===== page 1 =====
“Nuvolau friend”
Vice snapshot with Vice palette
Made with the GIMP from a MB photo
and converted to C64 320x200
Hires Mode Bitmap
by Stefano Tognon
in 2006
“Water splash”
...
Free Software Group
1
===== page 2 =====
SIDin 11
version 1.00
14 February 2007
2
===== page 3 =====
SIDin Contents
General Index
Editorials...............................................................................................................................4
News.....................................................................................................................................5
SIDPlayerDS....................................................................................................................5
Goattracker V2.51-V2.59..................................................................................................5
Ninjatracker V2.0-V2.03...................................................................................................6
SID Factory 0.05 (alpha 1) ..............................................................................................7
SIDId V1.0x......................................................................................................................7
XMPlay SID plugin beta 24d.............................................................................................8
HARDshit 2.0....................................................................................................................8
Goattraker 1.53 stereo HARDSID.....................................................................................9
SOASC Collection..........................................................................................................10
SID Compo 6..................................................................................................................10
SidTool...........................................................................................................................12
HVSC update 46.............................................................................................................13
XMMS-SID v0.8.0beta17................................................................................................15
Laxity Interview!..................................................................................................................16
Tiny Sid 2 (part 2)...............................................................................................................19
BLOCK ACID DUB.........................................................................................................19
Resolution.......................................................................................................................26
Conclusion......................................................................................................................29
SID Factory vs Ninjatracker................................................................................................30
SID Factory.....................................................................................................................30
Ninjatracker....................................................................................................................35
Comparison....................................................................................................................37
Conclusion......................................................................................................................39
3
===== page 4 =====
Editorials
Stefano Tognon <ice00@libero.it>
Hi, again.
It's incredible how the free time is always so little and when a problem pop up, the solution can kill
it at a big rate. For me, an Internet connection problem in December due to a very low quality cable
signal in my country, cause 15 days of no connection, irrelevant telephones call to the provider that
does not resolve the problem, wire cable testing in the home, and finally only, only, by reprogram-
ming the modem to produce a non standard tone, I get it connect again but with a low, low quality
rate.
Actually it takes me 15 minutes of trying only for sending or receiving a mail, with lot of modem
hangup (even if I program it to terminate connection only if the provider is...exploded by a nuclear
war...), or having dialing plane signal for minutes. Sometimes the signal is stable for 30 minutes or
so, and it is for this that you are reading the magazine (so I'm able to upload it), but you have to
stress yourself for minutes by trying and trying before find this situation.
With this problem, the first article about the second part of Tiny Sid Compo 2 is not completed for
the 4 players I planned to discuss, so I split it and another part will be written in the next number.
The second article is about a comparison of two editors: SidFactory and Ninjatracker, that as you
will read in the news, were released in the same week.
Finally, this is the period I traditionally run the Tiny Sid Compo, but this time I don't start it as I
cannot grant to have all the time it could took me. Maybe I will revamp the Big Sid Compo instead
as having a more longer duration could be better managed.
However, don't blame for the compo not running, my time for sid activity will be shared with
HVMEC (I have lot of new material that I'm adding to it), some tunes ripping, new tunes, and next
number of SIDin, of course.
Bye
S.T.
4
===== page 5 =====
News
Some various news of players, programs, and competitions:
• SIDPlayerDS
• Goattracker V2.51-V2.59
• Ninjatracker V2.0-V2.03
• SID Factory 0.05 (alpha 1)
• SIDId v1.0x
• XMPlay SID plugin beta 24d
• HARDshit 0.2
• Goattracker 1,53 stereo HARDSID
• SOASC Collection
• SID Compo 6
• SitTool
• HVSC update 46
• MMS-SID v0.8.0beta17
SIDPlayerDS
On August 2006 a port for Dreamcast of Christian Bauer's SIDPlayer where done by Troy
Davis(GPF). Original SIDPlayer is made from Frodo emulator and use SDL library, so It runs in lot of
systems (Linux, Beos, Windows, ecc.).
Look at http://www.neoflash.com/forum/index.php/topic,3034.msg21494.html#msg21494 thread
or go to the homepage http://gpf.dcemu.co.uk
Goattracker V2.51-V2.59
New version of Cadever's PC tracker is available:
v2.51
• Fixed packing of empty patterns when not using any effects.
• Fixed differing gatetimer value in instruments causing playback going out of sync.
• /G command line option no longer has to keep gatetimer values the same.
• Added high bit of gatetimer value to control whether wavetable execution starts right on the
note init frame. If used, will cause significant rastertime increase.
v2.52
• SHIFT+ENTER in orderlist view takes the next available pattern if pressed on a repeat or
transpose command.
• Added SHIFT+SPACE in pattern editor to start playback on a specific pattern row.
v2.53
• Fixed F3 to always play the currently visible patterns.
• Added BACKSPACE in the orderlist editor to set playback end position.
• Only effective when starting playback from position, not beginning.
v2.54
• Fixed SHIFT+F3.
• Changed SHIFT+
, . to update the pattern view.
• Changed SHIFT+SPACE to remain in pattern playback mode, if it was active.
5
===== page 6 =====
v2.55
• Added BACKSPACE in the fileselector to go to the parent dir.
v2.56
• Song filename cleared on songdata erase.
• Added /F to set custom SID clock frequency.
• Graphics output routines no longer compare text screen buffer to previous to find out if it
should be updated.
v2.57
• Added alternative hardrestart method & playroutine that is used when HR attack parameter
is at maximum (FF00 or F800 for example) - this can in theory give better reliability.
v2.58
• Cleanup, removal of a few questionable features:
* Gatetimer high bit to start wavetable immediately
* Guard 1stwave parameter
v2.59
• Songformat version update (GTS5/GTI5)
• Gateoff timer parameter bits control now hardrestart disable ($80) & gateoff disable ($40).
• 1stFrame Wave can be any from $01-$8F, $00 to leave both waveform & gateflag un-
changed, or $FE / $FF to control gateflag but leave waveform unchanged.
Download at http://cadaver.homeftp.net/tools/goattrk2.zip
Ninjatracker V2.0-V2.03
Released on August and September
2006 the new improved version of Ca-
daver C64 tracker.
• V2.0 Changes to previous
versions include commands
(also used as instruments), 2-
column tables and a slide
function that stops at target
pitch. As before, allows to
save both normal executable
musicdata and gamemusic
data without the player.
• V2.01 has improvements in
the editor and in the sound effect playback of the gamemusic player.
• V2.02 has the same editor features as V2.01, but has 2-frame hardrestart and
reduces playroutine zeropage use to 2 bytes. However, the new playroutine is
slightly slower and bigger.
• V2.03 has "hifi" style hardrestart and more optimized playroutine. Note duration
range has also been changed to 3-65, but there are no limits on the duration of a
pattern's last note, like in previous versions.
http://covertbitops.c64.org/tools/ninjatr2.zip
http://covertbitops.c64.org/tools/ninjatr201.zip
http://covertbitops.c64.org/tools/ninjatr202.zip
http://covertbitops.c64.org/tools/ninjatr203.zip
6
===== page 7 =====
SID Factory 0.05 (alpha 1)
SID Factory is a new editor system,
developed by Laxity of Vibrants/Maniacs
of Noise. The editor has inherited a lot of
features and look from the JCH editor.
It's been programmed from scratch and
is very different from the JCH editor be-
hind the surface.
The tracker made use of loadable play-
ers:
● v5: multispeed
● v6: low rasterline
Download the staff from
http://www.m-studios.dk
SIDId V1.0x
In September Cadaver has released a small DOS utility letting you know which playroutine is
used in a SID file. You can add playroutine signatures manually to sidid.cfg.
