.
• Hunter's Moon: the "★ player" article is a sound-effects engine (Martin Walker), not a version of Gray's player; a comparison-table row for sid-player-routine is proposed in the package report rather than added. [[martin-walker]] is linked from hunters-moon without a page (lint TODO; he is now named on six pages).

2026-08-30 — ingest: SIDin Editorials and News, issues 1–15 (WP-N of the SIDin batch)#

• Source: the Editorials and News sections of all 15 SIDin issues (raw/sidin/sidin01-ocr.txt pp. 4–7, raw/sidin/sidin02-pdfkit.txt pp. 4–7, raw/sidin/sidin03.txt–sidin15.txt per inbox/sidin-toc.md), read in full — about 208 KB of text.
• Summary (1): s-sidin-news-digest — a dated timeline by topic (GoatTracker 1.33→2.69 and the stereo/HardSID builds, NinjaTracker 1.0→2.03, JCH NewPlayer 21.g4/g5, SID Factory 0.05, SIDId 1.0–1.06, Hermit's 3SID tracker, Polly Tracker, Asterion, TFX, SDI 2.0, X-SID, CRD-Tracker64, SIDwinder, SIDedit/Audio::SID, SidDasm, JC64Dis, TSID, HVMEC 0.4/0.9, PSID64 0.4–0.7, SID Known; HVSC 5.0 (2002-08-17) → #63 (2015-06) with sizes and the structural changes; CGSC 1.9 → 1.30 with file counts; Sidplay2/w, libsidplay, reSID 0.14–0.16, Lankila's 2007 6581 distortion model, XSIDPLAY/XSidplay2, ACID 64, SIDPlayer, SIDPLAY for Mac, NanoSIDPlay, JSIDPlay/JSIDPlay2, plug-ins, Sidplay64, SOASC; HardSID Quattro/PCI/4U/Uno/UPlay, Catweasel MK3/MK4/MK4plus, SwinSID/SwinSID2, MSSIAH; c64.sk SID Compos II–8 and the Cover Compo, GT2 compo, SidWine 2–3, Tiny Sid 1–2, Big Sid, the rip compos, 2010 SID COMPO, PollyTracker Compo, DMC compo 2007; composers.c64.org, kubarth.de, Reformation, The Piano Collection, Project Sidologie, Tel Me More, Back in Time Symphonic Collection) and a section on the magazine itself (OpenOffice .sxw→.odt, hosting on digilander → iceteam.altervista.org, the rights statements, the reasons for each gap).
• Updated (20, via ripple): goattracker (release history 2002–2010), ninjatracker, sid-factory-ii (SID Factory 0.05, 2007), sidid (1.0x and adoption), hermit (the 2008 3SID tracker), lasse-oorni, chordian (NewPlayer 21, SID Factory), hvsc (release timeline 2002–2015), sid (reSID milestones, Lankila's measured nonlinearities, $4CC7 cycles/frame, the first SwinSID), commodore-64 (HardSID, Catweasel, SwinSID, MSSIAH, playing .sid on the machine), sid-format (PSIDv2NG/RSID/BASIC flag/PSID v3/3SID dates, PSID64), laxity, matt-gray (the 2014 return), chris-abbott (Street Hawk; the c64audio Kickstarters), chris-huelsbeck (The Piano Collection), jeroen-tel (HVSC finds, Tel Me More), aleksi-eeben (compo placings), tiny-sid-compo (origin, HVSC 42, Big Sid Compo), hvmec (0.3/0.4/0.9, the 2015 port), stefano-tognon (his tools and compos). catweasel, aleksi-eeben and tiny-sid-compo (WP-E) already carried the news facts citing this digest, so the Catweasel call was dropped and the other two trimmed.
• Not rippled: sid-wizard (the news does not relate the 3SID tracker to it); sidin left to the maintainer. Discrepancies recorded in the digest's To verify and in the WP report: HARDshit 0.2 vs "2.0" in the TOC, SIDPLAY 3.1.2 vs "3.12", "SIDPlayerDS" is a Dreamcast port, GoatTracker 2.18 (not 2.17) in #9, the readme's "v2.4 pulse speeds doubled" vs the news' 2.0 RC1, CGSC 1.27's repeated totals, HVSC #50's printed count, ninjatracker's "early 2007 on the same day" vs the news' August–September 2006 / "same week"; the hvsc and hvmec unverified lines on numbered updates / songlength database / HVMEC versions are now confirmed by the news.

2026-08-30 — merge + lint + site: the SIDin batch closed#

• Merged the nine packages: 44 new summaries + s-sidin02-matt-gray-driller (already present) under a new index group "SIDin — Stefano Tognon's e-zine"; 27 new pages in their groups (7 tools, 13 people, 4 songs, 1 event, 1 publication, 1 concept); inbox/sidin-toc.md links 74 of 75 rows to a summary (WP-N's title corrections applied: HARDshit 0.2, SIDPLAY 3.1.2, GoatTracker 2.18, SIDPlayerDS = Dreamcast); the index's Inbox lines updated; sidin rewritten (issue table with dates, page counts, interviews and articles; facts from the editorials; the "Stefano Tognon"/"ice00" aliases moved to stefano-tognon).
• Consolidation after the ripple: a third driver table "Seven more from SIDin (1982–2010)" on sid-player-routine (Colin's Sheba routine, Dig Dug, Lazy Jones, Hunter's Moon SFX, Modules, xxlarge, JITT64 — the six SIDin sections the packages appended stay as detail); ninjatracker / sid-factory-ii release wording reconciled with the news (NT 2.0–2.03 August–September 2006, SID Factory 0.05 alpha in the #11 news, "released in the same week" per the editorial); hvsc's To-verify line on numbered updates / songlength database now cites the digest (only vsid stays unverified); hvmec's settled To-verify line removed; tiny-sid-compo's "placings stated in neither part" now points at the news-sourced results; stacked blank lines from 165 insertions normalised (79 files).
• Lint: 565 pages, no frontmatter issues, 0 broken links, 0 orphans, none missing from the index; (unverified) markers 140 across 81 pages (up from 65/50 — the new summaries' To-verify lists, all itemised in the package entries above). Site rebuilt: 565 pages, 30 cheat sheets (the seven new tool pages included). CLAUDE.md's map lists tools/pdf-ocr.swift. Plan file: step 2 marked done with the findings for step 4. Not committed (the owner commits on request).

2026-08-30 — ingest: first-hand interviews (plan step 3)#

• Sources fetched into raw/: C64.COM interview #8 (Charles Deenen, 2010), Recollection's Domination #1 Tel interview (1994), the two IMR questionnaires from the zakalwe.fi mirror (Tel 1996 — the plan's "zakalwe" candidates turn out to be Per 'ZiKE!' Carlbring's IMR project, ©1996–99; Hülsbeck 1996-08-29), GamesArt's Tel interview (2015), and three archive.org OCR cuts: Happy Computer 3/87 "Der Soundprofessor", 64'er 2/88 "Von der Spielidee ins Regal" (Rainbow Arts feature), C=Lehti 4/89 "SIDin hollantilaiset valjastajat" (Jori Olkkonen's contemporary MoN profile). All three planned scans were readable — nothing skipped. New raw collections: imr/, gamesart/, happy-computer/, c-lehti/.
• Summaries created (8): s-c64com-charles-deenen-interview, s-domination01-jeroen-tel-interview, s-imr-jeroen-tel-interview, s-imr-huelsbeck-interview, s-gamesart-jeroen-tel-interview, s-happy-computer-soundprofessor, s-64er-rainbow-arts-feature, s-c-lehti-maniacs-of-noise.
• Page created (1): scoop-designs (group) — 4+ sources now describe it; the MoN page's "the files do not say what Scoop Designs was" note now points there.
• Pages updated (18): jeroen-tel (birthplace contradiction SETTLED: born Eindhoven, raised Helmond, his own words; four-interview section), charles-deenen, chris-huelsbeck, maniacs-of-noise (1989 profile: version 78, two-channel digi, mailed-disk origin variant kept as attributed disagreement vs the PCW-show story), musicfile, soundmonitor, chip-samples, pulse-width-modulation, tracking-workflow, arrangement, melody-writing, johannes-bjerregaard (Deenen's 1989 inspiration credit), hewson, turbo-assembler, cybernoid, martin-galway (Hülsbeck's Bad Cat claim), rainbow-arts, 64er; index (new group line + 8 summary lines).
• For the step-4 lint pass: the Tel birthplace item is settled; the MoN founding story now has two attributed variants (mailed disk 1989 vs PCW show 2010); C=Lehti's "version 78" reading is flagged as OCR-uncertain on musicfile.
• Fixed tools/html-to-text.py: it was committed without its import sys,re,html line and crashed on every run (NameError).

2026-08-30 — lint: legends contradiction pass (plan steps 4–5)#

• Resolved: MoN driver authorship on charles-deenen — all four first-hand accounts (Deenen 1998/2010 "the driver that I wrote", C=Lehti 1989, Tel 2015, c64.com 2010) agree Deenen wrote it; VGMPF's "he used Jeroen Tel's sound driver" kept as an attributed outlier. Tel's Probe dates reconciled on jeroen-tel (VGMPF's January 1988 = the freelance-exclusive start; the six-month in-house contract "at eighteen" = 1990/91, the move the 1991 interview describes). Tel's birthplace was settled in step 3 (born Eindhoven, raised Helmond).
• Stated as unresolved (owner's rule: when we cannot resolve a contradiction, state it): TFMX C64 dating (1987 CSDb poster vs summer 1988 VGMPF/SIDId) on tfmx; Digitronix dating (1987 tracker-history, exact date "undefined", vs 1988 SIDId) on rockmonitor (s-sidid-player-list added to its sources); the two Battle Valley origin stories on maniacs-of-noise/hewson (from step 3).
• Verified already clean: the step-1/2-settled code readings carry no stale (unverified) markers on musicfile / matt-gray-player / hard-restart / orderlist; SIDwave's single-witness firsts are attributed on martin-galway, chord-arpeggio, hard-restart, maniacs-of-noise, musicfile; only rockmonitor carries the Rockmonitor/Digitronix aliases; the Wikipedia claims stay marked (unverified) on tfmx, chris-huelsbeck, tracker, tracker-scene-history. health.html: 0 orphans, 0 missing from index; the 30 stubs are known-small pages (TW articles, minor people); none of the 12 other To-verify sections is resolvable from raw sources on hand.
• Step 5: Forum64 re-checked — still Cloudflare 403; the Retro Hour episode and the Behind-the-notes videos need the owner's notes; the d64 diskmag interviews need an emulator. Everything open, plus the lint page candidates with mention counts (DMC 29, Ben Daglish 25, 4mat 23, reSID 22, Frantic 17, Tim Follin 13, HardSID 12, CyberTracker 12, Agemixer 11, Hawkeye 9, Marcel Donné 9, Sidplay2 8, SwinSID 6, Odin Tracker 6, Probe Software 6, ACID 64 5), moved to inbox/leftovers-c64-legends-2026-08-30.md.
• Plan complete: inbox/plan-c64-legends-followup-2026-08-30.md deleted per its own instruction; its index line replaced by the leftovers note.

2026-08-30 — create: the 17 lint-candidate pages (from existing summaries only)#

• The owner asked for the leftovers note's page candidates; four parallel agents wrote them from the wiki's own summaries (no new ingestion), the maintainer applied the ripple. New pages — tools: dmc, cybertracker, odin-tracker, resid, sidplay, acid-64, hardsid, swinsid; people: dag-lem, frantic, 4mat, agemixer, ben-daglish, tim-follin, marcel-donne; group: probe-software; song: hawkeye.
• 58 backlink edits across ~35 existing pages (first prose mention linkified; verbatim quotes untouched); 17 index lines; leftovers note's candidate section marked done.
• Citation check prompted by an agent: lasse-oorni's "Téli Sándor (HardSID 4U support)" credit IS in the raw GoatTracker readme ("HardSID 4U support by Téli Sándor."), so the s-goattracker-readme citation stands unchanged.
• Honest gaps recorded as To verify on the new pages: DMC's original author and name expansion, Odin Tracker's author/versions, CyberTracker's "Cyberbrain" (2010) vs "from NoName" (2004) attributions.

2026-08-30 — leftovers: Forum64 threads closed without ingest#

• The owner read both Forum64 threads in a browser: 45509 (Soundmonitor manual) has no useful content and no manual links; 17082 (Hülsbeck's first tracker) likewise nothing worth ingesting. Marked closed in inbox/leftovers-c64-legends-2026-08-30.md; the TFMX dating disagreement stays as stated on tfmx. (For the record: 17082 has a Wayback capture from 2025-11-30, fetchable if ever needed; 45509 is archived nowhere.)

2026-08-30 — scout: cheesecutter + defmon#

• Web search for sources on the two trackers; wrote inbox/scout-cheesecutter-2026-08-30.md (8 candidates) and inbox/scout-defmon-2026-08-30.md (9 candidates). Nothing ingested; index Inbox section updated.
• CheeseCutter: the GitHub repo is canonical (D language, reSID with Cadaver's sid.cpp, ACME 0.91, GPL v2, Version = 2.10, SDL2 merged 2026-03; only man pages in doc/); the homepage theyamo.kapsi.fi is 401 today but Wayback 2024-12-21 shows v2.9 zips in regular + stereoSID builds; Scarzix's tutorial videos found (video-only). Key finding: chordian = JCH, author of the JCH editor/NewPlayer that CheeseCutter descends from — his blog memoirs are candidate #7 and would upgrade both chordian and cheesecutter.
• defMON: frantic's original wiki (toolsforscholars.com, 2014) is 404; the live successor is defmon.vandervecken.com (docs hub updated 2025-03: iLKke's defMONing 101/102, FTC's super-commands and BPM-sync docs, stereo-SID and DIN/MIDI/LSDj/Nanoloop sync pages). Download page gives the dated release list ("Cracked by Hackers" 2013-08-03 → "Nudge Nudge!" 2020-10-08, prehistoric 2008). The 28-page chipmusic.org megathread was started 2014-06-08 by frantic himself (page 1 verified via curl; chipmusic 403s WebFetch). No defmon page exists yet — ingesting #1–#3 would create it and give sync its first C64 external-sync section.