V1.0
• Original
V1.01
• AND function added
V1.02
• Multiscan added
V1.03
• Listing of unidentified files added
• Scanning of all files added
V1.04
• Added searching only for specific player
V1.05
• Added option to not recurse subdirs
V1.06
• Directory to scan can be given as an argument
• Added option -c to specify the configfile
• Added option -? to show usage information
• Added environment variable SIDIDCFG to specify the configfile
Download at: http://cadaver.homeftp.net/tools/sidid.zip
7
===== page 8 =====
XMPlay SID plugin beta 24d
Released in October a new version of the sid plugin for the XMPlay player:
● fix for memory leak in STIL loading
● memory optimization for SLDB loading
● sids with all songs filtered out are now rejected
● removed double displaying of start song, song numbers, song lengths if weren't changed
(General Info window)
● added displaying amount of memory needed for STIL and SLDB in About window
Download at: http://www.pieknyman.tk/
HARDshit 2.0
This is a fake hardsid.dll allowing realtime monitoring SID chip(s) writes. It is created by Raf of
Vulture Design and was released in November.
version 0.2 changelog:
• output text have additional mark - with hardsid's deviceid.
• supports up to 4 sids (e.g. you can monitor stereo in VICE)
• realtime updated table with writeable SID registers
The main site is at http://www.rafalszyja.republika.pl and download the tool from http://non-
ame.c64.org/csdb/getinternalfile.php/34346/hardshit02.rar
8
===== page 9 =====
Goattraker 1.53 stereo HARDSID
Raf even release in November a modified version of GT 1.53 stereo for Harsid card.
Download from http://noname.c64.org/csdb/getinternalfile.php/34366/gt153sterohardsidwin32.rar
9
===== page 10 =====
SOASC Collection
The SOASC project is an automated recording technique invented by Stone Oakvalley in order to
mass record music from the Commodore 64 and its SID chips (6581 and 8580).
It records all Sid from HVSC (about 93336 tunes) in both 6581 and 8580 chips without looking at
what chip the tune is intended to be listen to. The
recording format is in MP3 with 224kbps, mono,
44100Hz.
The final size of collection will be something like
300GB.
Look at http://www.6581-8580.com/
SID Compo 6
The annual Sid Compo by http://www.c64.sk was done in ending of the year and result were avail-
able before Christmas.
Public voting results... (49 votesheets)
1. Arctic Circles by Dane 1269
2. I swore a vow on my dying breath by Linus 1227.5
3. Albino Human by Fanta 1205
4. Critical Grid Voltage by CreaMD 1167
5. Misunderstandings by Zabutom 1122
6. Teh Disco by Jammer 1119
7. everything seems 2 be alright by Randall 1074.5
8. Electric Jesus by Intensity 1062
9. quickndirty by Syndrom 1048.5
10. Laserflake by Monk 1043
11. Delight by Nata 1037
12. White Light by A-Man 970.5
13. Autumn Memoir by Psycho 968
14. a Tune by Hein 938
15. Kollision by Maktone 932
16. Triskelion by Asterion 926
17. Tuesday Weirdness. by NO-XS 924
18. Illumination by Surgeon 921
19. Shortcut by Dr.Voice 902.5
20.
_
hidden (XeO3 Main Theme) by Luca 868
21. Mountain Morning by Sidder 859.5
22. INCD418 by Richard Bayliss 852
23. Carrier Lost by RaveGuru 834.5
24. c606 by Alex B 826
25. Breath by Rusty46 790
26. Lazy day by Rayne 786
27. alt.nerd.obsessive by Lordnikon 759.5
28. Summer SID by Uneksija 732.5
29. Paisley by Dalezy 693
30. Ex Inferis by Cadaver 685.5
31. Zuo3 by Raf 650.5
10
===== page 11 =====
32. City Waltz by Peter Bergstrand 33. Random Acts of Cruelty by Mermaid 34. Break beep by Vitnaque 457.5
434
414.5
35. @yourface by Josstintimberlake 370.5
Jury Rank: PTS
1 Arctic
Circles 157
_
2 Critical
Grid
_
_
Voltage 128
3 I
swore
a
vow
on
_
_
_
_
_
my_
dying_
breath 127
4 Albino
Human 111
_
5 Electric
Jesus 95
_
6 Teh
Disco 88
_
7 everything_
seems
2
be
_
_
_
alright 65
8 quickndirty 58
9 a
Tune 55
_
10 Triskelion 49
11 shortcut 48
12 Carrier
Lost 44
_
13 Delight 43
Autumn
Memoir 43
_
15 kollision 33
16 Mountain
_
Morning 31
17 Laserflake 29
18
hidden 23
_
19 lazy_
day 22
20 paisley 21
21 INCD418 20
22 alt.nerd.obsessive 19
23 c606 17
24 Summer 15
25 exinferis 9
zuo3 9
27 @yourface 7
28 Break-beep 2
Those authors didn not vote as jury
misunderstandings 85
White-Light 51
Tuesday_
Weirdness 44
breath 34
Illumination 31
Random
Acts
of
_
_
_
Cruelty 6
City_
Waltz 5
more at: http://www.c64.sk/index.php?content=article.php&articleid=115&id=4075
11
===== page 12 =====
SidTool
SidTool is a SID frontend developed by Madman that has:
● Support Sidplay2/w for playback
● Support TinySID for playback
● Easy access to the entire High Voltage Sid Collection
● Easy single-click interface
● Songlength database support
● Continuous play to discover SIDs unknown to you
● Playlist support with ability to add certain subsongs
● Experimental sid2midi support (sid2midi.exe not included)
● Built in search function
● Build on .NET framework 2.0 (required)
look at http://sidtool.drool.de/
12
===== page 13 =====
HVSC update 46
High Voltage SID Collection Update update #46 was released in January 21, 2007
After this update, the collection should contain 33,000 SID files!
This update features (all approximates):
1014 new SIDs
71 fixed/better rips
2 fixes of PlaySID/Sidplay1 specific SIDs
13 repeats/bad rips eliminated
664 SID credit fixes
160 tunes assigned a sidmodel flag
24 UNKNOWN demo tunes identified
As usual, the update and the all-in-one packages are available from http://www.hvsc.c64.org
This update features the tunes from the following parties:
X'06, c64.sk Sidcompo 6, Big Floppy People 2006, Breakpoint 2006, North Party 10, Primary Star
2006 and of course all the tunes from the 10 Years HVSC music collection.
Main Composers featured in this update: (Artists marked with NEW are either completely new to
the HVSC or they get their own directory in this update)
✗ Alien/WoW
✗ A-Man -> with previously unreleased stuff exclusively available in HVSC
✗ Art of Noise
✗ Richard Bayliss -> we've dug up 67 new tunes from him, with his kind help
✗ Arman Behdad -> the Trance master is active again, check out Electric
Jesus.sid
_
✗ Glenn Rune Gallefoss -> he kindly packed up some unreleased previews for us
✗ Chapelier -> one of the guys behind www.c64.com, he sent us his complete works
✗ CRD -> NEW! A promising new sid musician enters the scene!
✗ Dalezy -> still going strong with mellow minimalistic melodi
✗ DRAX / Maniacs of Noise -> that old melody master just won't stop composing candy for
our ears!
✗ Ed/Wrath Designs -> the Aphex Twin of the sid scene is back with an activity peak
✗ Fanta -> some fantastic goodies in there again by Onkel Fanta!
✗ Ghormak -> NEW!
✗ Hein Holt -> One of the multitalents of the 64 scene: He pixels, he codes and he composes
brilliant music
✗ Jac -> old Riffs hero from the 80s, good oldschool stuff
✗ Jammer -> always pushing the sid to new limits. This time with special focus on analog
speech synthesis (no sampled digis, everything made "by hand"!) His HVSC.sid is probably
the best tune
ever
_
_
in that department
✗ Josstintimberlake -> another avantgardistic sid musician brings us more data vibes
✗ Laxity -> yeah, you read correctly, the old famed composer is churning out blips and blops
again, as if it was 1989!