2026-08-30 — ingest: defMON (all scout ★) + goto80#

• Sources fetched into raw/: the whole defMON wiki (14 DokuWiki pages via export_raw + defmoning-102.rtf + defbpm.zip + the seqED key-map PNG → raw/defmon-wiki/), the 28-page chipmusic.org megathread (444 posts, 2014-06-08 → 2026-05-16 → raw/chipmusic/defmon-t14505-p01..28.txt), electro.pizza's VICE guide, Wikipedia Goto80, three goto80.com posts + the Improv page, and the Remix64 Goto80 interview (2002). All recorded in raw/sources.md. The three video tutorials were not saved (metadata only, noted in sources.md).
• 12 summaries written by three parallel agents (per the batch-ingest pattern; agents wrote only new files): the seven s-defmon-wiki-* pages, s-chipmusic-defmon-thread, s-electropizza-defmon-vice-chip, s-wikipedia-goto80, s-goto80-site-defmon-posts, s-remix64-goto80-interview. The keys_seqed_v3.png key map was transcribed into s-defmon-wiki-manual (one label, CTRL+Q "S:QNORR", marked unreadable rather than interpreted).
• New entity pages: defmon (tool, full cheat sheet: sidTAB columns + key table) and goto80 (person).
• Ripple (12 existing pages): frantic (refined defMON timeline ~2006/2008/2013, design-philosophy quote, 2021 withdrawal, Mad Frantic/FTC aliases, the Kommando Knorr alias-vs-duo tension stated per the owner's contradictions rule), sync (new defMON section — first C64 external-sync coverage; notes MARiO64's SID-Wizard 1.7 sync mod speaks the same protocol), hard-restart (### defMON: manual HR, the −1 step, iLKke's ADHR-by-artist table), multispeed (### defMON: editor-side CIA timers, $1003/$1006), instrument-tables (defMON as the single-table limit case), filter-programming (### defMON: RE mask 8, CP+ACID, $0200 8580 floor, no $D415 writes), sid-player-routine (defMON player interface), vice (defMON setup guide + community version preference), lsdj (defMON as sync master), live-tracking (goto80's improv practice; platform c64 added), ring-modulation-and-sync (### defMON: WG additivity, ORA'd columns, channel pairing), pulse-width-modulation (### defMON: PW/PS, the $YXY expansion).
• Index: 2 entity lines, 12 summary lines (new "defMON and Goto80" group), scout-defmon inbox line updated; scout file marked with status.
• Not created (left as red links / candidates): jch-editor (recurs across cheesecutter/chordian/goto80 material — strong candidate), hack-n-trade, scannerboy, anarkiwi. Open scout rows: CSDb pages, Lemon64 thread, chipmusic sync thread.
• Contradiction handling: none found between the thread and the wiki docs; the SIDId "made 2008" reading refined, not contradicted. Kommando Knorr stated both ways on frantic.

2026-08-30 — ingest: sid-wizard-vessel (owner identity revealed)#

• The owner identified themself as m64 / Mario 64 / MARiO64 — the MARiO64 of the defMON thread's SID-Wizard-1.7 sync mod. Saved to auto-memory (m64-identity).
• Ingested https://github.com/M64GitHub/sid-wizard-vessel README (verbatim → raw/github-m64/; recorded in raw/sources.md; feature-demo video and screenshots left as links). Summary: s-sid-wizard-vessel-readme.
• New entity page: sid-wizard-vessel (tool; lineage Hermit 1.8.7 → anarkiwi's Vessel port → M64's NMI sync 2021-09 + live features 2023-04; cheat sheet of the live-act keys).
• Ripple: sync (new "MIDI sync IN" subsection under SID-Wizard; the MARiO64 sentence in the defMON section now identifies the owner), sid-wizard (fork lines added under "Newer versions and reimplementations"), live-tracking (the vessel performance grammar beside the goto80/defMON section).
• Index: tool line, summary line, vessel added to wanted pages (anarkiwi's MIDI interface — future ingest: github.com/anarkiwi/vessel).

2026-08-30 — ingest: CheeseCutter + the JCH lineage (scout picks, no videos)#

• Committed the day's defMON/goto80/vessel work first (commit 888c465), then ingested the CheeseCutter scout's text sources: the GitHub repo docs (man pages ccutter.1/ct2util.1; AUTHORS/NEWS empty — the tool has no manual anywhere), the v2.9 release zip via Wayback (14 example .ct tunes; README only), all 10 CSDb release pages 0.4.0 (2011) → 2.9.0 (2017) plus the discovery of CheeseCutter 2.11R "Hypercut" (arch0N, 2026-07-25, credits "Abaddon of Artline Designs"), Triad's pages (Abaddon: Finland, Damage → Fairlight 2003 → Triad 2012–2017, the "Release The Cheese" compo), the chipmusic tutorial thread (4th Scarzix video found), and Chordian's two JCH memoirs (~22000 + ~1900 words).
• 4 new summaries: s-cheesecutter-repo-and-releases, s-chipmusic-cheesecutter-tutorial-thread (by the maintainer), s-chordian-my-computer-chronicles, s-chordian-player-editor-notes (one agent).
• New page: jch-editor (fills the wanted link; OldPlayer 1987 → NewPlayer 5/10.G0/15.G7/17.G1/20.G4/21 → editor v1 1988-11 … v3.04 1991-08; design pre-dates SoundTracker; byte-coded user list). cheesecutter rewritten (release table, ct2util, Hypercut, lineage); chordian reframed as Jens-Christian Huus = JCH of Vibrants (handle from the 1990 tune Chordian, adopted 2006).
• Contradiction stated per owner rule: SIDwave's "JCH is the inventor of hard-restart" vs JCH's own disclaimer ("It was Jesper [JO of Amok] who first made me aware … previously used by Rob Hubbard and also Laxity had figured it out"; implemented in NewPlayer 17.G1) — both accounts now side by side on hard-restart and chordian.
• Ripple: hard-restart, laxity (the 1988 player episode, Vibrants, NewPlayer 21), johannes-bjerregaard (test-bit discovery at JCH's, "by far the best C64 composer"), scarzix (JCH-side corroboration, 4 videos, v2.9 example tunes), sid-factory-ii (the full JCH → Laxity → Chordian line), goto80 (jch-editor linkified), multispeed (JCH's 1989 2× experiment), ring-modulation-and-sync (test-bit provenance), future-composer (JCH's one-tune rejection).
• Index: 2 tool lines, chordian people line, 4 summary lines (new "CheeseCutter and the JCH lineage" group), jch-editor removed from wanted, triad added to wanted earlier; scout-cheesecutter marked with status (videos remain open).

2026-08-30 — ingest: the recovered CheeseCutter manual (Wayback CDX)#

• On the owner's hunch, enumerated the dead theyamo.kapsi.fi via the Wayback CDX API (the single-URL availability check had missed everything): found ref.html — the "Guide", CheeseCutter's actual manual (~20 snapshots 2013–2019, content stable; recovered from the oldsite mirror capture 2023-02-17) — plus about.html, changes.html (2-series changelog from 2.0.0 = 2011-12-09) and the older ct00ref.html player doc. Saved as text to raw/theyamo-kapsi/; manual-illustration PNGs noted but not saved.
• Summary: s-cheesecutter-guide (by the maintainer). Key recoveries: the full command system (sequence command column ranges, 3-byte command table with HR ADSR in row 0 default $0F00), instrument bytes A–H with FOUR restart types (0x note / 4x soft / 8x hard / Ax "Laxity restart"), wave/pulse/filter/chord table formats (10-bit filter sweeps "4 times as fine as JCH players"), swing tempo = chord program 0 at speed 0/1, editor keys (F1/F3/marks/Ctrl-L), and the exact player lineage: "based on NP21.G4 by Laxity / Vibrants / MoN", 2.x = abaddon's NP21.B6 fork.
• cheesecutter page: cheat sheet went from 3 rows to the real thing (keys, command column, command table, instrument format, tables, workflow tips); intro lineage corrected to NP21.G4/NP21.B6; 0.x-vs-2.x facts (32 subtunes / 48 instruments / 128 sequences, multispeed ≤4x).
• Ripple: hard-restart (### CheeseCutter, the four-type table), shuffle-funktempo (### CheeseCutter — swing via chord program 0), chord-arpeggio (### CheeseCutter — indexed chord table, negative transposes), filter-programming (### CheeseCutter), pulse-width-modulation (### CheeseCutter), instrument-tables (CheeseCutter section), jch-editor + laxity (NP21.G4/B6 lineage precise). Scout #2 marked resolved; index summary line added.

2026-08-30 — query: how to make a song like Last Ninja 2 in SID-Wizard#

• Answered from last-ninja-2, matt-gray, matt-gray-player, sid-wizard, pulse-width-modulation, chord-arpeggio, wavetable-programming, vibrato. Translated the documented Gray techniques (chord plex + PWM sweep, drum tables, Sanxion-style hat-on-bass, raw-value vibrato, no filter, ADSR values from the sidin09 dump) into SID-Wizard chord-table / WF-ARP / pulse-table settings. Offered to save the synthesis as a technique page (not yet created).

2026-08-30 — scout: the DUET repo (local)#

• Scouted m64's DUET repo at /Users/m64/m64/space_64/duetracker (commit 85155dd, v0.1.0-alpha, not yet published) for the docs ingest that s-duet-readme left open. docs/ holds 19 end-user pages + 11 dev pages, each mapped one-line in the repo's docs/README.md; internal_docs/ holds ~2.5 MB of working notes plus two SID-Wizard research files (RESEARCH-SIDWIZARD.md, RESEARCH-SIDWIZARD-DEEP.md) that document SW 1.97's formats byte-exact (validated 324 SWI / 143 SWM files) and the whole player.asm runtime with line refs and quirk lists.
• Wrote inbox/scout-duet-2026-08-30.md: 10 candidate groups (docs grouped thematically, the two research files flagged ★★), skip list (DESIGN/HISTORY/GOTCHAS etc., generated keys.md/cli.md, dev process pages). Proposed ripple set includes a new swm-format page and deep updates to sid-wizard, instrument-tables, and the SID technique pages. No ingest yet — m64 picks.
• Naming: m64 prefers "m64" over "the owner"/"the wiki owner" in pages and prose; use it on all pages touched from now on.

2026-08-30 — ingest: the DUET docs + the SID-Wizard 1.97 research (scout phases 1+2)#

• Snapshotted m64's DUET repo (commit 85155dd, v0.1.0-alpha, not yet published) into raw/duet/: the full docs/ tree (19 user + 11 dev pages) and internal_docs/RESEARCH-SIDWIZARD{,-DEEP}.md; recorded in raw/sources.md. Two parallel agents wrote new files only; ripple applied serially by the maintainer (subagent-batch-ingest pattern).
• 7 new summaries. Docs (agent A): s-duet-docs-grid-and-sides, s-duet-docs-formats, s-duet-docs-dust-dusk, s-duet-docs-clipboard, s-duet-docs-sounds-mixer-bounce. Research (agent B): s-duet-research-sidwizard (byte-exact SWI/SWM formats validated 324/324 + 143/143, subtune truths, player runtime model) and s-duet-research-sidwizard-deep (6502-exact algorithms: SETPWID, FILTPRG arbitration, WFARPTB, vibrato/slide formulas, full FX dispatch, quirk lists) — player.asm line refs kept throughout.
• New page: swm-format — the .swm/.sws/.swt/.swq one-container family (+.swi): magic-driven layouts, shrinking caps (37→23 instruments, 8→1 subtunes), backwards parsing, subtune-major orderlists.
• duet rewritten from "basics only" into the full tool page: the two engines with their validation numbers, the four session badges and the .zon container law, the SID columns' language, co-play, DUST/DUSK (with the frozen v1 patch contract), the born-bypassed console and sends a–d, the clipboard dialects (DUET SID ≡ DUET SW — corrects the README-era "DUET SW"-only phrasing), headless/libduet, deliberate divergences from SW 1.97 ($12/$15 operand 0, SUBTUNES SPLIT, multi-subtune .swt/.swq refused), and a key cheat sheet.
• Ripple, 20 pages. From the clipboard docs: openmpt (the ModPlug clipboard text format, field by field — documented almost nowhere else), schism-tracker, 037-lab (producer paste rules). From the research: sid-wizard (new "Inside the 1.97 formats and player" section), instrument-tables, pulse-width-modulation, filter-programming, vibrato, pitch-slide-and-portamento, chord-arpeggio, hard-restart, multispeed, shuffle-funktempo, adsr-envelope, orderlist, detune, ring-modulation-and-sync — each a focused "What the player actually does (SW 1.97 player.asm)" block. From the docs: it-format (DUET's IT 2.14 writer choices), hard-restart (### DUST — restart as a patch parameter with look-ahead), wavetable-programming (### DUSK).
• Manual-vs-player discrepancies stated on the pages (per m64's contradiction rule; both sides kept): the WF-table $FE rule (manual/charts "≥ $40 = hold" is SWMconvert's GT-export heuristic; the player halts only ≥ $80, player.asm 2422), the PW set-row high nibble (true at the SID output, but the ghost keeps 7 bits and later sweeps run from them), the vibrato $20/$30 naming inversion (traced trajectories opposite to the source comments; research marks it an open question — an A/B against real SW playback would settle it).
• Index: duet tool line expanded, swm-format under Formats, new "DUET" summaries subgroup (8 lines), scout-duet marked phases 1+2 done (open: dev engine docs #8–#9). Skipped as summaries: keys.md/cli.md (generated), troubleshooting/visuals/getting-started/glossary (context only), dev process pages, DESIGN/HISTORY/GOTCHAS internals.

2026-08-31 — ingest: libduet (DUET scout phase 3, m64's pick)#

• Ingested raw/duet/docs/dev/libduet.md (read in full): the engine as a library — Zig module + libopenmpt-shaped C ABI, "bytes in, samples out", opens no file / creates no thread / owns no device. Summary: s-duet-docs-libduet.
• Noteworthy captured: the three render laws and the 8-song byte-identity A/B; the resolver; the loop law (SW modules never end — cap); the frame-stamped event ring (4096×24 B, row/order/note/effect/marker events with dialect) and sync markers as the demo/game sync road; instruments-as-sound-effects by family; format_cell drawing cells exactly as the grid does; the threading contract; the wasm web player (AudioWorklet reactor, 22× real time at 2.4 MB, 90-song gallery with provenance) and its A/B verdict that reSID's sample path is fully integer.
• Ripple: duet (Headless and libduet section rewritten in depth), sync (new "Syncing visuals to the music" section — the code-follows-tracker direction), resid (the integer-sample-path datum from the wasm/native byte-identity).
• Per m64: the remaining dev engine docs (#8: sid-wizard-engine / sid-and-resid / it-engine) and architecture.md are closed as summaries ("not so important") — they stay citable from raw/duet/docs/dev/. Scout-duet now fully resolved; index summary line added, scout + inbox lines updated.

2026-08-31 — scout: LSDj "ghost channel" (m64's request)#

• Wrote inbox/scout-lsdj-ghost-channel-2026-08-31.md — 7 candidates. Nothing ingested.
• Confirmed the technique m64 described: two waveforms superimposed additively inside the wave channel's 32×4-bit RAM (a hand-drawn sine plus a small square riding at 2×/3×/4× the fundamental), so one channel voices a bass note plus a second note — which must be a harmonic of the first (2 = octave, 3 = octave+fifth …). Named and popularised by TylerBarnes on chipmusic.org, 2013-07-11; nitro2k01 added the 32/x sample-spacing rule and the overtone→interval table in the same thread; bryface flagged the headroom cost. In practice only the octave and the octave+fifth are usable.
• Primary source is a video (Tyler Barnes, 22:15); the forum thread carries most of the technical content in text. Both YouTube candidates have auto-generated English captions and yt-dlp is available, so video ingest is possible here — flagged for m64's decision rather than skipped by default.
• Dead end recorded: the demo song file (Ghost Channel.lsdsng, app.box.com) is 404; a 2025-07-20 Wayback capture still names it. The same file is already cited inside raw/chipmusic/lsdj-complete-drum-kit-on-noise-channel-t15681.txt.
• No wiki page covers this yet; proposed new technique page [[ghost-channel]], plus entities for nitro2k01, Tyler Barnes and bryface on ingest.