✗ Linus -> bringing us more follin'ish gangsta tunes! :-)
✗ Maktone -> with some good mood party tunes again, hooray!
✗ Adam Morton -> NEW!
✗ Markus Müller -> Superbrain sent us some old GCF tunes, cheers mate!
✗ Ne7 -> composer of the DTV Hummer game soundtrack, and generally a very sophisticated
musician (Electronat will blow you away!)
✗ Psycho-> ventures into the deep trancey department
✗ Radiantx -> A comparably new sid composer brings us some complex and dense oldschool-
ish tunes
✗ Randall -> As usual MC Randall can't miss to place at least one or two smash hits in every
13
===== page 14 =====
HVSC update. Very Polished stuff, in every meaning of the word.
✗ Rio -> NEW! Another youngster in the club of C64 composers.
✗ Rupert Dissident -> NEW! Another one of those home composers who just pop up now and
then and ask us if we have use for some twenty+ unreleased tunes. Pretty good stuff he did,
placative and complex.
✗ Rusty46 -> We bring you 52 tunes from this swedish guy, all done in 2005-06. Innovative el-
ements coupled with catchy hooklines!
✗ Sidder -> Smart arrangements, unusual instruments - Sidder is always exploring different
paths, from conventional to experimental.
✗ Stainless Steel -> Joe cooked up some brilliantly arranged remixes of recent pop hits for us.
✗ Jeroen Tel -> Jeroen once again opened his treasure chest of unreleased tunes and gave
the scene three "new" songs. But we also found a bit of old and forgotten stuff from him.
✗ Xiny6581 -> Another one of those rather new sid musicians, but a valuable enforcement.
Give these tunes some time, they don't release their message in the first 3 chords.
✗ Zabutom -> Ha, only about a year ago he entered the stage and now he's working the sid
like an orchestra that was made just for him! Wow!
Fundamental directory structure change
In this update the HVSC Crew decided to fundamentally overhaul the HVSC directory structure
and re-sort it.
There are a few disturbing things in the current HVSC directory structure:
1. Inconsistencies to our own rule as to what is root and what is VARIOUS. There are com-
posers in VARIOUS that, according to our rules should be in the root and vice versa.
2. With more and more composers surfacing and others getting their own directories by gather-
ing their 3 tunes, the directory trees grew to huge sizes (several screens high), which made
navigating and maintaining a bit cumbersome.
3. The term VARIOUS and the fact that the tunes inside that directory were 3 directory levels
further down than the glorious root composers was considered unfair by a lot of people. Dis-
cussions went in circles thinking about a better name (SCENE? POST
_
1993?) or to organize
the root dir similiarly (CLASSICS? PRE
_
1993?), so we finally decided to go for a completely
new sorting structure:
/DEMOS
/0-9
/A-F
/G-L
/M-R
/S-Z
/UNKNOWN
-> Sorting structure is now similar to the /GAMES directory. No changes to the rules were
made as to what goes into DEMOS. It's still the same, it contains tunes from demos made
by composers that are either unknown or that have composed less than 3 tunes overall.
/DOCUMENTS
-> No changes here
/GAMES
/0-9
/A-F
/G-L
/M-R
/S-Z
-> Again, nothing changed apart from the added /0-9 subdirectory. The GAMES directory still
contains tunes from games made by composers that are either unknown or that have com-
posed less than 3 tunes overall.
14
===== page 15 =====
/MUSICIANS
/0-9
/A
/B
/C
-> Basically we've merged the root directory and the old VARIOUS directory giving up the A-
F/G-L/M-R/S-Z sorting and going for a plain alphabetic flat sorting structure. A composer still
needs 3 or more tunes to get into the MUSICIANS directory, in order to avoid cluttering up
HVSC with hundreds of directories that only contain one or two tunes.
We think we found an unbiassed and futureproof solution with that new sorting structure: It's
easier to maintain. It's easier to navigate. It's more fair: No more discussions about "why is
composer x in the root, shouldn't he be in VARIOUS?" or "composer y made 3 gametunes
before '93, shouldn't he be in the root?"
. In addition, an almost flat structure like that hopeful-
ly encourages one or the other guy out there to program a proper frontend for HVSC one
day. Be it in form of a tool or a website interface, so people can conveniently switch between
several HVSC "views"
, assign "favourites" flags to tunes and composers etc. If you've got
the skills and the time/motivation to help us in making this vision reality, drop us a mail at
hvsc@c64.org. Especially a motivated webdesigner with PHP/MySQL knowledge would be a
welcome addition to the team.
HVSC has grown almost too big to jump around in Windows Explorer without getting lost, so
we hope you get used to the new structure and like it just as we do. There were concerns
that people new to HVSC would get lost in the new structure as they are usually searching
for their favourite game tunes from back then, which basically boils down to: Hubbard, Gal-
way, Follin and Tel. ;) We think that people will still find what they're looking for, only it takes
them a little longer than before. And who knows how many gems they may find on their hunt
for that old tune by randomly clicking into other composers directories, which were previous-
ly buried in VARIOUS.
XMMS-SID v0.8.0beta17
New version of XMMS-SID, a plugin for UNIX audio player XMMS, was released in end of Jan-
uary:
• Structural cleanup of the code was made; some previous artificial limitations were removed
in the process.
• Dynamic allocation is now used in all structures of SonglengthDB and STILDB, which should
improve memory usage efficiency.
• Documentation was updated and extended. Code was re-indented and commenting was im-
proved.
• Internationalization support via GNU gettext was added, with initial (but largerly unfinished)
Finnish translation.
• Some portability bugs were fixed, including a non-aligned memory access crash on certain
non-x86 platforms.
Download from: http://www.tnsp.org/xmms-sid.php
15
===== page 16 =====
Laxity Interview!
by Stefano Tognon
This time I go to interview Laxity, a old composer that I think you already know, as it is a member
of Vibrants and Maniacs of Noise groups. He recently started to compose again with sid. The inter-
view was done in January by mail.
Hello, could you introducing yourself and what you do in real life?
My name is Thomas Egeskov Petersen. I'm 33 years old, married and have two children, a boy and
a girl. I live near Aarhus in Denmark, where I work as a game programmer. I have worked as a pro-
grammer since 2001, where I was hired to program a game on the GameBoy Advance. Since then
I've worked with games on PC/X-Box, Nintendo GameCube and presently mobile phones (Java..
yak!).
You started composing in C64 where you were younger using Hubbard player. Wasn't this a
difficult task (and maybe a good way to learn coding)?
Yes, I started out making music that way. It was relatively difficult, but not an experience that taught
me coding but merely a basic understanding of how pointers work etc.. At the time I had pretty basic
coding skills, so I was able to figure out where pointers were read, parsed and so on and so forth.
The whole thing was mostly a deductive process - basic learning by doing in the context of trial and
error, I'd say. The most difficult was actually making the music, since I had no experience with that
and found it rather difficult to make it sound as I imagined it.
You code (and you are still coding) many music players (and editors), would you tell us
something about all your work in this field?
There's so much to tell. I think what drives me on the c64 is the fact that I'd like to not only do the
music, but the whole thing. It satisfies me to both program the driver AND compose the music, as I
somehow regard that as the full package of making music on the c64. Thereby not saying that those
who "only" use tools by other people aren't doing it right - I'm merely trying to say that I enjoy doing
the whole thing.
Looking at my source files I've identified 10 sound drivers for the c64 that I've written over the years,
where the first 4 were done in the period of 87-91 (3 publicly used). There rest of them we pro-
grammed after 2005. One is the JCH player (which took a long time to do), 3 for SID Factory and 2
for the Tiny Sid Competition Challenge last year (2005).
In the period of 1991 to present I've also programmed sound drivers for the Amiga, GameBoy, PC
and GameBoy Advanced, which all (except the PC one) have inherited a lot of structure and idea
from the basics done with the c64 sound drivers. The GameBoy Advanced driver was probably the
most complex of them all, and done in a hurry while I programmed my first game.