2026-08-31 — ingest: the LSDj ghost channel (m64's pick: scout candidates 1, 4, 3, 2)#

• New raw collections. raw/chipmusic/lsdj-ghost-channel-t11649.txt (16 posts, 2013-07-11); raw/hackaday/{channelhacking-the-gameboy,splitting-pulse-waves}.txt (Quint with utz, 2019); raw/youtube/{tylerbarnes-advanced-lsdj-ghost-channel,defensemech-learning-lab-overtones-ghost-channel}.txt. All recorded in raw/sources.md.
• New tool + rule: video ingest. tools/yt-transcript.py (yt-dlp) writes a provenance header + description + caption transcript in [m:ss] paragraphs, undoing YouTube's rolling-caption repetition. CLAUDE.md now carries the rule under Operation: ingest: auto-captions are a machine transcription, so hex values, wave/frame numbers and command names from them are unverified until checked against a text source, and on-screen values that were never spoken go to ## To verify rather than being guessed. This closes the "video only — skip" default that earlier scouts used.
• 4 new summaries: s-chipmusic-lsdj-ghost-channel, s-tylerbarnes-ghost-channel-tutorial, s-defensemech-ghost-channel-allpass, s-hackaday-gameboy-channelhacking.
• New technique: ghost-channel — two waveforms added inside one 32×4-bit wave so a Game Boy wave channel voices a bass note plus an independently steerable harmonic. Carries nitro2k01's overtone→interval table with the 16/x half-wave spacing, three recipes (hand-drawn 2013, allpass + high Q 2020, crunched from audio), and the two constraints: the ghost note must be a harmonic (only the octave and octave+fifth survive equal temperament) and the sum costs amplitude.
• New concept: harmonic-series — partials as intervals in LSDj note names, the overtone-vs-partial naming clash between the sources (this wiki writes the multiplier), just vs tempered intervals, CUTOFF calibrated in harmonics.
• New entities: tyler-barnes, nitro2k01, bryface.
• Ripple, 9 pages: wavetable-programming (LSDj section), channel-interleaving (the frequency-domain sibling), chord-arpeggio (what C cannot do), octave-bass (new ### LSDj), filter-programming (new ### LSDj (the wave synth's filter) — the page had been SID-only; allpass/Q and CUTOFF-as-harmonics), game-boy-apu (new "What 32 × 4 bits buys you": the additive budget, the wave-swap click dead end, the pulse-splitting alternative at 7 levels instead of 15), lsdj, lsdj-wave-cruncher (the tool the 2013 thread asked for), defense-mechanism (the LSDj Learning Lab streams), chipmusic-org (thread list and members).
• Corrections made while writing: nitro2k01's rule is 16/x samples between alternations (a full rider cycle every 32/x) — the scout file's first phrasing was fixed in place.
• Dead end kept on the pages: Tyler Barnes's demo save Ghost Channel.lsdsng (app.box.com) is 404; only a Wayback capture of the file page survives.
• Index: new lines under Channel economy, Theory in semitones, People and the LSDj summaries group; the scout file listed under Inbox. tools/lint.py clean (631 pages, no broken links, no orphans, no frontmatter issues).

2026-08-31 — ingest: LSDj wave-channel envelopes, articulation and the synth frame chart (m64's picks 1, 2, 4)#

• New raw: raw/chipmusic/lsdj-wav-channel-release-t11066.txt (18 posts, 2013-05-13/14), raw/chipmusic/lsdj-wav-instrument-tutorials-t14721.txt (12 posts, 2014-07), raw/youtube/defensemech-lsdj-wave-channel-ramble.txt (40:47, 2020-04-08 — the description with the frame chart is author-written, the transcript auto-generated). Recorded in raw/sources.md.
• 3 new summaries: s-chipmusic-lsdj-wav-channel-release, s-chipmusic-lsdj-wav-instrument-tutorials, s-defensemech-wave-channel-ramble.
• Ripple, 12 pages. instrument-envelopes — new section "LSDj: the wave channel has no envelope — you draw one" (frame envelope, F01 per row, 16 steps per synth, F10 chains two synths, the additive F, the synth-screen overwrite). lead-articulation — new ### LSDj (wave channel): loud first frame as attack, slurring by leaving the instrument number off, flutter tonguing with F01/FFF from a second table started by A, noise-burst breath, varied V speeds; the page had been sample-tracker-only. sidechain-pump — the LIMIT ramp (0→F across the synth, LOOP with the loop point at the end) as the wave channel's real duck, since E jumps 50 %→100 %. lsdj — the LENGTH → generated waves chart as a table (a fact the manual does not state; Defense Mech wrote it hoping it would reach the manual). filter-programming — "the filters are emulated, this is a wavetable synthesizer"; the whole harmonic series walked up CUTOFF by ear (90 major second … C0 fifth, then nothing) and the square's missing even partials. wavetable-programming — hand-drawn banks and their hazards. harmonic-series, ghost-channel (the synth screen wipes a hand-drawn bank — the hazard that matters most there), volume-slides, stereo-panning (M + hard panning = 8 volume steps for one channel), instrument-design, game-boy-apu (the wave volume is a bit shift: 3 and 2 bits of resolution at the low settings — nitro2k01), plus chipmusic-org, defense-mechanism, tyler-barnes, nitro2k01.
• Contradiction recorded, not smoothed: jefftheworld gives the wave E steps as 0/25/75/100 %; the manual, Pan Docs and Defense Mech all say 0/25/50/100 %. The wiki follows the manual and the summary says why (s-chipmusic-lsdj-wav-channel-release To verify).
• Scout lead noted for later: roboctopus's WAV-channel article series (noisechannel.org, being ported to tumblr in 2014).
• Index: three summary lines under LSDj course and forum threads. tools/lint.py clean (634 pages, no broken links, no orphans).
• Scout written: inbox/scout-lsdj-advanced-techniques-2026-08-31.md — 7 rows. Deliberately written before ingesting the Learning Lab playlist, because 12 videos is a collection to triage, not one source: rows 2–4 list every episode with runtime and id so the cost per pick is visible. Finds worth naming: roboctopus's LSDJ & You survives on tumblr (noisechannel.org is a splash page now) with a hand-drawn wave-channel PWM recipe — pulse-width-modulation has no Game Boy section at all — and Pain Perdu's six-part Hidden LSDJ Tricks (2016), covering retriggering, sidechaining, grooves and workflow, all with auto-captions. Blocked/dead sources recorded so they are not re-checked: littlesounddj.fandom.com (403/402), wiki.littlesounddj.com, noisechannel.org, the Box song file.

2026-08-31 — ingest: roboctopus's LSDJ & You + Pain Perdu's Hidden LSDJ Tricks 1, 4, 6#

• New raw: raw/roboctopus/ (4 articles, via the tumblr API — noisechannel.org is a splash page in 2026) and three raw/youtube/painperdu-hidden-lsdj-tricks-* transcripts (EP1 20:31, EP4 10:40, EP6 27:56). In raw/sources.md.
• Tool fix: tools/yt-transcript.py asked for --sub-langs en.*, which pulled every machine-translated English track and hit HTTP 429. It now requests <lang>-orig,<lang> only, tolerates partial failures, reports the video's spoken language in the header, and refuses to hand over a machine translation of a machine transcript unless --allow-translation is passed (Pain Perdu's EP4 looked French in --list-subs; the metadata's en-orig proved it is spoken in English).
• 7 new summaries: s-lsdj-and-you-collapse-of-the-waveform, s-lsdj-and-you-pwm, s-lsdj-and-you-echo-chamber, s-lsdj-and-you-supersaw-bass, s-painperdu-01-instrument-retriggering, s-painperdu-04-sidechaining, s-painperdu-06-grooves.
• 2 new entities: roboctopus (the hand-drawing school; the LSDJ & You column, 2013–2014) and pain-perdu (with the six-episode table — 2, 3 and 5 left as scout candidates).
• Ripple, 8 pages. pulse-width-modulation — new #### Hand-drawn PWM on the wave channel (16 waves narrowed by hand, SELECT+B B / SELECT+A to copy between frames, PINGPONG + REPEAT F + slow SPEED; the audible stepping and where it hides). fake-echo — the typed LSDj echo and its two-position lag (the echo after B is the C), panned and table-panned echoes, the pulse-2 chorus echo at FINE TUNE 2. retrigger — "the instrument number is itself a retrigger": envelope, table and wave cycle, AUTOMATE stepping per instrument value, and the kick-interrupts-the-lead fix. shuffle-funktempo — tempo is not BPM, the hourglass, per-channel grooves with matched tick totals, one-value grooves, groove as a per-step tempo map, triplet arithmetic, G in tables (Joe Toye's trick, which the wiki had from Mrwimmer without the attribution). sidechain-pump — Pain Perdu's full method. instrument-design — "shape is timbre" with three drawn instruments and the mixing note (keep pulse under A8). stereo-panning, instrument-tables (new section on the two clocks a table can run on).
• Index: two people, seven summaries. tools/lint.py clean (643 pages, no broken links, no orphans). Scout rows 1 and 2's picks are done; EP2/EP3/EP5, the Learning Lab shorts and the save-file walkthroughs stay open.

2026-08-31 — ingest: Pain Perdu's Hidden LSDJ Tricks EP2, EP3, EP5 (the series completed)#

• New raw: raw/youtube/painperdu-hidden-lsdj-tricks-{02-wav-channel-walkthrough,03-isometric-sav-walkthrough,05-shortcuts-and-workflow}.txt (47:42 / 54:48 / 31:02, auto-captions). raw/sources.md row rewritten to cover all six episodes.
• 3 new summaries: s-painperdu-02-wav-channel-walkthrough, s-painperdu-03-isometric-sav-walkthrough, s-painperdu-05-shortcuts-and-workflow. pain-perdu rewritten — all six episodes summarised, songs named.
• Ripple, 10 pages. tracking-workflow — the clipboard (and the SELECT+A cut/paste trap that overwrites it), slim vs deep cloning, "everything you can highlight you can copy" including waveforms, the cross-song clipboard carrying values only, transpose-column over cloning, KEY DELAY/REPEAT. live-tracking — new "Preparing a save for the stage": marking with B B B, empty rows so blocks loop back on themselves, transposition applied on A release, the select+start row stutter. wavetable-programming — the instrument side (LENGTH/REPEAT/SPEED, SPEED = the groove's ticks), the MANUAL myth corrected (hand-drawn waves work in ONCE/LOOP/PINGPONG too), F bleeding into the next synth, and waveform copy-paste. chord-arpeggio — two-tick four-note arp tables, the five-note chord from a non-root bass, and two pulse channels on the same table at different duties offset by three ticks. instrument-design — the noise "explosion", S as a relative shape sweep driven by a hop, K trimming successive kicks, tuned snares, "go quiet to go loud". arrangement — the hand-written tempo curve (100 → 150 → 180 → 80) and empty phrases as structure. fake-echo — the one-channel three-note delay with L/R instrument pairs. vibrato — A+UP/DOWN inverts vibrato phase; fast V on a raspy wave instrument as a flanger (and its CPU and per-pitch caveats). sidechain-pump, filter-programming (Q as the audible "sonic cursor").
• Cross-source resolutions recorded: EP2's unanswered on-camera question — which physical waves a LENGTH-3 synth plays — is answered by Defense Mech's 2020 frame chart (0 5 A F), and both are cited on the same pages. EP2's "PHASE 1F is silence" is true for 2016 but not for 9.x, where 1F "is not total silence as it used to be" — kept as a version difference on both summaries rather than smoothed.
• The scout's rows 1 and 2 are now fully ingested; open: the Learning Lab shorts (row 3), Defense Mech's tutorials and save-file walkthroughs (row 4), rows 5–6.

2026-08-31 — ingest: Defense Mech's Learning Lab shorts, the kick compendium and the 9.2.L noise tutorial#

• Looked at six, kept five (m64's standing permission to drop an item once seen). Skipped: Learning Lab – Adding Waveforms (4:14) — it is a demonstration in Audacity, not LSDj: sine waves mixed to show reinforcement, phase cancellation and a saw assembled from harmonics. No tracker content, no values, and ghost-channel / harmonic-series already carry the additive principle from three stronger sources. Recorded in raw/sources.md so it is not re-checked; no raw file saved.
• New raw: three Learning Lab shorts, defensemech-making-every-lsdj-kick.txt (2024-02-02, on LSDj 9.4.0) and defensemech-noise-channel-92L.txt (2022-06-05, 9.2.L).
• 3 new summaries: s-defensemech-learning-lab-shorts (the three clips together), s-defensemech-every-lsdj-kick, s-defensemech-noise-channel-92l.
• Ripple, 9 pages. wavetable-programming — LIMIT vs VOLUME (VOLUME scales the signal before the chain, so turning it down destroys the clip/wrap you liked; LIMIT constrains the range afterwards and keeps the shape), the VSHIFT rule the wiki had been missing (it is to DIST what Q is to the filter: useless with CLIP, interesting with FOLD/WRAP), and Hypnogram's multi-synth long filter sweep. instrument-design — every kick by channel: L80 as the kick that cannot wrap, transient tuned by the starting note, the groove-split 808 (3 9 / 3 9 6), the PU1 hardware sweep, the R80 gabber kick reaching ~F1, SAFE-pitch noise kicks and the hi-hat hidden in the kick table (nintendo kwondo). retrigger — R with a volume digit as the noise channel's only dynamics control, and the 9.4.0 change that makes R reset the pitch. vibrato — the self-resonating V bass. filter-programming, pulse-width-modulation (VSHIFT+FOLD as a poor man's PWM), hypnogram, defense-mechanism, lsdj (a note that 9.4.0 exists and what it changed — the version history otherwise stops at 9.2.6).
• An open contradiction moved forward. game-boy-apu had recorded that the within-mode noise mutes of 2021 "have no hardware explanation in the sources" and that BGB/SameBoy did not show them. The 2022 tutorial supplies an account — the LFSR "sometimes randomly produces zeros, and that results in silence" — which is not Pan Docs' all-ones lock-up, and reports the emulator situation reversing (current SameBoy and BGB reproduce the cutout). Both accounts now stand side by side on the page, per m64's rule.
• Scout row 3 is closed apart from the clips judged redundant; row 4's save-file walkthroughs and interviews remain open, Telemelt deliberately not pursued.

2026-08-31 — ingest: LSDj kits, the speech instrument and live mode (m64's pick from scout row 5)#

• New raw: raw/little-scale/how-to-prepare-samples-and-create-lsdj-kits.txt (the 2008 tutorial the scene still points at — found by following the link in thread 258, which is itself a pointer and was not saved) and four chipmusic.org threads: kit creation (t12080), kit volume (t10868), speech synthesis (t9199), live performance (t530).
• 4 new summaries: s-little-scale-lsdj-kits, s-chipmusic-lsdj-kit-threads (the two kit threads together), s-chipmusic-lsdj-speech-synthesis, s-chipmusic-lsdj-live-performance. New entity: little-scale (Sebastian Tomczak).
• Ripple, 7 pages. chip-samples — the sample preparation chain the wiki was missing: the 3FA0-byte kit budget, cropping as the discipline (500 ms → 75 ms in the worked example), 11,468 Hz and its 5,734 Hz Nyquist lowpass as the cure for kit hiss, DC-offset removal, edge fades against clicks, and the two opposed volume answers — little-scale normalising then adding 3–6 dB of deliberate clipping, versus the 2013 consensus of mixing the other channels down to about 7X; plus the 2× speed-up / SPEED 0.5 trick for long samples and the patcher-preview click that is not on hardware. lsdpatch — the byte budget and the pointer to preparation. lsdj — SPEECH in practice: unforgiving ("you can't delete wrong letters"), scope it to a phrase, write the allophones in STSPEECH on an Atari ST emulator first, keep it centred in the mix. live-tracking — new "What live mode actually adds": the one unique thing (starting and stopping channels), per-channel chain queueing, chain-loop islands built with empty rows, BR1GHT PR1MATE's vamp sections and resultant melodies, H polyrhythms, live Q sweeps, the sync-to-slave drone as both trick and hazard. stereo-panning, chipmusic-org, roboctopus.
• Contradiction recorded: herr_prof gives the Game Boy sample rate as 8192 Hz where LLTH cites the manual's 11,468 Hz; the wiki follows the manual and the summary says so.
• Scout row 5 is now closed. tools/lint.py clean (654 pages, no broken links, no orphans).

2026-08-31 — scout: little-scale's blog (m64's request)#

• Enumerated the whole blog through the Blogger feed — 2907 posts, 2006–2024 — and saved the index (date, title, labels, URL) as inbox/little-scale-post-index-2026-08-31.tsv so it never needs re-fetching. Wrote inbox/scout-little-scale-2026-08-31.md: 8 rows, nothing ingested.
• Shape of it: chipmusic 668 posts and chiptech 260 are this wiki's territory; Ableton Live (471), Max for Live (330), Max/MSP (309), Arduino (222) and Teensy (159) are not.
• Best finds: a self-contained 2009 cluster on Game Boy microtuning and under/overclocking (LSDj tuning tables with separate GB/GBC/GBA and Super Game Boy tables, LSDj at 15-TET, credits to Abrasive and nitro2k01) — the wiki has nothing on non-12-TET chip music; the 2008 Game Boy sync series (ten-plus posts, from a ~AU$15 MIDI-clock box to fifteen Game Boys off one clock), which would give sync its missing Game Boy half; All Classic LSDJ Kits as WAV Files plus a Loud TR808 kit from 2014, which is precisely the stock-kit remaster herr_prof wished for in 2013 (s-chipmusic-lsdj-kit-threads); and ~24 Creative Commons chip sample packs, two of them chips the wiki documents (SID 8580, SP0256-AL2).
• Caveat recorded in the scout and on little-scale: his tables, schematics and waveform shots are images, so a text ingest gets the method and the provenance but usually not the numbers.
• Plan written for a fresh session: inbox/plan-little-scale-2026-08-31.md — read order, phases, definition of done. It records a capability verified rather than assumed: little-scale's data lives in images, and they can be read by fetching at full size (swap /s400/ for /s1600/ in the Blogger image path) and opening them with the Read tool. The test image reframes candidate 1: the "LSDJ Tuning Tables" post is a hex-editor screenshot of lsdj 3 8 9.gb ("216 bytes selected at offset 27,962"), i.e. the pitch table as 16-bit period values around $6D30 — so microtuning LSDj is a ROM-patching technique, not a settings one. Images will be saved under raw/little-scale/img/. Start with phase 1 (microtuning); then the 2008 sync series, kit archaeology, and the sample packs as one summary.

2026-08-31 — ingest: little-scale phase 1, Game Boy microtuning and clocking (plan inbox/plan-little-scale-2026-08-31.md)#

• The plan's image route works, and paid for itself. Every number below was read out of a picture at /s1600/ with the Read tool; three of the images needed a sips crop-and-upscale before the hex was legible. The images are saved under raw/little-scale/img/ and described per-file in raw/sources.md.
• New raw: seven posts under raw/little-scale/ — lsdj-tuning-tables.txt, abrasive-on-lsdj-microtuning.txt, underclocking-game-boy.txt, retuning-underclocked-game-boy.txt, tune-game-boy-like-japanese-koto.txt, nanoloop-tuned-to-18-tet.txt, microtonal-chipmusic-compilation.txt — plus eight images. The scout's URL for the Nanoloop post 404s; the real one is /2009/03/18-tet-18-divisions-of-octave-instead.html, corrected in raw/sources.md and the scout's row 1. The compilation post was added to the plan's six because it is where little-scale states the point of the whole cluster.
• 6 new summaries: s-little-scale-lsdj-tuning-tables, s-little-scale-abrasive-lsdj-tune, s-little-scale-underclocking-and-retuning (the two underclocking posts as one, as the plan allowed), s-little-scale-koto-tuning, s-little-scale-nanoloop-18-tet, s-little-scale-microtonal-compilation.
• 2 new pages instead of the planned one. The material split cleanly along CLAUDE.md's own templates, so the theory went to the concept page alternative-tunings (cents, EDOs, just intonation, why a chip's note grid is software, the 64 Hz floor and the coarse top, four ways off the grid) and the how-to to a new technique page rom-retuning (the offsets, the arithmetic, hand-patching, lsdj_tune, Nanoloop, underclocking, pitfalls). 2 new entities: abrasive and nanoloop — the latter was overdue, being named in seven summaries with no page.
• What the images gave that the prose did not. The two LSDj screenshots' status bars read "216 bytes selected at offset 27,962" and "…at offset 28,178 out of 1,048,576 bytes" on lsdj 3 8 9.gb: two 216-byte pitch tables back to back at $6D3A and $6E12, 108 notes each, the second recomputed for the Super Game Boy — which is why the KiGB bug (nitro2k01 checked it) matters and why a retune must patch both. The Nanoloop shot gives nl13d edit.gb, "144 bytes selected at offset 29,450 out of 32,768". The koto spreadsheet gives the ratios, the frequency and data-value columns, the DEC2HEX formula, the 2-3-2-3-2 mapping of five degrees onto twelve keys, and the red rows below 64 Hz where the register value goes negative.
• Everything numeric was cross-checked before it went on a page. Decoded through Pan Docs' f = 131072 ÷ (2048 − v) (game-boy-apu), LSDj's stock table is 12-TET from C2 (65.41, 69.28, 73.39 …); the SGB table gives the same pitches against a clock 2.4 % faster, and its first entry is 00 00 because C2 is below what that clock can express. The 108 replacement words printed in the retuning post count to exactly 216 bytes and decode to 12-TET at a 4 MHz master — i.e. an exact octave down on a 2 MHz machine, in tune. The 18-TET Nanoloop values have a mean successive ratio of 1.0393 against 2^(1/18) = 1.03926. The koto sheet's hex (3E, 1CB, 22F, 358, 3DD, 41F) matches its own frequencies.
• Ripple, 9 pages. game-boy-apu — "the note grid is software": the tables' addresses and sizes, and the two ends of the range now shown with real values (the top of LSDj's own table repeats period values across semitones). game-boy — a new Underclocking: which crystal section distinguishing 2.097152 MHz (exactly half, no patch) from the 2.000000 MHz crystal shops actually sell (an octave and 82 cents, hence "some disgustingly out of tune B"), and the SGB's second pitch table. lsdj — an Inside the ROM entry with the offsets, the signature bytes and the version warning. nitro2k01 — three new facts: the KiGB double-check, his own just-intonation LSDj retune ("it gives each scale it's own character"), and the "elusive 2.097152mhz crystal". little-scale — a new section on the 2009 tuning cluster. scales-and-modes — scales that are not made of semitones, and the warning that 0 2 4 5 7 9 11 only means "major" on a 12-TET grid. semitone-math — cents, and the per-step arithmetic for n-EDO. harmonic-series — the series is just intonation. sync — the forum's "free sub-octave" claim corrected: in tune only with an exactly halved clock or a rebuilt pitch table.
• Contradictions and gaps recorded, not smoothed: little-scale's frequency formula uses the piano-key convention (Note − 49) while calling the column "MIDI notes" (s-little-scale-nanoloop-18-tet); his 144-byte / 72-entry Nanoloop table against "rows 27 – 90" (64 rows); "a C3 will play back on the unmodified Game Boy as a C2", whose reading is reconstructed from the arithmetic; and the 2010 commenter whose 3.8.9 table bricked a 4.0.5 cartridge, now a pitfall on rom-retuning.
• python3 tools/lint.py clean: 664 pages, no broken links, no orphans, no frontmatter issues, index complete.
• Plan phases 2–4 (the 2008 sync series, kit archaeology, the sample packs) remain open; inbox/plan-little-scale-2026-08-31.md updated with what was done.

2026-08-31 — ingest: little-scale phase 2, the Game Boy sync series (2006–2011)#

• The cluster was bigger than the scout's nine posts. Filtering the TSV index for Game Boy sync turned up fifteen, including two the scout missed that matter: syncing-nanoloop-to-midi-clock (2008-05-06), which is where the 12 ppqn figure comes from, and overclocked-game-boy (2008-05-01), which partly answers a To verify left open by phase 1.
• New raw: fifteen posts under raw/little-scale/, plus nine images — only the data-bearing ones (the device chart, the link-plug pinout, five schematics, the parts photo). The 20 step-by-step soldering photos of the AU$15 build and the finished-unit shots were deliberately not saved; recorded in raw/sources.md.
• 6 new summaries: s-little-scale-sync-options-for-game-boys, s-little-scale-arduino-gameboy-sync-boxes (five posts of one project line), s-little-scale-1-dollar-nanoloop-sync, s-little-scale-teensy-usb-midi-sync, s-little-scale-syncing-things-to-lsdj (LSDj as master, plus the 2006 un-sync piece), s-little-scale-overclocked-game-boy. Skipped: game-boy-pushpin-midi-interface (2008-06-02) — a build note on someone else's schematic; its only wiki-relevant facts (Pushpin exists, Game Boy Color only, full MIDI control, nitro2k01 published the interface) are already carried by the census summary. Saved in raw/, cited from nitro2k01, not summarised.
• The deliverable the plan asked for: sync now has a real Game Boy half — what is actually on the wire (the pin table, the plug pinout with wire colours, the ppqn rates), driving LSDj from a DAW step by step, the 2008 device landscape, making the Game Boy the master, and deliberate drift.
• The fact the wiki most needed: lsdj and nanoloop listen on different pins of the same connector — LSDj on SC (serial clock, with a 22k–100k pull-down), Nanoloop on SI (serial in, plus a series resistor for 2.x). That is why every sync device has separate LSDJ and Nanoloop columns, and it was stated in exactly one schematic note in 2008 and confirmed on different hardware in 2011.
• Read out of images: the 18-device × 11-column comparison chart (transcribed in full on the summary and in shape on sync) and the link-plug pinout 1 VDD35 no wire · 2 SO red · 3 SI orange · 4 SD brown · 5 SC green · 6 GND blue. From the LED schematic, the divider tap is 4040 Q5 (÷32) into a 4017 wrapped at Q2 — which makes LSDj's master line 32 clock transitions per sixteenth-note step (inference, marked as such).
• Ripple, 10 pages. sync (the new Game Boy section), lsdj (an On the wire paragraph: SC, master-mode rate, 35–200 BPM, WAIT, and that it does not resume after a stop), nanoloop (SI, 12 ppqn, Wittchow's cable, and that nanoGroover existed because Nanoloop could not do triplets), game-boy (the plug pinout; 8 MHz added to the crystal table; a ProSound entry), nitro2k01 (he fact-checked the chart, and his Pushpin schematic), little-scale (a new section on 2006–2011), live-tracking (the Game Boy as the rig's clock, and as its drift), sid (a new Clocked by a Game Boy section — the C64 filter and nanoSync24), midi-in-trackers (a Game Boy section), rom-retuning + recording-samples.
• A phase-1 To verify moved forward. Asked in 2009 whether his replacement pitch table would work at 8 MHz, little-scale said "in theory… because I haven't actually tried it out with an overclocked GB" — but he had built an 8 MHz DMG in May 2008. The overclock is now documented (an octave minus 82 cents, the 2 MHz error mirrored) and rom-retuning states the rule; the table question itself stays open, and the summary says so.
• Contradictions recorded, not smoothed: the link-cable wire colours here disagree with nordloef's DMG-07 colours already on sync (brown and yellow vs brown and orange; red, green and blue agree) — different cables, neither corrected; electri-fire's unanswered objection that the $1 circuit's 4017 has its clock tied to 5 V and MIDI on CLOCK INHIBIT, against little-scale's "I have it working in a very stable fashion"; and the chart cell where nitro2k01 said "Arduinoboy+mGB: The LSDj notes column says yes", little-scale replied "it didn't. trust me.", and the saved image says yes.
• python3 tools/lint.py clean: 670 pages, no broken links, no orphans, index complete. Site rebuilt.
• Plan phases 3 (kit archaeology) and 4 (the sample packs) remain open.

2026-08-31 — ingest: little-scale phases 3 and 4 — the kits and the chip sample packs (plan complete)#

• Filtering the index again paid off, in both directions. Phase 3's scout list had 3 posts; the TSV has 11, including the whole 2008–09 kit run (Street Fighter, Debussy, Chopin, Mozart, Scriabin, Ellington) that the 2013 forum threads were still naming. Phase 4's sample-pack label also turned up an SN76489 technique post and a Max/MSP piece about the wave channel that no list had mentioned.
• New raw: 26 posts under raw/little-scale/, one image. Deliberately not fetched: any of the .zip kits and sample packs (they are the material, not the documentation), and the ~35 non-chip "Completely Free Sample Pack" and impulse-response posts of 2012–2019 (ukulele, toy keyboards, Roland JV-30, VCS3, Pocket Operator, Erica Synths…) — listed from the TSV, outside the wiki's focus, recorded in raw/sources.md so they are not re-checked.
• 4 new summaries: s-little-scale-lsdj-kit-releases (phase 3, eleven posts as one), s-little-scale-chip-sample-packs (phase 4, twelve packs as one, with the SID 8580 and SP0256-AL2 called out as the plan asked), s-little-scale-rom-sample-extraction and s-little-scale-all-4-bit-waveforms.
• The best thing in phase 3 was a screenshot. All Classic LSDJ Kits as WAV Files is a bare download link, but its image is a Finder window — "22 items" — giving LSDj's complete stock-kit list: AA-EL ACIEED ANIMLZ CR-78 CR8000 DMX DR-55 DR-110 DRUMUL EL-KK2 GHETTO KK3-TH KR-55 LINNDR RHYACE TOM TR-606 TR-707 TR-727 TR-808 TR-909 TT-ZZ. s-lsdj-manual had only a partial list ending in "…"; this completes it, and it is now on lsdj and chip-samples.
• herr_prof's 2013 wish, and the honest version of the story. He wanted "a remaster of all the stock kits… they really are quiet for no reason"; little-scale posted a Loud TR808 in April 2014. The 2014 post does not mention the thread, so the summary records the sequence and explicitly declines to claim a causal link.
• Phase 4 turned out to be documentation, not audio. Every pack post states the recording method, the coverage, the naming convention, the format and the licence, and those are worth more than the samples: sounds he generated on the hardware ("not sounds or samples recorded from games or demo programs"), recorded dry before the console's output filter, normalised and trimmed but never EQ'd. Counts: 1344 Atari 2600 TIA (3 envelopes × 14 waveform types × 32 pitches), 945 SN76489, 316 POKEY noises, 245 YM2413, 112 SAA-1099, 72 YM2612 across its eight algorithms, 59 SP0256-AL2 allophones. The licence moved: CC BY-NC 3.0 in 2008–09, relicensed "copyright-free and license-free" by March 2010, "restriction free" by 2012.
• Two findings nobody was looking for. (1) The SN76489's noise channel is a bass voice: it has no frequency divider of its own and can be clocked from pulse channel 3, so "the output noise frequency is octaves lower than the current pulse three frequency" — you spend channel 3's pitch and channel 4's volume to get one bass line. sn76489 added to the index's Wanted pages rather than writing a thin entity page. (2) The wave channel's design space is 16^32 ≈ 3.4 × 10^38 waveforms; little-scale's patch auditioning 100 a second would take "1,079,028,307,100,000,000,000,000,000 centuries" — his figure, checked here and correct.
• Ripple, 8 pages. chip-samples (a new Where to get chip sounds without the chip section: the packs, the stock-kit list, and that piano survives the 11.468 kHz crush better than expected), sample-ripping (a new Getting audio out of a ROM section with the Audacity settings — signed 8-bit PCM, mono, 0 offset, 100 %, 12000 Hz — and the trick of finding audio by eye), wavetable-programming (the 16^32 figure, and why it is the argument for recipes), lsdj (the 22 stock kits), sid (a Sampled section on the 8580 pack and what it deliberately omits), little-scale, game-boy (Pocket Music's 100 samples), nanoloop.
• An observation recorded as an observation. LSDj's SPEECH instrument builds from 59 allophones; the SP0256-AL2 has exactly 59 sounded allophones. No source connects them, so it is stated on lsdj and in the summary as (unverified) rather than as a finding.
• python3 tools/lint.py clean: 674 pages, no broken links, no orphans, index complete. Site rebuilt.
• inbox/plan-little-scale-2026-08-31.md is now complete — all four phases done. Open on this blog and left in the scout: ProSound measurements (row 5), tuning the Sega and Atari chips (row 6), C64 hardware interfacing including the Prophet 64 sync posts (row 7). Row 8 stays out of scope.