Most recently I've programmed my first ever music tracker. I was motivated to program it while I was
developing the music driver for the JCH editor. Therefore SID Factory, as it is called, takes off in the
JCH Editor platform - and the idea was to take it a bit further. I've basically tried to elaborate on the
driver/editor integration protocol and enhanced tracks to have an instrument column and a com-
mand column instead of just one unified one. Those were the basic idea for SID Factory, and has
proven to work pretty well as I can create table setups defined in the drivers, instead of having to
write a driver that uses tables as defined in the editor. (Makes sence?.. hmm)...
16
===== page 17 =====
What are your plans for the future of SidFactory? New features?
Well, I don't really have big plans for it, as it is mostly a spare time project (spare time, which there's
VERY little of actually). When I released the version 0.5 of SID Factory, there were some brief com-
ments on CSDb, where most of them were concerning missing features, such as STOP command in
the tracks, multiple tune support, etc.. Especially the latter is a bitch to do, but I will eventually have
to do it, otherwise I'll have another stranded project on my HD. Besides that, some editing features
were requested, which I ought to implement. Especially the nursing features, such as press-button-
while-on-pointer-move-cursor-to-position-in-table-pointed-to seem to be what some people want. A
bit annoying actually, and initially I though I wouldn't implement that. But I think I have an idea which
would work on commands too, so that if a command is of a type that point to a location in a wave ta-
ble, hitting the magic-button will transport the cursor to correct location in the wave table. etc. etc.,
all without structural changes to the driver/editor integration.. But it's hard to say when I'll have the
time to do those changes. At the present I have a version 0.6 which features some changes to the
driver/editor integration, and I'd rather like to have all eventual changes done before I release anoth-
er version, so it's not too much of a bother to users (if there are any at all. :)..)
What do you think about speech synthesis using sid voices?
I think it's interesting, but not a thing I have any experience at all.
Sometimes it sound really cool, but it's not always that impressive.
Now some quick final (standard) questions:
Real machine vs emulator: what do you think about?
Well, since the emulator is by definition trying to emulate the real thing and not vice versa, I'd have
to go for the real thing - since it's the real thing :)
6581 vs 8580 chip: any (musical) preference?
I like to use the 6581 myself, but a tune written for the 8580 I prefer to listen to on the 8580 and ditto
with tunes written for the 6581.
What is the worst and the better sid you composed?
I think I wrote a whole bunch of awful sid tunes, but I did some conversions like: James Bond, The
Pink Panther and I wanna dance with somebody, that suck big-time.
Who are your best sid authors?
I think there are a lot of good sid composers. I'm pretty fond of the stuff composed by Hein Holt -
he's really good and doesn't do that much mainstream stuff. I like that a lot. Goto80 sometimes
rocks my boat. He's really good with rhythm and sound design. Aleksi Eeben is a really good com-
poser, also alternative - a brilliant guy. I've bumped into him numerous times over the years, doing
GameBoy Color, GameBoy Advanced stuff and now c64.
From the "classic" c64 composers, my favourites are probably to be identified as the usual sus-
pects. Hubbard (all time favourite), Galway, Fred Gray - those sorts of composers.
17
===== page 18 =====
What are the best sids ever in your opinion?
I'd have to pick some and be concerned that some are missed out that deserve to be mentioned
too. Here are a few I especially like for one reason or another, presented in no particular order:
Master of Magic by Rob Hubbard
The Way of the Exploding Fist by Niel Brennan
Myth Demo by Johannes Bjerregaard
Wiz by Rob Hubbard
Daley Thompsons Decathlon By Martin Galway and David Dunn
Crazy Comets by Rob Hubbard
Body Slam (tune 4) by Tim Follin
Legend of Kage by Fred Grey
Ghouls'n Ghosts (tune 1) by Tim Follin
Implosion (tune 1) by Fred Grey
Finally, many thanks for the time you give for this interview, and now would you say some-
thing else to the our readers?
Any time and thanks for your interest, Stefano. Thanks for reading, still nice to have a c64 family out
there.
18
===== page 19 =====
Tiny Sid 2 (part 2)
by Stefano Tognon <ice00@libero.it>
This is another part about the Tiny Sid 2 compo. Here I go to show you two more entries, looking
at their engines. Next time we will see the last 2 entries.
BLOCK ACID DUB
This is the 256 bytes entries by Frantic/Hack'n'Trade and the player is well documented with
comments into the source code, so here I go to sum its features and working.
● It starts up at 326, by setting the IRQ address, and then disable (SEI) the interrupt.
● Notes frequencies are used only with the high value of frequencies byte (low value is never
set, and so it is 0). It was only used notes that give less error.
● Attack/Decay is never used for voice 1 and 2 (so it is 0)
● Use of undocumented instructions SBX, ANC & ASR for save space
● Use of hardrestart for drum
● Use random sequences for made the tune more longer and different
● Use only SID data values for both instruments and notes that are fitted together
The later point need some further comments.
The pattern of music data for each voices is done by a byte with this format:
Bits value Description
dddp.pppp
ddd=Duration
ppppp=Pointer to probability "nodes"
Probability node is a group of 4 bytes that contain a sequence (instrument/note) to play and it is
random selected as soon as a pattern is processed. Each of this value is a pointer to the instru-
ments/notes declaration. Look at this table:
Bits position Description
00000000 Delay of one frame
10000000 Set Sustain/Release
01000000 Set Attack/Decay
00100000 Set Control of voice
00010000 Set pulse high value
00001000 Set pulse low value
00000100 Set frequency high
00000010 Jump to another instrument
00000001 Stay:skip pointer update (=> end of instrument)
Those bits can be combined together and then a byte for each bit (but not for Stay) is follow in the
sequence.
19
===== page 20 =====
Here some examples, with a single note, an arpeggio (look at the Jump instruction) and a Bass
that have a noise in the first two ticks:
single note to play:
.byte SID
.byte $06
_
FRQHI | STAY
arpeggio:
.byte SID
FRQHI
_
.byte $0c
.byte SID
_
FRQHI | JUMP
.byte $09, arpeggio
Acid Bass 2:
.byte SID
_
SR | SID
CTRL
_
.byte $b8, $81
.byte TICK
.byte SID
_
CTRL | STAY
.byte $20
Now the source code. It is interesting that you look at the point where random sequence is taken
using $DC04 values. Here is when lot of undocument instructions are used to limit the choose for 4
patterns.
;*******************************************************************************
;
; Title: "BLOCK ACID DUB"
;
; Author: Frantic/Hack'n'Trade
; alias Glenn Again/Kommando Knorr
;
; Size: Code + data + load address = 254 bytes
; Two bytes reserved for the file system,
; so the tune will fit in a single disk
; block.
;
; Chip: 6581! (PAL)
;
; Format: DreamAss was used to assemble this source.
;
;*******************************************************************************
;
; Last Tiny SID compo I aimed to make a long composition with harmonies and a
; "real" melody, rather than creating advanced SID sounds. This time I aimed for
; fatter sounds instead. Having explicit or implicit functionality for hardrestart,
; arpeggio, wavetables and so on. The player is capable of quite flexible creation
; of sounds. In fact, any combination of SID register values each frame is
; possible (but only one value per frame, and excluding lobyte of frequency).