2026-08-31 — plan written: consolidation (m64 asked what to do next, in a fresh session)#

• Surveyed what is open and measured the wiki's shape rather than recommending by feel. At 674 pages: 363 summaries, 248 entities, and only 61 technique + concept pages — six source pages for every page a musician actually reads. The imbalance shows in the wrong place: the thinnest reader-facing pages sit on the best-sourced topics (octave-bass 3.4 KB against 19 summaries that discuss basslines; multisampling, sidechain-pump, song-init-and-looping, sample-offset all under 5.2 KB; the theory pages still resting on s-037-lab plus one or two TraxWeekly articles although 909 TraxWeekly articles are in raw/).
• Also counted the backlog: 49 To-verify bullets on technique/concept/entity pages, 177 (unverified) markers across 107 pages. Sampled them — a real fraction is now answerable from sources ingested since they were written, which is how this session closed the 8 MHz question on s-little-scale-underclocking-and-retuning.
• Recommendation: stop ingesting for one session and harvest. Written up as inbox/plan-consolidation-2026-08-31.md — four passes (enrich the thinnest reader pages from the summaries that already exist; the two or three genuinely missing pages, sound effects being the clearest at 34 summaries and no page; triage the To-verify backlog into resolvable / correctly-flagged / needs-a-source; then contradictions, duplicates and stale claims). It touches no raw/ files and needs no new sources. The alternatives, if m64 would rather ingest, are listed in the plan in order of value.
• Index Inbox section refreshed: the little-scale plan is marked complete, its scout marked rows 1–4 ingested / 5–7 open, and the advanced-LSDj scout's status corrected (it still said "nothing ingested yet" for both).

2026-08-31 — consolidation: harvest the summaries into the pages people read (inbox/plan-consolidation-2026-08-31.md, all four passes)#

No new sources. Nothing in raw/ touched. Everything below came out of summaries that were already here.

Pass 1 — the thinnest reader pages, all ten enriched. Sizes before → after: octave-bass 3.4 → 9.1 KB · multisampling 4.6 → 11.1 KB · sidechain-pump 4.1 → 9.7 KB · song-init-and-looping 4.1 → 8.0 KB · sample-offset 5.2 → 7.0 KB · sample-looping 6.0 → 8.7 KB · semitone-math 4.6 → 8.7 KB · melody-writing 4.7 → 7.7 KB · diatonic-chords 4.7 → 7.5 KB · chord-inversions 3.0 → 5.2 KB. What the harvest actually turned up:

• CC0 plays 0, C, 0, C… — LSDj has a one-command octave bass and the wiki did not say so; the manual's neighbouring values (C0C, CCC, C00) went onto octave-bass and into s-lsdj-manual's command line.
• The table-octave pitfall, from SID Factory II, generalises: a wave table's semitone offset must sit on the short attack rows, because "slide/portamento/vibrato only work while the offset is 00" (source: s-chordian-sf2-instruments). Now on octave-bass and chord-inversions.
• Multisampling is not only a PC-tracker idea. An LSDj kit instrument names two kits — "the first kit will be used in the left note column … the second kit will be used in the right" — 15 slots each inside 3FA0 bytes; and the C64's Digi-Organizer chains (start page, end page, rate) triples. Both are note-to-sample maps; multisampling gained platforms: [pc-tracker, gameboy, c64]. little-scale's packs supply the sampling density question, answered two ways: every C and G per octave for the SID and YM2413, every one of the 83 or 32 playable pitches for the SN76489 and the TIA.
• The unverified C64 section of sidechain-pump is now sourced. The SID has no per-voice volume — one global 4-bit $D418 nibble, "4 bit precision = 16 steps" — so ducking is either global (sid-wizard small-FX $Ax and orderlist $A0–$AF, goattracker DXY) or per-voice through the ADSR sustain ($5x/$6x, 5XY/6XY). And a real pitfall fell out of it: on a 6581 a fast $D418 pump is the digi trick — the register leaks a DC offset, so a per-frame envelope buzzes.
• Chip trackers do not reset state at the loop, and the drivers agree: GoatTracker's orderlist transposes are "reset only at song start"; Deenen's Maniacs of Noise player restarts the sequence without resetting the transpose; Bjerregaard's $FF restarts all three voices at once; the Dominator player fades on $FE and repeats on $FF. song-init-and-looping's (unverified) chip paragraph replaced by those four cases and one rule: re-state sticky state at the loop target, not at the end.
• A cross-tool table that only this wiki could write: where the minus sign lives in a semitone offset — SID-Wizard $E0–$FF down, GoatTracker 60–7F down, LSDj two's complement (F4 = −12), IT Jxy up-only, Deenen's driver up-only 0–31, Bjerregaard's biased at $C0 = 0. With the warning that SID-Wizard and GoatTracker agree on $00/$80/$81–$DF and disagree on the down range, so a table copied between them transposes wildly.
• Smaller: LSDj Fxx on a kit is a relative, backwards-capable sample offset (80–FF steps back) and Sxy moves the loop points live; IT's Mxx/Nxy channel volume is the closest thing a tracker has to a sidechain bus; three-note chord commands (Cxy, Jxy) cannot spell a seventh, which is a real constraint and now stated on diatonic-chords; melody-writing gained an "On three voices" section (Hubbard's fills in the melody's rests, Galway trading drums for melody channels, Huelsbeck's checklist).

Pass 2 — the missing reader pages, both written.

• sound-effects (14.2 KB, new). 34 summaries mentioned SFX and no page owned it. The finding is that four independent C64 drivers converge on the same design: an effect is an instrument program plus a frequency, with no note and no duration, because the instrument's own table decides when it ends. Galway's Arkanoid says it outright ("the sound instruments table is a perfect copy of the instrument table, but with two extra bytes at the end"), Fred Gray's Mutants adds a next effect byte so effects chain, Maniacs of Noise and JCH both ship a separate SFX player, and SID-Wizard exposes the same triple as an API at LoadAddress+12 (X note, Y instrument, A length). Martin Walker's Hunter's Moon engine gives the recipes — six pitch-ramp modes, byte-swapped frequencies, EOR scrambles, oscillator-3 noise added to pitch, ring-modulated voice pairs — in about 250 lines, and costs the in-game music.
• levels-and-headroom (11.6 KB, new). The material was scattered across the kit-volume thread, the ghost-channel threads, the wave-synth articles and the 1996–98 mixing columns. Its spine: on a chip the ceiling is a 4-bit number, so lowering a voice costs resolution, not just level (three of the wave channel's four volume steps discard bits), every additive trick is a loan against the mix, and the correct move is almost always to turn the others down. Plus LIMIT-versus-VOLUME with the calibration numbers, normalise-before-bit-reduction, and Zinc's monitoring checklist.
• sn76489 left in Wanted pages, as the plan asked — it should come from an ingest, not from nothing.

Pass 3 — the To-verify backlog. Four markers removed (178 → 174 on the metric as it stood). A note on that metric: it was counting log.md and index.md, which talk about the markers rather than carry them, so every session's write-up inflated it. tools/lint.py now excludes those two bookkeeping pages, and the honest figure for the wiki's content is 153 markers across 101 pages — the plan's 177 was about 24 too high for the same reason. The movement matters more than the count either way:

• Resolved from sources already in raw/: DMC = "Demo Music Creator", stated in the Brief history of SID lecture, which also places it in the "DUR editor" lineage from UBIK's Music (1987). VICE = "the VersatIle Commodore Emulator", spelled out in a 2017 defMON setup guide. cRSID is Hermit's — the DUET research notes name libcRSID as "Hermit's C64+SID emulation". And the 59-allophone "coincidence" is not one: LSDj's manual, Appendix A, names the SP0256-AL2 as the machine behind its speech kit, so s-little-scale-chip-sample-packs's recorded observation is now a sourced fact.
• Narrowed rather than closed: the IFF/8SVX question on recording-samples — OpenMPT loads "Amiga IFF (8SVX / 16SVX / MAUD)" and ProTracker saves "IFF-loops", so Cowboy's "IFF only supports 8kHz mono" is about his converters; what is still unsourced is narrower (does an 8SVX chunk state its own rate). Same treatment for x64sc: named in the wild as a much heavier VICE binary, still not stated to be the cycle-exact one.
• Left alone, correctly flagged: the cent arithmetic on harmonic-series and semitone-math, and the entity bullets that genuinely need a new source (Nanoloop's version history, siddump's author, Odin Tracker's provenance) — each now says what would settle it.

Pass 4 — contradictions, duplicates, staleness.

• A real error found and fixed on harmonic-series. The partial table's note names are internally inconsistent: with A4 = 440 and octave numbers changing at C, 660 Hz is E5 and not E4, 1100 Hz C#6 and not C#5. The source's names are kept (they are what it says) with the arithmetic stated under the table and a second reason not to lean on them — LSDj's own octave display shifted by one between versions, which is why lsdj_tune for 3.9.9 takes --base A5 440 where version 6 shows the same pitch as A4.
• The 8192 Hz / 11,468 Hz disagreement is now on the reader page, not only in the summary: the thread's own author corrects himself and cites the manual, so the wiki follows 11,468 — while noting that the 9.2.6 manual prints no rate at all, so the citation runs through an older edition.
• Checked and found already handled: the two link-cable wire-colour tables on sync (both stated, neither corrected, "check with a multimeter before you cut") and the LSDj version history (the "Later than the manual" note already carries 9.4.0). No duplicate entities — every near-match among the 248 slugs (lok/loki, matt-gray/matt-gray-player, the whole sid* family) is a genuinely distinct page.
• A fabrication caught in the act and removed. Adding drax to Wanted pages, the first draft gave his real name as Thomas Egeskov Petersen — which is laxity. No ingested source names Drax; the entry now says so explicitly.
• New standing check in tools/lint.py: unlanded ripples. A summary lists a page under "## Pages touched" but the page never cites that summary — the ingest stopped at the summary layer. 462 such claims across 133 pages, worst on orderlist (17), wavetable-programming (16), sid-player-routine (16), instrument-design (11), vibrato and detune (10 each). It is a review list, not a defect list ("Pages touched" sometimes means "relevant to"), but it is the sharpest measure the wiki now has of where ingests have outrun synthesis.
• Landed a first slice of it: opl-fm-instruments gained the JCH AdLib story (a November 1991 player with C64-style arpeggio tables and "register bypassing" — cache every OPL register write, skip unchanged values; the EdLib editor folding out one of nine channels; only two Laxity tunes ever converted from SID), and detune gained Pain Perdu's "drunk" unison and three-tick phasing, roboctopus's failed supersaw, and the fact that a MOD's finetune is a signed nibble in the sample record, not a channel property.

State. 676 pages (674 + 2 new technique pages); python3 tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete; python3 tools/build-site.py rebuilt (2131 files). Reader-facing pages: 61 → 63, and the ten thinnest are no longer the thinnest. Backlinks added to every summary harvested, so the citation runs both ways.

---

2026-08-31 — Unlanded ripples, pass 1: eight pages worked top-down#

Operation: consolidation (the follow-on the previous session's plan named — no new sources; nothing in raw/ touched).

inbox/plan-consolidation-2026-08-31.md was already done. Its closing note asked for the next session to work the unlanded-ripple list top-down, one page per sitting, the way Pass 1 worked the size list. Eight pages, all taken from the top of that list, each cross-checked against every summary that named it and each now citing what it drew from in sources: and ## Sources.

Total: 462 → 364 unlanded claims, 133 → 125 pages. tools/lint.py clean throughout.

What was actually sitting unharvested#

The list turned out to be a good detector, not just a bookkeeping metric. Three kinds of gap:

• Whole tools missing from a concept page. orderlist documented SID-Wizard, GoatTracker, LSDj, the 1980s drivers, the PC trackers and four SIDin formats — but not defMON's seqLIST or CheeseCutter's sequences, the two C64 editors in current use, both of which had summaries waiting. defMON's whole order language is one command (FF xx 00 jump, FF xx yy finite loop, target read from the next voice's column), plus a rule nobody would guess: a sequence that appears nowhere in the list is not saved to disk. The same gap on vibrato — CheeseCutter puts vibrato in the command table, and defMON has no vibrato at all on purpose.
• Forward references that were never landed. wavetable-programming said "the angular saw, the one-sample pulse and the notched sine are worth writing down" and then never wrote them down — the roboctopus hand-drawing posts had been summarised but never harvested. They are there now, with the two rules that make them teachable (volume is the wave's overall height; "the more angular the wave, the sharper its edges, the sharper and harsher the sound is") and the method that generalises them: google-image a waveform and approximate the shape.
• Primary research not reaching the reader page. m64's own DUET research into SID-Wizard 1.97's player.asm was cited on orderlist but not on sid-player-routine, which described the player types from the manual. It now carries the runtime: channels processed 3→2→1, the deliberate SR→AD→freq→PW→waveform write order ("keeping distance" between ADSR and waveform), the uniform two-frame delay from a row's tick 0 to the audible note that makes 2-frame hard restart possible, sweep counters ticking per player call rather than per frame, and the filter's leftmost-priority being emergent from the dispatch order rather than coded.

Findings worth keeping#

• The per-voice orderlist has an author and a date. JCH wanted "tracks to follow each other so that multiple voices could be compared easily" and disliked SoundTracker's voices "fused together in the same pattern"; v1 of his editor (Nov 1988) shipped with "infinitely long voices", v2 (Dec 1988) recoded them into flexible per-voice sequences. He states the difference again from the far side in 1995, on moving to FastTracker 2: "much of the same principle as my own music editors of the past, except the channels did not use separate sequences." That is the C64/PC split in one sentence, from someone who worked both. Now on orderlist and tracker.
• Two new sections that had no home before. orderlist gained LSDj: the song screen as a live instrument — per-channel chain queueing, and chain-loop islands (the live sequencer stops rewinding at an empty step, so blocks separated by blank rows each cycle on themselves), which Bit Shifter and Pain Perdu arrive at from performance and workflow respectively. sid-player-routine gained Loadable drivers: one editor, many players — CheeseCutter's .ct carrying its own player, SID Factory's v5.0x/v6.0x split (the fast one capping the spike at "around $18 scanlines"), Ninjatracker's many v2.x builds, and Cadaver's reason for all of them: "each time, when starting a new game project, I also wanted to rewrite the playroutine for memory or speed (rastertime) gain."
• The PSID/RSID contract was nowhere on sid-player-routine, although every exported tune meets it: song number in A at init, playAddress 0 meaning the init installs its own IRQ, the per-song speed bit choosing VBI or CIA, and RSID's whole point — an old emulator must refuse a tune that needs a real C64 rather than play it wrongly.
• A (unverified) resolved. s-chipmusic-lsdj-wav-instrument-tutorials flagged roboctopus's WAV-channel series as a lost scout candidate (noisechannel.org gone, tumblr mirror unknown). Four of those posts are now sources in this wiki; the bullet is replaced by a ## Resolved section naming them. Marker count net zero (153): one resolved, one added — the honest new one being that roboctopus's 2014 PWM stepping was never re-tested after LSDj 7's silky wave.
• Vibrato as a routine you had to write is now a section on vibrato, because three sources say it from three angles: Matt Gray ("even putting a vibrato on a sound needed a mod routine to do it. There was no button to push"), his own player's five state variables (VIBDELAY, VIBSTEP, VIBTIME, VIBD, VIBDIR), and the Maniacs of Noise driver packing delay + growing depth into two nibbles — which is a swell, the string player's vibrato, encoded in an instrument byte.
• A porting trap now stated: S3M's effect memory is global to the column, so a bare J00 recalls the last non-zero parameter of any command in that column — where MOD has no arpeggio memory and IT keeps it per command. And ProTracker 1.1A doubled its vibrato depth scale: "Double all your depth numbers if you load old protracker songs."

State#

676 pages. python3 tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete. Unverified markers 153 across 101 pages (unchanged, and correctly so). Unlanded ripples 364 across 125 pages, down from 462/133. python3 tools/build-site.py re-run.

Next, if the list keeps being worked top-down: the two big entity hubs lsdj (24) and sid (22) — both heavily read and both large jobs — then commodore-64 (10), sidid (9), shuffle-funktempo (8), multispeed (8), sid-format (8), instrument-tables (8), adsr-envelope (8), pitch-slide-and-portamento (7).

---

2026-08-31 — Unlanded ripples, pass 2: lsdj#

Operation: consolidation. No new sources; nothing in raw/ touched.

The largest single entry on the unlanded-ripple list: 24 summaries named lsdj and the page cited none of them. All 24 landed; 29.6 → 46.1 KB. Total now 364 → 340 unlanded claims across 125 → 124 pages.

Why so many, and what was actually missing#

Nearly all 24 are tutorial sources — pain-perdu's six episodes, roboctopus's four LSDJ & You posts, two Defense Mechanism videos, four chipmusic.org threads, four little-scale posts. The page was built from the manual and the intense-tech articles, so it had the documented behaviour of every screen and command and almost none of the behaviour that only shows up in use. Deciding what belonged here rather than on a technique page was most of the work: recipes went out as pointers, semantics stayed.

The single biggest gap was a core rule the manual only implies: the instrument number is the trigger, the note is not. A bare note plays with whatever settings are already running; the instrument number re-reads the settings and restarts the envelope, the table and the wave synth's cycle. Three consequences now on the page — omitting it is legato; a per-trigger table gives L R L R when every note is triggered and L L R R when only every second one is; and a slow PINGPONG wave cycle is never heard in full unless triggered sparsely, which is why kicks sharing WAV force you to size LENGTH × SPEED to the gap between hits.

Other tool-level facts that had no home:

• The synth's instrument side. LENGTH = how many of the sixteen waves the cycle uses, LOOP POS = how many end frames repeat, SPEED = ticks per frame and therefore a groove decision (SPEED 6 under groove 6/6 = one wave per step). MANUAL is the quiet mode because no frame change occurs. Hand-drawn waves are not restricted to MANUAL — "one of the most common misconceptions". F past a synth's sixteen waves bleeds into the next synth's. VOLUME scales before the chain (turning it down destroys the clip or wrap that made the sound) while LIMIT constrains after and keeps the shape — and there is no LIMIT command for the phrase screen, so different levels mean different synths. CUTOFF 00 is the low end, which is why a high-pass instrument at the default FF seems silent.
• The clipboard survives loading another song, but carries values only — chain 1B pasted into a second song points at that song's 1B, so it works properly only between songs from the same template save. Named by Pain Perdu as the easy way to build a remix.
• Live transposition applies on release: hold A and press arrows in the project screen and the new TRANSPOSE lands only when A is released, so a key change sits exactly on a bar line. Same for TEMPO.
• What live mode is for: per-channel queueing, not muting — "PU1 can be playing the chain at position 01 while PU2 plays the chain at position 2F" — and chain-loop islands, where the live sequencer stops rewinding at an empty step, so blocks separated by blank rows each cycle on themselves.
• Kit budget in numbers: 3FA0 bytes, "just over two seconds worth of audio per kit", 11,468 Hz / 8-bit / mono, three-character names. Plus two quirks — a longer sample can be sped 2× and played at kit SPEED half, and the patcher's preview lies (the click at the end of a long sample is not there on hardware).
• Two retuning caveats, from the worked koto retune: a ratio scale with fewer than twelve degrees maps several adjacent semitone keys to the same pitch, and every frequency below ~64 Hz gives a negative period value, so a retuned ROM's lowest octave and a half is unplayable. The same ROM can be read as audio through Audacity's Raw Data import (signed 8-bit PCM, mono, 12000 Hz) — how the stock kits were extracted.
• A hardware cross-check that closes a loop: Pan Docs notes a Super Game Boy 1 plays sharp, which is exactly why the ROM carries a second 216-byte note table recomputed for its 2.4 % fast clock — two sources, one fact, now stated together. Also VIN, the cartridge's analog input that could have been a fifth channel: "no licensed games used this feature".
• History the page had none of: goto80 "made one of the first Gameboy live performances with LSDj together with Role Model" around 2001, and played a whole Gothenburg set on a Game Boy in 2007 after his laptop broke.

One marker added, deliberately#

153 → 154. The chain-loop passage the live-performance thread quotes comes from an older edition of the manual ("3.5.1 Chain Loops"); the 9.2.6 manual in raw/lsdj/ has not been checked for the current wording, and the page now says so rather than implying the citation is current.

State#

676 pages. tools/lint.py clean. Unlanded ripples 340 across 124 pages. tools/build-site.py re-run.

Next on the list: sid (22), then commodore-64 (10), sidid (9), shuffle-funktempo (8), multispeed (8), sid-format (8), instrument-tables (8), adsr-envelope (8).

---

2026-08-31 — Unlanded ripples, pass 3: sid#

Operation: consolidation. No new sources; nothing in raw/ touched.

The second of the two big hubs: 22 summaries named sid and the page cited none of them. All 22 landed; 43.0 → 66.0 KB. Total now 340 → 318 unlanded claims across 124 → 123 pages.

Why so many, and what was actually missing#

The page was built from the datasheet, the Programmer's Reference Guide, the Yannes interview and Alstrup's measurements — so it was strong on what the silicon does and almost silent on what forty years of players did to it. The 22 unlanded summaries are mostly editors (defMON's four wiki pages and its megathread, CheeseCutter, Soundmonitor) and reverse-engineered drivers (Galway's Arkanoid, Matt Gray's Serpent Demo, the Tiny Sid 256-byte entries), plus four interviews. Deciding what was chip and what was tool was most of the work; recipes went out as pointers, register-level behaviour stayed.

Five things that had no home anywhere in the wiki:

• A section for $D418. The register does two unrelated jobs — volume in the low nybble, filter mode in the high — so every program that touches one has to preserve the other, and three sources thirty years apart show the same defensive idiom: Soundmonitor's volume byte "almost identical to SID register 24, but without the filter nibble", Galway's lda #$0F / ora TEMP / sta $D418, defMON's two self-modified operands ($10B1, "Make sure that the upper nibble is always zero"). The volume-register fourth voice then gets its three generations in one place: 1986 Galway, who explicitly disclaims the invention ("Never would I claim to have invented that technique, I just got it published first") and whose Arkanoid percussion turns out to contain no PCM data at all — six routines adding a constant to a 4-bit accumulator through a halving delay table; 1987 Matt Gray's fourth pattern channel streaming nibbles from a CIA-2 NMI while the raster IRQ plays the SID voices; 1991 JCH hacking digis into his own editor and making exactly four tunes with them.
• Two SIDs, or three. The page treated the chip as singular. Now: the chip is mono, so stereo means a second chip; defMON's selectable bases $D4xx/$D5xx/$DExx/$DFxx (and the warning that a wrong one corrupts RAM); the cost in software — a second full copy of the player, no packing for 2SID tunes, and frantic's "about 95% of all bugs I have fixed in defMON over the years were related to that"; and hermit writing a 3SID tracker on a real C64 in four months of 2008, four years before sid-wizard, because "DMC 4.3 and Prophet64" were the only 2SID editors he could find.
• Software can ask which chip it is on. defMON's init detects 6581 vs 8580 (disabling interrupts while it does, which is an integration trap) and then acts on the answer, hard-capping the filter cutoff at $0200 on an 8580 and $0000 on a 6581 "to make the chips behave more alike". TLR's sid-detect2.prg finds a second chip at an unknown address.
• The filter clicks when you switch it — less on the 8580. soundmonitor 1986 gives a sound a FF "touch no filter register" value "which avoids the SID's click when filters switch"; freedom 2010 gives the positive form — on the 8580 the click is faint enough that "it is possible to change filter type on an instrument while a note is being played". Twenty-four years, two sources, one artefact, and a technique that only exists on the newer chip.
• The MIDIbox SID. The page had the chip driven from a PC (Catweasel) but not the chip in a MIDI synthesizer. It also supplies the wiki's definition of "wavetable" in the C64 sense, and the envelope bug with a number on it: "The gate is delayed by up to ca. 30mS!"

Findings worth keeping#

• The noise-reset trick has a discoverer and a date. The page already documented that the test bit restarts the noise shift register (Alstrup's measurements, 4mat's $f9 then $81 recipe). JCH's memoir attributes it: johannes-bjerregaard, visiting him around 1988, "discovered how to use a test bit in the SID chip to reset the random numbers used internally by the noise waveform", making noise "more predictable … slightly more metallic" — and JCH's scanned papers include a draft magazine article about using the test bit when composing. frantic states the mechanism cleanly for tracker users: it "resets the oscillator that produces the waveform, and that is completely independent of the volume envelope generation".
• A contradiction narrowed by released music. The page carried "the OSC3/ENV3 readback was never used" (Furnace; Yannes's own "probably never used") against the PRG's examples and Alstrup's measurements. Three 2005 tiny-sid-compo entries settle it as released music: $D41B picks note positions (lda $d41b / and #7 / sta pos) and self-modifies waveforms, and aleksi-eeben derives the filter cutoff from $D41C (lda $D41C / lsr / adc #$20) — the datasheet's own "WAH WAH" suggestion, in a 256-byte tune twenty-three years later. Read "never used" as "rarely, and then mostly where code size mattered more than a voice".
• Two tools agree on the silent combined waveforms. CheeseCutter's cheat sheet — "Waveforms $3, $6 and $7 are audible only on new SID (8580)" — is exactly Chordian's 31/61/71 audibility table, reached from a different editor in a different decade.
• Microtuning is free on the SID and not on the Game Boy, and the reason is one line of the datasheet: 16 bits of frequency divider. little-scale's 2009 compilation call names the SID among chips with "a frequency resolution that is fine enough for the composition of microtonal music"; lsdj needs its ROM note table patched, a SID editor needs nothing.
• The cutoff's low three bits can simply go unwritten — anarkiwi's register dump shows defMON never touches $D415, "so the rightmost two ACID digits of a set value do nothing".
• A 1983 game that pokes the SID from BASIC. Telengard zeroes all 25 registers, computes a 60-entry note table by one multiplication per semitone (r=61176/64814), and times notes by how long the interpreter takes — the Programmer's Reference Guide style as shipped software, and the reason its rip had to become an RSID.
• Emulation, four verdicts, four different missing pieces: hermit 2008 (the analogue filter, and "the note-frequency dependent samplerate of the SID"), goto80 2002 ("Don't imitate, innovate — fuck emulators"), frantic 2020 ("It's like a fractal maze that drives you more mad the closer you try to get to it", and defMON is native so you work on the real chip), and electro·pizza 2019 arriving from the other side — VICE on a Pocket CHIP is for learning defMON, "especially useful in experimenting with and constructing instruments and effects in the SID tab", not for performing.

State#

676 pages. tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete. Unverified markers 154 across 101 pages (unchanged). Unlanded ripples 318 across 123 pages. tools/build-site.py re-run.

Next on the list: commodore-64 (10), sidid (9), shuffle-funktempo (8), multispeed (8), sid-format (8), instrument-tables (8), adsr-envelope (8), pitch-slide-and-portamento (7).

---

2026-08-31 — Unlanded ripples, pass 4: commodore-64#

Operation: consolidation. No new sources; nothing in raw/ touched.

10 summaries named commodore-64 and the page cited none of them. All 10 landed; 11.0 → 22.5 KB. Total now 318 → 308 unlanded claims across 123 → 122 pages.

What the ten were, and what they exposed#

Nine of the ten are the defmon/goto80 seam (the megathread, three defMON wiki pages, the electro·pizza VICE guide, goto80's own site, the 2002 Remix64 interview, Wikipedia) plus JCH's memoir, the Telengard RSID rip and the furnace C64 docs. On a platform page they turned out to be about one thing the page had never said out loud: the machine outside the chip. The page went straight from the SID's registers to composer anecdotes, with nothing in between about the clock, the memory map, the ports, the keyboard, or what it takes to record or emulate the thing. Sorting machine from chip was again the whole job — the 2SID address map, the register table and goto80's list of loves and hates all belong on sid and stayed there, referenced rather than repeated.