;
;*******************************************************************************
; Comments on some of the methods used to decrease size:
;
; - Using "random" recombinations of one single data sequence to make the tune
; somewhat varied and "longer" (filter type is also changed over time). A
; timer value is used as "random number" and this causes the tune to appear
; different each time. This player concept also allows one to use only one
; stream of data for all channels (only one "sequence") that just loops all
; the time to avoid seq-tables and code for handling such things. The
; possibilities of recombination are, however, configurable, so there is a
; certain degree of control of what might happen. (Jucke cited som dub guru
; the other day. This guru said that one part of the brain follows what is
; steady in the music, and another part follows that which is fleeting and
; passing, and therebetween a tension arises. According to the guru, that is
; what constitutes dub. Hehe.. Anyway, I hope that description fits this
; dubish likkle tune of mine. ;)
;
; - I am only using notes which can be approximated by using SID_FRQHI while
; letting Sid_FreqLo remain set to $00 all the time. That is, using base note
; of $0400, $0800, $1000, $2000 and so on. See end of source for some more
; info on that. There is a html/javascript attached in this package which
; was used for frequency calculation. Just enter a 16-bit base frequency
; in the source of the script, and view the results in a www browser.
;
; - AttackDecay is never changed in Voc 1 & 2 and remain set to $00 all the time.
;
20
===== page 21 =====
; - Using illegal ops whenever suitable (and not just for the sake of doing it
; of course). SBX, ANC & ASR are used here and they all do "two things in one"
; apart from SBX, which is useful for subtraction with the X register, rather
; than the A-register affected by standard SBC.
;
; - Not using different code structures for "notes" and "instruments". It's
; all just SID data, and freq-setting can be included in the "instruments"
; (in the case of a bass drum for example) or just be set on it's own, like
; a "note".
;
;*******************************************************************************
; GREETINGS TO NINJA/THE DREAMS, THE 252 BYTE PUNK!
;*******************************************************************************
; CONSTANTS
INSTPTR = $02 ;= $03 +7 +7 (X offs is 1...)
TICKCOUNT = $40 ;= $41, $48, $4f (X offs is 1...) ..is set to zero at startup.
SONGPOS = $99 ;= is set to zero at startup
CTRLBYTE = $60
NUMBEROFVOICES = 3 ;All three voices, of course!
;---------------------------------
; CODE
* = $326
@binbegin:
.word @start .word $f6ed
;Four byte init code to make it an executable.
@start: sei
;--
; Main player loop
@wrast: cpx $d012 ;X=?? after init and X=$fb after loop.. both values ok..
bne @wrast
;Loop once for each SID voice.
ldx #((NUMBEROFVOICES-1)*7)+1 ; +1 because we skip the Sid_FreqLO register altogether. Freeing
one control bit.
@vocloop:
lda TICKCOUNT+7+7+1;,x ;..but first, Filter Sweep - controlled by tickcounter from drum channel,
to achieve variable behavior.
asl
asl
asl
sta $d416
@parseseq:
@nonewinst:
;Handle tick counter
dec TICKCOUNT,x
bpl @parseinst
;--------------------------------------------
;FETCH NEXT INSTRUMENT (PARSE NEXT PROBNODE)
lda SONGPOS
inc SONGPOS
and #%00011111 ;There are 32 seqdata bytes so just throw away the three upper bits to take
care of wrapping.
tay ;A reg is set some lines up..
;Chose one of 4 possible instrument pointers
lda $dc04 .byte $2b, 3 adc @compositiongrid,y
pha
and #%00011111 ;Clear 3 msb
tay
pla
lsr
;Get "random" value from $dc04
;ANC instruction. Same result as "and #3 / clc" in this case.
.byte $4b, %01110000 ;Illegal ASR (AND first, to remove instrument ptr bits, and then LSR result,
in one instruction.)
ora #%00000111
sta TICKCOUNT,x ;Store the new tick counter value
lda @probnodes,y ;Get instrument number
beq @nonewinst ;Instrument table pos 0 has the special meaning NO INSTRUMENT CHANGE (and
contains no instrument data).
sta INSTPTR,x ;Store new instrument ptr..
;--------------------------------------------
;PARSE INSTRUMENT DATA AND WRITE TO SID HERE..
@parseinst:
;HARD RESTART - only for drum channel..
21
===== page 22 =====
ldy TICKCOUNT+1+7+7
cpy #2
bne @nohr sty INSTPTR+1+7+7
;Skip instrument read and use #2 as instptr instead.
@nohr:
;Fetch instrument data
ldy INSTPTR,x
lda @instdata,y ;Fetch ctrlbyte.
sta CTRLBYTE
@parslp:asl CTRLBYTE
stx @sidwlo txs ldx #5 bcc @nxtbyt
iny
;Save voice offset directly in SID-write code.
;Store X in stackptr temporarily, for later retrieval.
;6 registers used for each SID Voice.
lda @instdata,y
@sidwlo = *+1
sta $d4ff,x
@nxtbyt:dex
bpl @parslp
tsx ;Restore voice pointer in X..
;Check for STAY parameter. If it's set, then just skip pointer update.
asl CTRLBYTE
bmi @noptrupdate
;Save updated instptr (either just increased or from a JUMP argument).
;This one points either to first step in next instrow or to the jump argument.
iny tya
bcc @nojmpfetch ;Skip Jump-destination fetch if jumpbit was set (carry set by the ASL
instruction above).
lda @instdata,y ;Fetch instjmp argument
sta INSTPTR,x ;Now points to next instrument row to be parsed..
@nojmpfetch:
@noptrupdate:
@loopend:
;Loopcounting (X) and filter type changes.
lda SONGPOS
.byte $4b,%11100000 ;Use ASR to "calculate" filtertype. Don't change filtertype too often, as a
straight AND would give.
bne @dontforce
lda #$10 ;Force lopass filter if no other filter type is set.
@dontforce:
ora #$0f ;Set Global Volume (also used as AXS and-value) ;REMOVED.....and
always turn bandpass on
sta $d418 ;Filter type / Global volume
.byte $cb,7 ;AXS #7 or SBX #7 depending on who you ask.
bpl @vocloop
inx ;x becomes = $fb
stx $d417 ;Set filter resonance an voice input to #$fb (doesn't matter that
we also filter external input by this)
bmi @wrast
;=============================================================================
; Composition Data
;
; Contains a table of pointers to "probability nodes". One row for each voice.
;Duration constants.
D1 = 0 << 5
D2 = 1 << 5
D3 = 2 << 5
D4 = 3 << 5
D5 = 4 << 5
D6 = 5 << 5
D7 = 6 << 5
D8 = 7 << 5 ;Max duration..
@compositiongrid:
; |Voc1 |Voc2 |Voc3 |SeqStep|
;-------|---------------|---------------|-----------|-------|
.byte D2 | @pbd, D5 | @pabnt, D2 | @pnote2; 0 # |first row must set all three
channels.
; .byte ; 1
.byte D1 | @psd, D4 | @pabAAB ; 2
.byte D1 | @psd ; 3
.byte D1 | @pbd ; 4 #
.byte D1 | @pbdl, D4 | @pabAB ; 5
.byte D2 | @psd, D7 | @pabBl ; 6
; .byte ; 7
.byte D2 | @pbd ; 8 #
.byte D6 | @synx ; 9
.byte D2 | @psd ; a
; .byte ; b
.byte D3 | @pbd ; c #
22
===== page 23 =====
.byte D5 | @pnote ; d
; .byte ; e
.byte D1 | @psd, D7 | @pabAB ; f
.byte D1 | @pbd ;10 #
.byte D3 | @psd ;11
.byte D6 | @pabAl ;12
; .byte ;13
.byte D3 | @pbd ;14 #
; .byte ;15
.byte D1 | @pabBl ;16
.byte D1 | @psd, D6 | @synn ;17
.byte D2 | @pbd, D6 | @pabBl ;18 #
; .byte ;19
.byte D2 | @psd ;1a
; .byte ;1b
.byte D3 | @pbdll ;1c #
.byte D3 | @pabnt ;1d
.byte D2 | @pabAAB;1e
.byte D1 | @psd ;1f
;-------|---------------|---------------|-----------|-------|
;==========================================================================
@probnodes: ;Probabilitiy "nodes" (4 bytes each) for instrument trig.
;Max 32 bytes in this table!