Five sections that had no home anywhere in the wiki:

• Clock, frame and memory map. PAL Ø2 = 985248 Hz, NTSC = 1022727 Hz (defBPM's constants, and they agree with the book's 17734475 / 18), one PAL frame = 19656 cycles — which is exactly why defMON's 1× timer preset is $4CC7, and why it can sit on an exact BPM where frame-locked editors cannot. Then the eight addresses that recur across every C64 music source in one table: $0801, $1000–$1FFF, $D011, $D012, $D400–$D41C, $D418, $DD00/$DD01, $DE00–$DFFF.
• The crowded neighbourhood. $DE00–$DFFF is 512 bytes wanted by a MIDI cartridge, a fastload cartridge and the second SID at once. The page already carried the SID-Wizard book's warning that cartridges clash with MIDI interfaces there; frantic's reason for putting defMON's sync on the user port instead — MIDI carts collide with the 1541 Ultimate and Action Replay whose fastload routines defMON uses for disk I/O — is the same collision from the other side, and the two now sit together.
• Music from BASIC, and what an interpreted note costs. The sid page already had Telengard as a shipped game that pokes registers from BASIC; the number that belongs to the machine is what it costs — one POKE ≈ 5 200 cycles ≈ a quarter of a PAL frame, for qq=1 to 600: next ≈ 672 000. That is the argument for raster-interrupt players in one figure. Also its four-byte sequencer rows, where 0 = hold and 255 = gate off, i.e. a tracker's empty row and note-off in 1983.
• The keyboard. The C64 has keys a PC does not (RUN/STOP, C=, CONTROL, the back-arrow ←, £, ↑, RESTORE) and C64 editors lean on all of them — defMON is driven almost entirely from them. That is why every emulator guide in the wiki opens with a key-mapping table, and why FlexiArcade sells a 3D-printed defMON keyboard overlay. The mapping tables themselves stay on vice.
• Noise and clean recording / running it in emulation / two machines at once. The book's video-port pin-5 ground fix now sits beside frantic's video-port recording and 4mat's POKE 53265,65 screen blanking. Emulation gets the fact that Debian's vice package ships no C64 ROMs (copyright), the three practitioners' verdicts on what emulation is good for, and furnace as the machine modelled rather than emulated. And two-machine work — goto80 and frantic's weekly two-C64 defMON jams from 2008, defSYNC's $DD00 CLK-bit poll, and JCH carrying tunes from an Amiga 500 to the C64 on paper before cross-assembling with EASS from December 1990.

One To-verify item narrowed#

The page's single marker asserted two things at once. The PAL/NTSC half is now resolved: a tune does play at a different pitch and tempo on the other standard unless the player compensates — defMON is PAL-tuned and needed a rebuilt frequency table for its 2018-02-11 NTSC build, its tempo math takes the clock as a parameter, and the .sid header carries CIA timer values for cross-standard playback. What remains unverified is only whether Drean is the Argentine PAL-N variant, and the marker now names the kind of source that would settle it.

State#

676 pages. tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete. Unverified markers 154 across 101 pages (unchanged; one narrowed, none removed). Unlanded ripples 308 across 122 pages. tools/build-site.py re-run.

Next on the list: sidid (9), shuffle-funktempo (8), multispeed (8), sid-format (8), instrument-tables (8), adsr-envelope (8), pitch-slide-and-portamento (7).

---

2026-08-31 — Unlanded ripples, pass 5: the whole "next on the list"#

Operation: consolidation. No new sources; nothing in raw/ touched.

Six pages in one sitting, clearing every page the previous plan named: sidid 9→0, shuffle-funktempo 8→0, multispeed 8→0, sid-format 8→0, instrument-tables 8→0, adsr-envelope 8→0, pitch-slide-and-portamento 7→0 (with commodore-64 10→0 earlier the same day). Total 318 → 252 unlanded claims across 123 → 115 pages. Sizes: sidid 4.9→13.6 KB, shuffle-funktempo 12.1→17.7, multispeed 11.9→16.5, sid-format 14.1→18.4, instrument-tables 28.1→35.4, adsr-envelope 25.5→30.2, pitch-slide-and-portamento 21.6→25.9.

What each page was actually missing#

• sidid had a cheat sheet of names and nothing about what a name means. The nine summaries were the realdmx driver sources, so the page gained what a signature keys on — Fred_Gray is the note-trigger idiom sta V1CTRL,x / inc V1CTRL,x, David_Whittaker the gate write stx SIDV1CTRL / inx / stx SIDV1CTRL, Jeroen_Kimmel its instrument-load code, and Martin_Galway's published signature keys on the CMP #$C0 in an interpreter whose music can re-point the play vector — and one name, a family of players: Hubbard's driver "with small modifications" across ~30 tunes and two disassemblies that already differ, Ocean's driver compiled two ways per game, Bjerregaard's two files with different data encodings, David_Whittaker covering both the trivial Lazy Jones routine and the full Panther command language. The find of the pass is one line in johannes-bjerregaard's own 1989 ;IMPROVEMENTS: list: ;RECOGNITION BYTES CHANGED! — a working composer tracking which bytes of his player identified it, in the same file as a Maniacs of Noise job sheet. Also: the name credits the driver's author, which is why Paul_Hughes exists and Jonathan_Dunn does not.
• shuffle-funktempo had no defMON section on a page about swing. defMON has no groove list at all — per-step, per-channel speed columns (ticks − 1), so drums can run 5 3 5 3 against a straight melody — and under hardware sync the adapter's six pulses per two steps leave exactly one swing available, 4+2. Beside it now sits the smallest possible swing: a tiny-sid-compo entry XORs the duration byte with $08 after every note. From LSDj: the groove reaches into the instruments (wave SPEED should equal the groove's tick count), it is a resolution knob as much as a feel (groove 3 = 32nds for sidechain recovery), two grooves can run at once if the tick totals match, and a tempo curve can be written per phrase. New tip: a swung song is a swung clock — chip sync counts steps, not milliseconds.
• multispeed was missing the file layer: a .sid header cannot say "four calls a frame", which is what RSID is for, and the CIA constants a multispeed tune ends up depending on. The consequence, dated: HVSC update 40 shipped "about 500 multi speed fixes" for one wrong constant — "the C64 has $4CC7 and not $4CF9 cycles per frame, due to running at 50.125Hz instead of 50Hz". Plus a CheeseCutter section ("doesn't work well past 4x") and what the tools supported, 2002–2015 — GoatTracker to 16×, TFX 2.97 claiming "up to 10x on all channels, and 72x on a single channel", ACID 64's "exact timing for … multi speed songs".
• sid-format covered SID-Wizard and GoatTracker exports and no others: CheeseCutter writes .sid only through ct2util (whose import and init commands move a tune between player versions or wrap a custom player), and defMON exports no .sid at all — packs to PRG, strip the two load-address bytes, and 2SID tunes cannot be packed. General warning now stated: the export is not the editor's playback. Two smaller finds: the header's SID-model bits are what a two-socket catweasel uses to pick its chip, and Hunter's Moon uses the songs field as a sound-effect bank — 16 "songs" that are 16 effect routines.
• instrument-tables gained the size-limit extreme (Tiny SID 2005: Laxity's two parallel wave tables, 4mat's five bytes with "only one counter for both data & waveform", and a 3-byte instrument that carries the note duration), Galway's Arkanoid instrument as memory the music pokes while it plays (29-byte definition, 34-byte running copy, SETNI/INSTR/SETCI), the LSDj table economy (32 tables against far more phrases, S being relative, the instrument number as the table's trigger), and a date for when a private driver acquires a wave table at all: JCH's NewPlayer, July 1988.
• adsr-envelope had no defMON or CheeseCutter section. defMON's field guide states the trap plainly — sustain "is the loudest = the VOLUME (not the time)" — and that the envelope runs on regardless of the gate column; CheeseCutter is unusual in setting one envelope parameter per pattern command (A0–AF…D0–DF) and in naming four restart types including the "Laxity restart" that leaves AD alone. Two more: JITT64 makes AD and SR tables of their own, and Lazy Jones writes AD $66 once for the whole tune and never writes SR at all, so the gate retrigger is the entire articulation.
• pitch-slide-and-portamento gained CheeseCutter (portamento works only on tied notes and must be activated before them) and defMON (no slide command by design; the AF column packs finetune, bend-up speed and bend-down speed into one byte), the 2024–25 LSDj kick material (L80 always lands on C1, so the phrase note tunes the transient; 9.4.0's R finally retriggers pitch), and why PITCH SAFE exists — the noise LFSR can produce zeros and go silent mid-slide on a DMG, so SAFE re-inits the channel at the cost of texture.

Housekeeping#

Six pages had summaries in sources: that were missing from their ## Sources line (s-lsdj-manual three times, s-painperdu-06-grooves, s-it-manual, others); all reconciled — the two lists now match on every page touched today.

State#

676 pages. tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete. Unverified markers 158 across 104 pages (four added deliberately: two auto-transcript values on LSDj pages, two SIDId questions that name what would settle them). Unlanded ripples 252 across 115 pages. tools/build-site.py re-run.

Next on the list (the long tail starts here — no page above 7): filter-programming (7), hvsc (7), hard-restart (6), detune (6), rob-hubbard (6), game-boy-apu (6), fasttracker-2 (6), melody-writing (6), arrangement (6), soundmonitor (5), martin-galway (5).

---

2026-08-31 — Unlanded ripples, pass 6: the whole long tail down to 4, and a new lint check#

Operation: consolidation. No new sources; nothing in raw/ touched.

Fifteen pages, clearing every page the previous plan named plus the first four of the next tier: filter-programming 7→0, hvsc 7→0, hard-restart 6→0, detune 6→0, rob-hubbard 6→0, game-boy-apu 6→0, fasttracker-2 6→0, melody-writing 6→0, arrangement 6→0, soundmonitor 5→0, martin-galway 5→0, volume-slides 4→0, ring-modulation-and-sync 4→0, recording-samples 4→0, chip-samples 4→0. Total 252 → 170 unlanded claims across 115 → 100 pages.

Sizes: filter-programming 34.0→39.9 KB · hvsc 12.0→15.9 · hard-restart 29.3→34.9 · detune 8.8→12.8 · rob-hubbard 16.9→20.3 · game-boy-apu 16.9→21.9 · fasttracker-2 15.5→18.2 · melody-writing 7.6→12.1 · arrangement 13.4→19.0 · soundmonitor 12.8→15.7 · martin-galway 17.1→19.4 · volume-slides 9.1→11.5 · ring-modulation-and-sync 14.6→16.6 · recording-samples 9.1→12.3 · chip-samples 31.1→34.2.

What each page was actually missing#

• filter-programming was strong on registers and tools and had no entry earlier than 1986. It now opens the driver history with Dig Dug (1984), where the filter is not a table but a set of step commands in the note stream ($28 cutoff, $30 resonance, $50/$60 route in and out, $58 mode) with four zero-page shadows written as $D417 = $FC + $FA and $D418 = $FB + $FD — the shadow-register pattern every later driver repeats. Then the size-coded end: a whole filter as a 17-byte cutoff table (Tognon's 15BB), ENV3 into the cutoff a year before Resolution (aleksi-eeben, lda $D41C / lsr / adc #$20), and two sweeps for free — Agemixer's song counter and Frantic's jsr $b5ff into BASIC ROM, which also fixes his tune's 399-second length. The find worth repeating is in Tips: the filter is not only the chip, it is the board — Tognon's checklist for hosting a SID outside a C64 requires filter capacitors "like in C64, otherwise the filter (of the same chip) will have different behavior", and on his catweasel a low cutoff did not colour the sound, it swallowed whole voices. With it, his diagnosis method: dump, read the routing, patch every $D417 write, play again.
• hvsc had the collection's history and none of its machinery. Added: sidid as the tool HVSC and DeepSID use, with the observation that its .nfo cites players by collection path (/MUSICIANS/H/Huelsbeck_Chris/Starball.sid) — the archive is stable enough to be a code reference; hvmec, Tognon's companion collection of the engines rather than the tunes, and the source of the wiki's editor-versus-tracker distinction; what a hard rip costs (The Trivia Arcade: three entry points, four raster calls a frame, hence RSID); the prehistory from a composer's side (JCH's floppy-swapped 1990 collection, "later dwarfed by HVSC"); and the archive in the 2020s, where packed defmon tunes arrive through DeepSID and martin_demsky's were accepted in the 2020-12 update.
• hard-restart gained a section it had been missing entirely: what a decade of release notes settled. GoatTracker 1.4b makes the restart programmable (1 or 2 frames), 1.4c/d run the wavetable through the restart frame, 1.5 switches to the test bit "for much sharper sounds" and makes the HR ADSR configurable ("previously only AD"), 2.05 adds legato, 2.59 gives the gate-off timer $80/$40 disable bits, and 2.47 adds /Z — a deliberate random delay on reSID writes "in order to make potential ADSR-bugging on C64 audible", the closest thing in the wiki to a test for whether a restart will hold up on hardware. NinjaTracker converges on the same points from 1.1 to 2.03. Also: Galway's Arkanoid instrument ends in two timing bytes ($1B release delay, $1C restart delay, $FF = none), Matt Gray's V4.2 byte 6 as "a 'basic' restart of note", and MoN's gate-off point as a per-instrument nibble with $fb "connect" as its legato opposite.
• detune had no primary-source authority and no C64 tracker sections. The sid datasheet already recommends it under Unison — "a slight detuning of each oscillator or tuning to musical intervals" for "a rich, animated sound". Added CheeseCutter (command 3, 16-bit) and defMON (the AF byte: $00–$19 finetune where $19 = a semitone, so one step ≈ 4 cents, and bend speeds above that — detune and portamento share a column). Then two ends nobody had written down: detune as a size-coding device (a fixed $1E added to voice 3's frequency low byte so two voices can share one track) and when the whole tune is detuned — Matt Gray's player runs at A4 ≈ 424 Hz on PAL, "Some of Galway's tracks were the same, Green Beret for instance was at 424hz", with his advice to retune such material as samples with fine-tune rather than a pitch-shifter.
• rob-hubbard: what Matt Gray actually borrowed — "I did borrow his idea of drum tables but it was always coded from scratch … My vibrato routine was very different to his and Martin Galway's". Plus the hard-restart provenance chain from JCH's memoir (Hubbard used it, Laxity figured it out, JO of Amok taught JCH), his own summary of the job as Tel quotes it — "It was like making music with boxing gloves on" — and Commando as the scene's benchmark cover tune in the 2002 iseq argument.
• game-boy-apu: the crystal as the tuning reference ("the crystal oscillator is responsible for controlling all timing aspects … the underclocked Game Boy plays back just under an octave lower"), and why the fix is to retune the ROM table for a round 4.000000 MHz so the machine lands an exact octave down. The wave channel's budget from both sides: 16 levels against 65,536 is the crunch, and 16^32 ≈ 3.4 × 10^38 waveforms means searching the space is not a method (little-scale's patch would finish in "1,079,028,307,100,000,000,000,000,000 centuries"). Plus why ghost-channel intervals are harmonic — a rider must complete a whole number of cycles in 32 samples, and 16/3 = 5⅓ is why even the useful one aliases — and three new rows in the quirks table (hardware sweep and envelope "take like no CPU whatsoever"; a stutter kick cannot outlive its envelope or LENGTH; the kit rate as a property of LSDj, not of the chip).
• fasttracker-2: the FT2-vs-ST3 argument as the participants themselves ended it — vadimvs's correct prediction that the feature set, not the interface, would decide it, and daedalus's "let that stupid old argument DIE!" — and a C64 editor author's structural verdict: JCH found FT2 "much of the same principle as my own music editors of the past, except the channels did not use separate sequences", which is the orderlist difference in one clause.
• melody-writing gained the scene's own argument about whether any of it matters (TraxWeekly #39–40: basehead's actual claim, maelcum on chords as one tool, trixter refusing the framing) and a "what the tools let a melody do" section: IT's Alt-I templates and sustain loops, LSDj's shared phrases and per-chain transpose, GoatTracker's +++ that cannot re-open a gate the wavetable cleared, SID-Wizard's absolute pitches in the arp column, and which waveforms make a lead that survives both chip revisions.
• arrangement: the song screen as the arrangement — empty rows between sections (visible blocks, and in live mode islands that loop back on themselves), marks via three presses of B, per-channel queueing, and a clipboard that survives a song load if both songs share a template. Then "two intensities of the same part" (one verse kick, one chorus kick, no pattern change), why ducking channels the kick never touches keeps a breakdown's pulse, and echo as a final pass over free space.
• soundmonitor had its model and history but not why people left it: it was everybody's sound (JCH: the prestige was in your own player; its tunes went "trite"), its data was uncompressed and its player "way too slow … for rastertiming" (scarzix), and Matt Gray's summary — "to get your own sound you need to write your own routines I feel".
• martin-galway: Tel's characterisation as technique ("playing around with the 'octaves' all the time"), Gray's consistent second place and the 424 Hz tuning he shares with him, and what Ocean did with his idea after he left — jonathan-dunn's driver as the same macro-language with lead-sheet comments.
• volume-slides: LIMIT versus VOLUME stated as a rule (VOLUME scales before CUTOFF/Q/PHASE/DIST and destroys the shape you liked; LIMIT constrains after and is therefore the wave channel's fader), OpenMPT's Fast Volume Slides as a per-song flag, and E02→E03 as a 16th-note pump.
• ring-modulation-and-sync: a CheeseCutter section (the gate/sync/ring/test nybble, and "Waveforms $3, $6 and $7 are audible only on new SID"), the defMON manual's column split, and Galway's Rambo loader as heard-in-code — ring modulation as an instrument ($41 | RINGBIT) and as a mid-phrase poke into the playing instrument's control byte.
• recording-samples gained the method for recording a sound chip you are playing yourself: build the interface and take the signal before the machine's analogue stage ("you can always add a digital filter via software if you so desire"), generate the material yourself and say so, capture at 16-bit/44.1 and process at the point of use, state the coverage. Plus a section on recording toward a target smaller than you are.
• chip-samples: pulling samples straight out of a ROM with Audacity's Raw Data import (signed 8-bit PCM, mono, 12000 Hz as a starting point) and finding the audio by eye, because PCM looks like a waveform where code looks like a block; Schism's Alt-Y text-to-sample; "volume is height" for hand-drawn waves; and JCH's own 1991 digi voice, which stayed private because the editor was "too hairy to be used by anyone else but me".