@pbd = *-@probnodes
.byte @bd
@pbdl = *-@probnodes
.byte @bd
@pbdll = *-@probnodes
.byte @bd
@psb = *-@probnodes
.byte @bd
@psd = *-@probnodes
.byte 0; ;...means no instrument change.
.byte @sd
@pabAl = *-@probnodes
.byte @n_c2
.byte @sd
@pabAAB = *-@probnodes
.byte @syn
@pabAB = *-@probnodes
.byte @abA
@pabBl = *-@probnodes
.byte @n_e
.byte @cpb
@pabnt = *-@probnodes
.byte @abAx
.byte @abB
@synx = *-@probnodes
.byte @syn
@pnote = *-@probnodes
.byte @abA
.byte @n_dar
@pnote2 = *-@probnodes
.byte @n_g
.byte @n_e
@synn = *-@probnodes
.byte @n_c2
.byte @n_gd2
.byte @syn
.byte @syn2
;==========================================================================
; "INSTRUMENT EDITOR"
;
;
; Almost looks like real wavetables, although being a "text file instrument
; editor", doesn't it? ;)
; Some control contants.
TICK = %0 ;An empty control byte simply causes a delay of one frame.
SID_SR = %10000000
SID_AD = %01000000
SID_CTRL = %00100000
SID_PULHI = %00010000
SID_PULLO = %00001000
SID_FRQHI = %00000100 ;Note that FreqLO is not used at all
STAY = %00000001 ;STAY overrides JUMP, so only use one at a time. Else data will get out of
sync.
JUMP = %00000010
@instdata = *-2: ;Read comment for hardrestart below to get the idea with the "-2" here.
;Also, position 0 in the table is used as a "NO INSTRMENT CHANGE FLAG"
;and shall not contain data anyway.
;--- HARDRESTART
instrument trig.
@hardrestart: ;This one should be placed the same ammounts of bytes into the instdata table as ticks to preceed
.byte SID_SR | SID_CTRL |
23
===== page 24 =====
STAY
.byte $00, $10
@bd = *-@instdata
.byte SID_SR | SID_AD | SID_CTRL
.byte $fb, $04, $49
.byte SID_CTRL | SID_PULHI | SID_FRQHI
.byte $41, $08, $08
.byte SID_CTRL
.byte $11
.byte SID_CTRL | SID_FRQHI
.byte $10, $07
.byte SID_FRQHI | STAY
.byte $04
;----
@sd = *-@instdata
.byte SID_SR | SID_CTRL
.byte $33, $49
.byte SID_CTRL | SID_FRQHI
.byte $41, $0c
.byte SID_CTRL | SID_FRQHI
.byte $80, $a0
.byte SID_CTRL | SID_FRQHI |
STAY
.byte $40, $18
;---- Mongo Bass
@cpb = *-@instdata
.byte SID_SR | SID_FRQHI
.byte $54, $06
.byte SID_CTRL | SID_FRQHI | JUMP
.byte $11, $05, @cpb
;--- NOTE E (round error 10)
@n_e = *-@instdata
.byte SID_FRQHI | STAY
.byte $05
;--- NOTE G (round error -2)
@n_g = *-@instdata
.byte SID_FRQHI | STAY
.byte $06
;--- ARPEGGIO G - D2 (D2 round error -6, G error -4..)
@n_gd2 = *-@instdata
@n_dar = *-@instdata
.byte SID_FRQHI
.byte $0c
.byte SID_FRQHI | JUMP
.byte $09, @n_gd2
;--- NOTE C2 (round error 0)
@n_c2 = *-@instdata
.byte SID_FRQHI | STAY
.byte $08
;--- ACID BASS #1
@abA = *-@instdata
.byte SID_FRQHI
.byte $04
@abAx = *-@instdata
.byte SID_SR | SID_CTRL | JUMP
.byte $9a, $21,
@abB_entry2
;--- ACID BASS #2
@abB = *-@instdata
.byte SID_SR | SID_CTRL
.byte $b8, $81 ;Just to add some subtle filtered "percussion" here and
there.. ;)
@abB_entry2 = *-@instdata
.byte TICK
.byte SID_CTRL | STAY
.byte $20
;--- Tri Synth
@syn2 = *-@instdata
.byte SID_FRQHI
.byte $18
;High G note..
@syn = *-@instdata
.byte SID_SR | SID_CTRL
.byte $fc, $11
.byte SID_CTRL | STAY
.byte $10
;==========================================================================
;PRINT SIZE (Adding two for the obligatory load adress + two for the file system).
24
===== page 25 =====
;#print (*-@binbegin)+2+2
;==========================================================================
; Here is a list of calculated notefrequencies, assuming that base notes
; of the scale start at $0400 (C), $0800 (C2), $1000 (C3..), $2000, $4000...
; Some of them can be approximated by setting the hibyte of freq to
; the value, and just having lobyte freq set to $00 (as it is when the
; C64 is reset.) These values are calculated with the html/javascript
; file which is attached in the package.
;.byte 4 ; C error: 0 <------------
;.byte 4 ; C# error: 60
;.byte 4 ; D error: 125
;.byte 4 ; D# error: 193
;.byte 5 ; E error: 10 <------------
;.byte 5 ; F error: 86
;.byte 5 ; F# error: 168
;.byte 6 ; G error: -2 <--------------
;.byte 6 ; G# error: 89
;.byte 6 ; A error: 186
;.byte 7 ; A# error: 32
;.byte 7 ; H error: 141
;.byte 8 ; C error: 0 <------------
;.byte 8 ; C# error: 121
;.byte 9 ; D error: -6 <-------------
;.byte 9 ; D# error: 131
;.byte 10 ; E error: 20
;.byte 10 ; F error: 173
;.byte 11 ; F# error: 80
;.byte 12 ; G error: -4 <------------
;.byte 12 ; G# error: 178
;.byte 13 ; A error: 116
;.byte 14 ; A# error: 65
;.byte 15 ; H error: 26
;.byte 16 ; C error: 0 <------------
; byte 16 ; C# error: 243
;.byte 17 ; D error: 245
;.byte 19 ; D# error: 6 <-- (Useable, but not used here..)
;.byte 20 ; E error: 40
;.byte 21 ; F error: 91
;.byte 22 ; F# error: 160
;.byte 24 ; G error: -7 <------------
;.byte 25 ; G# error: 101
;.byte 26 ; A error: 232
;.byte 28 ; A# error: 130
;.byte 30 ; H error: 52
;.byte 32 ; C2 error: 0
25
===== page 26 =====
Resolution
This is a 512 bytes entries by Eric Odland. Here I present a reverse engineering source of the en-
gine, but before, we analyze some features of the player:
● It starts with basic sys call
● Irq is set by adding the pointer to $314
● It initializes all the sid registers by a loop using a table of values
● The init phase (irq, sid and track/pattern setting) is repeated when the track is over: a way to
save code by reuse all code even if not all is necessary
● Tune is composed by 13 patterns to play (patTable that use relative offset, so one byte for a
pointer)
● The notes used by the tune are 24 and are coded with two tables for low/high values of the
notes frequencies
● For each note the duration is $0A, unless the high bit is 1 ($15 is so used for duration)
● All the 3 voices play the same notes, but voice 3 is used a little different in release phase
from 1 and 2. Also, an effect of adding 4 notes over the current is also used. This is what
make the instrument timbre you listen.