New in tools/lint.py: citation drift#

The previous pass noticed pages whose sources: frontmatter listed summaries their ## Sources section did not. That check now runs in the linter (both directions), and it found 35 pages — the frontmatter and the section disagreeing by up to nine entries (instrument-design), with lsdj drifting both ways at once. All 35 reconciled; the check now reports 0.

State#

676 pages. tools/lint.py clean — no broken links, no orphans, no frontmatter issues, index complete, citation drift 0. Unverified markers 158 across 104 pages (unchanged; one HVSC marker narrowed to name the article that would settle it — SIDin #15's "JSIDPLAY2", in raw/ and not yet summarised). Unlanded ripples 170 across 100 pages. tools/build-site.py re-run.

Next on the list (nothing above 4 now): vice (4), shades (4), rainbow-arts (4), goattracker (4), frantic (4), chipmusic-org (4), charles-deenen (4), 64er (4), then a long tail of 3s and 2s.

2026-08-31 — Ripples summary-first, clusters A and B (inbox/plan-ripples-summary-first-2026-08-31.md)#

Operation: consolidation, with the list inverted — read one summary, land it on every page that names it. No new sources; nothing in raw/ touched. 170 → 134 unlanded claims across 100 → 83 pages.

Cluster A — Hülsbeck and Rainbow Arts (170 → 155, six pages)#

Six interviews and press visits (1987–2011) read once and spread over the pages each serves a different slice of: 64er 4→0, shades 4→0, rainbow-arts 4→0, arkanoid 1→0, plus two neighbours the plan pointed at, great-giana-sisters 1→0 and chris-huelsbeck 1→0.

• 64er had the listing and the prize but not the competition. It now has the contest from four sides — Hülsbeck's own "Das war im Februar 1986. Ich habe von dem Wettbewerb gelesen und dachte mir, ich mache da mal mit" (Happy Computer 3/87, which also notes 64'er as "eine Schwesterzeitschrift der Happy-Computer"), the 16-year-old's aim ("determined to get at least noticed … I never expected to win"), and, in 1996, "I had no contact with the 'scene'" beside "Winning the first music competition in a german computer mag" as the proudest moment of his career. A dating conflict is now stated rather than smoothed: Wikipedia cites the announcement to 64'er 6/1986 p. 173, Hülsbeck says February 1986; both stand, and shades points at the page where they stand. The Soundmonitor listing came out of the prize ("Danach habe ich mit der Redaktion vereinbart, daß ich noch ein Programm schreibe") and was reprinted in "64'er Extra Nr. 13".
• shades gained what made him enter (Synthsample "brachten mich auf die Idee") and the order of events that the wiki had backwards-shaped: the MusicMaster play routine existed before the tune, and only the win made him build an editor around it.
• rainbow-arts now carries two accounts of the hiring side by side (VGMPF/Remix64: the company phones him; his own March-1987 telling: "In der Zwischenzeit habe ich mich mit Rainbow Arts in Verbindung gesetzt"; 2011: "first contracted, then hired"), the Bad Cat delay that cost him the sample-priority credit, and its 1990s DOS side — Lollypop, commissioned from Brain Bug in Aarhus with Vibrants doing the audio (JCH on AdLib SFX, Drax and Metal on music), PC 1994 / Amiga 1995.
• arkanoid gets one paragraph it was missing: the game is routinely named as the first with sampled drums, and the claim is disputed — Hülsbeck's 1996 "So Martin Galway got the fame with 'Arkanoid'" against Galway's own refusal to claim the invention. Both sides already live on martin-galway; the game page now says so instead of implying a settled first.

Cluster B — JCH's memoir and the editor lineage (155 → 134, eleven pages)#

charles-deenen 4→0, maniacs-of-noise 3→0, tracker-scene-history 3→0, sid-factory-ii 2→0, hewson 2→0, tim-follin 1→0, turbo-assembler 1→0, cheesecutter 1→0, cybernoid-ii 1→0, musicfile 1→0, jeroen-tel 2→0.

• The memoir's densest unlanded item was the prestige economy, and it belonged on tracker-scene-history: in 1987 "almost everyone was using a public domain editor called SoundMonitor", but "there was also an enormous prestige in using one's own player … I was among just two or three people in Denmark that could do this"; shared-tool sounds went "trite". Status, not convenience, produced the 1987–89 wave of private drivers. With it: the group-hopping chronology (New Men → Galaxy → 2000 A.D. → INXS → Jewels → Wizax → Dominators → Channel 42, late 1987 to 1989) and where the C64 line parts from the Amiga line — JCH wanting voices side by side with independent flexible sequences, against SoundTracker's voices "fused together in the same pattern".
• charles-deenen is the page the memoir changes most: he was the gatekeeper of the English market. He turned JCH down — "the English market not wanting the 'hit, snare' kind of music and that I should listen to what tim-follin had done" — took him on after the Push-It intro tune, and the one job that followed was Chase HQ II, Link composing while JCH multiplexed two SFX voices against two music voices. That single sentence also gave tim-follin the one thing his page lacked: evidence of what UK publishers were buying, not just what composers admired. Tel's 1994 and 1996 statements were added to the driver-authorship resolution, where they line up with every other first-hand account.
• turbo-assembler finally says why a music page cares about an assembler: JCH wrote NewPlayer and his editor in it and typed his first tunes as hex into Laxity's player listing — and it ran out. By December 1990 its memory footprint was the bottleneck, so editor v3 moved to Henning Andersen's EASS cross-assembler, edited on an Amiga 500 and dumped into C64 RAM.
• cheesecutter gets the provenance of two things it displays: the stacked view (JCH's paper sketches predate his seeing SoundTracker — "the notes in horizontal bars" first, downward columns after) and the NP21.G4 in its own player credit (JCH's NN.GN scheme, dated on paper at 10.G0, 15.G7, 17.G1 = hard restart). sid-factory-ii gets the same chain from the other end — Laxity's new player version for the JCH editor, and SIDId's five separate signatures (JCH_NewPlayer V1–V20, Vibrants/Laxity, Laxity_NewPlayer_V21, SidFactory/Laxity 2006, SidFactory_II/Laxity 2020) that make the line legible inside hvsc.
• musicfile gained its design constraint in Deenen's own numbers — "six to eight kilobyte of memory for music and sound effects", $200–1000, two to six days — which is why its tables are terse. cybernoid-ii and jeroen-tel gained the credit split his 2010 list makes explicit: on the famous Tel titles Deenen puts himself down for sound effects, not music.
• Two "merely relevant" calls, written as pointers rather than paragraphs: chris-abbott's 2003 interview says nothing further about maniacs-of-noise (the page now says exactly that and points at the remix economy on his own page), and goto80's Wikipedia entry is the after-story of tracker-scene-history, not part of it — one sentence, at the end of the 1997–98 debates.

Cluster C — the defMON / CheeseCutter tool family (134 → 95, sixteen pages)#

The most mechanical of the four: tool and person pages that the tools' own documentation had never reached. frantic 4→0, chipmusic-org 4→0, goattracker 4→0, live-tracking 4→0, vice 4→0, goto80 2→0, siddump 2→0, ninjatracker 2→0, lasse-oorni 2→0, defmon 2→0, midi-in-trackers 2→0, game-boy 3→0, dag-lem 1→0, hardsid 1→0, resid 1→0, tiny-sid-compo 1→0.

• One README served four pages. CheeseCutter's credits — "reSID engine by Dag Lem & A. Lankila. Parts of reSID interface (sid.cpp) by Cadaver / CovertBitops" — landed on resid (which gained ccutter's flags: reSID-fp is the default, -nofp turns it off, -fpr x picks a filter preset "0..16 for 6581 and 0..1 for 8580"), on dag-lem and on lasse-oorni, whose code is inside an editor he did not write. hardsid got the 2026 answer about USBSID-Pico support — "not a quick afternoon hack" — which is the honest state of real-SID output in editors.
• vice was the page with four sources and no account of what people do with it. It now has the 2002 Dig Dug rip's monitor workflow (breakpoints to log executed addresses, d, or save a memory range for jc64dis), the assemble-load-listen loop of Gray's public player (64tass → a.out → x64), composing in emulators since 2013 (scarzix, with his caveat that 6581 revisions differ so much "everyone will say their own C64 had the right sound"), and the humblest use of all — a tool author's release note reading "untested beyond starting in VICE in NTSC mode".
• ninjatracker finally says why it exists, in Cadaver's own 2002 words: GoatTracker "was fast & easy to compose on, but finally I had to realize that music made with it takes too much memory (because of the patterns…). So I started the quest for the extremely optimized playroutine/musicdata combination". goattracker gets the same comparison from the other side — CheeseCutter's author choosing "features/comfort over speed", so GoatTracker's player is the lighter one on rastertime — plus what it does not have (the stacked track view Scarzix went looking for).
• frantic's design philosophy is now documented from his program as well as from the forum: "Power through simplicity and flexibility rather than a large number of 'special functions'", hard-restart and vibrato left to the user, the player's coder interface ($1000 / jsr $1003 once a frame / $1006 for the extra multispeed calls) and defBPM — a formula and a printed table instead of a sync feature.
• live-tracking gained both sides of the clock. On the C64: mute keys, follow-play, and the three levels of defMON sync (Scannerboy's user-port adapter, defSYNC over a serial cable, or manual BPM matching with ±1-tick nudges live). On the Game Boy: little-scale's operating procedure in order, including the trap — stopping the MIDI clock stops LSDj, and CONTINUE is not a resume; you must reselect the position by hand. game-boy took the wiring consequences (a 22k–100k pull-down on the LSDj clock line, six machines per program per box or 15 in LSDj-only mode, ~35–200 BPM on a DMG, and a $1 4017 that clocks Nanoloop but never LSDj).
• chipmusic-org was an index of threads and is now also an account of the forum as a distribution channel: the CheeseCutter Facebook-group detour for a Mac build, the defMON thread as "the de facto support channel" with named testers, Scannerboy's sync adapter with no shop but a thread, and little-scale's loud TR-808 rebuild answering a kit-volume wish from a year earlier. Two missing thread index lines were added while there.
• siddump gained a use it did not know it had: defMON's sidTAB is "quite similar to siddump output", and the community's hard-restart survey was compiled "from CSDb and siddump analysis" — the dump as a table format, not just a diagnostic.

Cluster D — Matt Gray, SIDin #14–15 (95 → 80, nine pages)#

matt-gray 3→0, system-3 3→0, driller 2→0, matt-gray-player 2→0, tracking-workflow 2→0, dominator 1→0, codemasters 1→0, last-ninja-2 1→0 (and vice, already cleared in C).

• The plan predicted this material was "already read", and it was — what was left was player analysis that belongs on the person and game pages. driller now measures itself against the later engine: one-byte durations instead of $FD kk, arpeggio moved from the instrument byte to a pattern command, per-instrument pulse limits ("in Driller it was fixed into the source"), the hat and the end-of-tune fade-out both absent. dominator gets Tognon's ordering of the three engines he has read — Driller → Serpent Demo → Dominator, "as features maturated from one to another".
• tracking-workflow gained the one genuinely transferable thing in SIDin #14: the six-table order of work for any assembler-era driver (speed → instruments → drum tables → arpeggio tables → patterns → tracks), and the habit that makes it bearable — naming the bytes (INSTR = $FA, ARPEG = $FD, DUR5 = $75) so a pattern reads as music. Beside it, from the LSDj side, Pain Perdu's "TEMPO is not BPM": grooves govern structure, sixteen steps are a bar only at groove 6, and numbering grooves by tick count makes the table number tell you the feel.
• system-3 got the Bangkok Knights audition told three times, and the three tellings do not overlap: the vacancy was rob-hubbard's (leaving for EA), the turnaround was "a few days", the offer came "within the hour", and the call came from Mark Cale.
• Two honest small calls: codemasters gets one pointer sentence for a 2015 interview that only lists it among remembered companies, and last-ninja-2 gets Gray's refusal of a formula — "If there was a formula I'd have bottled it by now."

State after all four clusters#

170 → 80 unlanded claims across 100 → 59 pages. tools/lint.py clean: no broken links, no orphans, no frontmatter issues, index complete, citation drift 0. Unverified markers unchanged at 158 across 104 pages (one wording fix on siddump, whose "none of the three sources" no longer matched a five-source page). tools/build-site.py re-run.

What is left is the genuine one-off tail the plan predicted: nothing above 3 claims, and the top of it is now channel-interleaving, ghost-channel, impulse-tracker and it-format (3 each), then a long run of 2s and 1s — the OpenMPT manual pages, the Pain Perdu LSDj series and the TraxWeekly interviews.