Now the source:
.org $0801
.byte $0B, $08, $D6, $07, $9E, $32, $30, $36, $31
.byte $00, $00, $00 ; basic call: 2006 sys 2061
.org $080D
Init:
ldy #$18
loop: ; init SID registers
lda sidTable,y
sta $D400,y ; Voice 1: Frequency control (lo byte)
dey
bpl loop
jsr initPointer
lda #$15
sta $F7
sta $F8 ; store note duration
; set irq
sei
lda #<irq
sta $0314 lda #>irw
sta $0315 cli
rts
; Vector: Hardware Interrupt (IRQ)
; Vector: Hardware Interrupt (IRQ)
irq:
inc $0420
dec $F9 bne setFiltCutHi
; dec actual duration
lda #$15
ldx $FE
cpx #$02
beq skipReadDur
lda $F8 ; read note duration
skipReadDur:
sta $F9 jsr player
beq setFiltCutHi
; store actual duration
jsr Init
jmp $EA31 ; Default hardware interrupt (IRQ)
;=================================
; set filter cut off freq according
; to voice 3 envelope output
;=================================
setFiltCutHi:
lda $D41C ; Generator output
lsr
sta $D416 ; Filter cut frequency: hi byte
26
===== page 27 =====
jmp $EA31 ; Default hardware interrupt (IRQ)
player:
ldy $FE
bne testRelease
jmp readNext
testRelease:
dey
bne skipOnly3
lda #$40 sta $D412 jmp readNext
; release for voice 3
; Voice 3: Control registers
skipOnly3:
ldy $FB
lda $FA ror
bcc makeRelease
tya
adc #$04 tay
; note index for effect
; uses 4 notes added
makeRelease:
lda #$20
sta $D404 lda #$40
sta $D40B sta $D412 inc $FA ; Voice 1: Control registers
; Voice 2: Control registers
; Voice 3: Control registers
; inc note index for effect
jmp outFreq ; y=note index
readNext:
ldy $FC ; load pattern index
inc $FC ; increment pattern index
lda patt,y ; read value
bne processValue
jsr getNextPattern
cmp #$00
beq readNext
rts
processValue:
tay
asl ; test high bit
bcc no8Set
; carry was set
clc
lsr
tay ldx #$0A
stx $F8 inc $FE
; store without high bit set
; store note duration
no8Set:
asl
bcc no4Set
; carry was set
clc
lsr
lsr
tay ldx #$15
stx $F8 ; store without bit set
; store note duration
no4Set:
asl
bcc outFreq
clc
lsr
lsr
tay
lda #$15
sta $F7
outFreq:
lda freqLow-1,y
sta $D400 sta $D407 adc #$08
sta $D40E lda freqHigh-1,y
sta $D401 sta $D408 sta $D40F lda #$21
sta $D404 lda #$41
sta $D40B ; Voice 1: Frequency control (lo byte)
; Voice 2: Frequency control (lo byte)
; Voice 3: Frequency control (lo byte)
; Voice 1: Frequency control (hi byte)
; Voice 2: Frequency control (hi byte)
; Voice 3: Frequency control (hi byte)
; Voice 1: Control registers
; Voice 2: Control registers
27
===== page 28 =====
sta $D412 dec $FE
bpl notNeg
lda #$02
sta $FE
; Voice 3: Control registers
notNeg:
ldy $FC lda patt,y ; load pattern index
; load pattern value
bne isRead
jsr getNextPattern
cmp #$00 ; value read?
beq isRead
lda #$01
rts
isRead:
lda #$00
rts
;================================
; Get ext pattern and read the
; value of pattern
;================================
getNextPattern:
inc $FD ; pattern index
ldy $FD
lda patTable,y
bne readValue
lda #$01 ; end: value not read
rts
readValue:
sta $FC tay
lda patt,y
sta $FB
inc $FC rts
; store patten index
; increment pattern index
lda #$00 ; value read
;=================================
; Init pointers to pattern
;=================================
initPointer:
lda #$15
sta $F9 ldy #$FF
sty $FD iny
sty $FA ldy #$02
sty $FE
jsr getNextPattern
rts
; store actual duration
; reset pattern index
; clear note index for effect
sidTable:
.byte $30, $04 .byte $00, $08 .byte $20 .byte $05, $9F ; low/high freq voice 1
; low/high pulse voice 1
; control voice 1
; attack/decay sustain/release voice 1
.byte $8F, $0A .byte $00, $04 .byte $40 .byte $08, $A8 ; low/high freq voice 2
; low/high pulse voice 2
; control voice 2
; attack/decay sustain/release voice 2
.byte $FF, $0F .byte $00, $05 .byte $40 .byte $0A, $B9 ; low/high freq voice 3
; low/high pulse voice 3
; control voice 3
; attack/decay sustain/release voice 3
.byte $00, $50 .byte $67, $1F ; low/high of filter
; resonance/filter mode/volume
patt:
.byte $FF
pat01:
.byte $11, $07, $0B, $06, $0A, $04, $09, $05
.byte $0A, $07, $0B, $06, $0B, $06, $07, $09
.byte $0A, $00
pat13:
.byte $0F, $0B, $07, $07, $04, $04, $05, $07
.byte $09, $08, $06, $03, $04, $04, $05, $07
.byte $09, $00
pat25:
.byte $13, $05, $07, $09, $0A, $05, $07, $0A
28
===== page 29 =====
.byte $0B, $00
pat2F:
.byte $12, $06, $08, $0B, $0C, $08, $0D, $8C
.byte $0B, $4A, $00
pat3A:
.byte $11, $09, $0A, $0A, $0B, $0B, $0C, $0A
.byte $09, $00
pat44:
.byte $10, $09, $08, $08, $07, $05, $04, $02
.byte $01, $00, $13, $07, $0B, $06, $0A, $04
.byte $09, $05, $0A, $00
pat58:
.byte $13, $8E, $0D, $0B, $84, $07, $89, $0A
.byte $89, $08, $86, $04, $87, $0A, $88, $4B
.byte $00
pat69:
.byte $13, $8E, $0D, $4B, $8B, $09, $48, $06
.byte $05, $04, $04, $00
pat75:
.byte $13, $13, $13, $00
patTable:
.byte pat01-patt, pat01-patt
.byte pat13-patt, pat13-patt
.byte pat25-patt, pat2F-patt
.byte pat3A-patt, pat44-patt
.byte pat4E-patt, pat4E-patt
.byte pat58-patt, pat69-patt
.byte pat75-patt, $00
freqLow:
.byte $E9, $61, $68, $8F, $30, $8F, $18, $D2
.byte $C3, $D1, $1F, $60, $1E, $31, $5A, $A3
.byte $CC, $23, $86, $B4, $47, $98, $47, $0C
freqHigh:
.byte $07, $08, $09, $0A, $0B, $0C, $0E, $0F
.byte $10, $12, $15, $16, $19, $1C, $02, $02
.byte $02, $03, $03, $04, $05, $05, $06, $07
.byte $00
Conclusion
This time we had looked at one 256 bytes and one 512 bytes engines and they are very differ-
ents. In particular the 512 bytes take emphasis to the notes to play and use the 3 voices together for
creating a timbre of one instrument. The 256 bytes instead uses random sequences to make the
composition more longer, but all 3 voices play different notes/instruments.
29
===== page 30 =====
SID Factory vs Ninjatracker
by Stefano Tognon <ice00@libero.it>
It is strange to see that two new editors were released about the same day.
The first is the Laxity tracker: SID Factory
The second is Cadaver editor: Ninjatracker V2
In this article I go to show you the two new editors and soft compare the features they supply.
SID Factory
The first thing that pop up by looking at
the tracker is the very similarity with JCH
editor, that can be summed as:
• One editor that can use many differ-
ent player drivers
• Tracker style editor based onto ta-
bles of values
• Very similar looking (even if you will
note that they are different)
However the program starts with the
main frame you see here in the right side: it
has the main menu (enter editor, disk,
packer and options) in the upper part and in
the lower part it has information about the
current loaded driver.
The JCH editor has the similar main
frame, but in it there is even the list of the
disk contents. In this editor you will see the
list of files in the apposite sub-menu, but on
the other side you have more information
about the current used driver (in JCH there
is only the last screen line for that informa-
tion).
Maybe (but this is just my opinion) it
could be interesting to have one frame as
in JCH, but with all information that SID
Factory has, in the main menu. That could
be possible by restyling the interface a little
bit.
Entering the editor itself, you will see that it looks similar to JCH here too: in the upper part there
is the patterns editor for each tracks and in the below part there are the tables for instruments.
Color is green/white like in JCH but here it uses even light green and blue, giving to the editor his
different look style from JCH.
Take present that for each track, you will set the pattern to use by entering it as a number, and
below you have the pattern to enter. So, all the tunes is viewed by go down in the pattern editor.
30
===== page 31 =====
In one pattern row you can insert 3
values:
➢ Instrument number (32 max)
➢ Command number (64 max)
➢ Note values
Instrument and command number
refers to the fixed table of values that
are in the low part of the editor.
A dynamic table (in center of editor)
is activated on demand and contains
other values used by the other tables.
Actually SIDFactory comes with two
different driver:
➢ V5.0x: standard music driver with calculated vibrato and no aim at execution speed (multi-
speed allowed)
➢ V6.0x: "fast" music driver without calculated vibrato and aims at a maximum execution spike
at around $18 scanlines
Here I'll go to report the actual meanings of values for the two drivers and so you can figure how it
works:
Driver v5.0x Driver v6.0x
Instrument: aa bb cc dd ee ff gg hh Instrument: aa bb cc dd ee ff
aa Attack/Decay aa Attack/Decay
bb Sustain/Release bb Sustain/Release
cc
Instrument properties 1:
• Restart settings (bit 7-6)
• Arpeggio mode enable (bit 5)
• Oscillator reset (bit 4)
• Hard restart table pointer (bit 3-0)
cc
Instrument properties:
• Restart settings (bit 7 / $80)
• Filter pointer set enable (bit 6 / $40)
• Pulse pointer set disable (bit 5 /
$20)
• Oscilator reset (bit 4 / $10)
(Changed from $08 in driver 6.02
and forward
dd
Instrument properties 2
• Pulse and filter settings (bits 7-6)
• Arpeggio delta delay (bits 0-5).
dd
Filter table pointer (Set if bit 6 of
Instrument properties is set)
ee Resonance setting (If this one is
non-zero, filter will enable for the
channel the instrument is set on)
ee
Pulse table pointer (NOT set if bit 5
of Instrument properties is set)
ff Filter table pointer ff Wave table pointer
gg Pulse table pointer
hh Wave table pointer
31
===== page 32 =====
Driver 5.0x Driver 6.0x
Wave table: aa bb Wave table: aa bb
aa
Note offset (If bit 7 is true, the note is
fixed)
if aa=$7x it's a wave table command
• 7f = Jump to position bb
• 7e = Set wave table delay to bb
• 7d = Wait bb ticks (not implemented
yet)
Note offset (If bit 7 is true, the note is
fixed)
if aa=$7f, Jump to $bb
aa
bb Waveform bb Waveform
Driver v5.0x
Arpeggio (not used in driver v6.0x)
aa
Note offset
if aa=$7x "repeat X steps from the top of the current arpeggio"-command.
Driver 5.0x Driver 6.0x
Pulse table: aa bb cc Pulse table: aa bb
if aa < $10 (Add to pulse)
• aa = add pulse high
• bb = add pulse low
• cc = execution time (in frames/updates)
if aa < $80
• aa = Execution time of current pulse
program step
• bb = Pulse sweep value (reversed nibble).
Pulse sweep is store in reversed nibble.
if aa = $7f (Jump to index)
• aa = $7f
• bb = unused (could maybe be used for jump
count?)
• cc = new index
if aa >= $80 (Set pulse value)
• bb = New pulse value (if bb = xy, then new
pulse width is: $0yx0)
if aa >= $80 & aa < $90
• aa = $8x where x is the new high pulse
value
• bb = new low pulse value
• cc = execution time (in frames/update -
effectively wait time)
if aa = $7f (Jump to index)
• aa = $7f
• bb = new index
32
===== page 33 =====
Driver 5.0x Driver 6.0x
Filter table Filter table: aa bb
if aa < $10 (Add to pulse)
• 0a = add filter low value (Beware, these are
the opposite of the pulse table, where this
one is the high part!)
• bb = add filter high
• cc = execution time (in frames/updates)
if aa < $80,
• aa = time to execute filter program step
• bb = add to current filter value
if aa = $7f (Jump to index)
• aa = $7f
• bb = unused (could maybe be used for jump
count?)
• cc = new index
if aa = $80 (Set pulse value)
• aa = $80
• bb = new pulse value
if aa >= $80 & aa < $90
• aa = $8x where x is insignificant
• bb = new high filter value
• cc = execution time (in frames/update -
effectively wait time)
if aa = $7f (Jump to index)
• aa = $7f
• bb = new index
Driver 5.0x Driver 6.0x
Commands Commands
0x aa ?b
Set slide
aabb is the added value to the current
frequency value (only additive)
0X XX
Set slide up
XXX is the value added to the
current frequency value
1x aa bb
Set vibrato
aa = Frequency
?b = Amplitude
1X XX
Set slide down
XXX is the value subtracted from
the current frequency value
2x ?? aa
Set arpeggio
pointer
aa = Arpeggio index
2X YY
Set vibrato
X = Frequency
YY = Amplitude, absolute value
added/subtracted to/from frequence
value
3x aa bb
Set filter
and/or pulse
table pointer
x - Bit 0 => AA is a filter table pointer
Bit 1 => BB is a pulse table pointer
Ex.
x=0 =>AA and BB remain undefined
x=1 =>AA is set as new filter table
pointer
x=2 =>BB is set as new pulse table
pointer
x=3 =>AA is set as new filter table
pointer AND BB is set as new pulse
table pointer
3X YY
Set Filter
parameters
X = Band width
YY = Resonance and filter select
bits
33
===== page 34 =====
4x aa bb
Set wave
and/or pulse
table pointer
8x AD SR
9x AD SR
c0 xx ?? f0 xx ??
f1 xx xx
aa bb cc
Driver 5.0x Driver 6.0x
x - Bit 0 =>AA is a wave table pointer
- Bit 1 =>BB is a pulse table pointer
Ex.
x=0 =>AA and BB remain undefined
x=1 =>AA is set as new wave table
pointer
x=2 =>BB is set as new pulse table
pointer
x=3 =>AA is set as new wave table
pointer AND BB is set as new pulse
table pointer
4x YY
Set Filter
program
pointer
YY = new filter program pointer
ADSR is set for the scope of the
current instrument, including the
current note event
5x YY
Set pulse
program
pointer
YY = new pulse program pointer
(Direct)ADSR is set for the current
note event only, and reset to the
previous ADSR value on next event in
the instrument scope
6x YY
Set wave table
pointer
YY = new wave table pointer
Set Volume
8D SR
Set (A)DSR
Attack is always set to 0.. Bummer!
Set new tempo.
xx = pointer into tempo table
Set portamento. (Set until vibrato,
slide or new instrument is set!)
xx xx = Frequency variable
add/subract
Driver 6.0x
Init table
aa = Tempo table index
==> first comes filtersetting, then volume.
bb = Volume / Filter bandpass setting (i.e. $1f = $f volume $1 bandpass!)
cc = Filter Resonance / Filter channel enabled (i.e. $f1 = $f resonance, $01 = enable filter
channel)
Note that filter enable settings are bit:
bit 0 / $01 = channel 1
bit 1 / $02 = channel 2
bit 2 / $04 = channel 3
34
===== page 35 =====
Ninjatracker
Ninjatracker V2.x is based onto
V1.x that Cadaver created years ago
for having an editor that had low
rastertime and memory usage that
can be used for the music of his
games.
The editor was quite hard to use,
but that were an utility that convert
Goattracker tune to Ninjatracker for-
mat, so one could be concentrate only
in instruments fixing upon the tune
was imported into Ninjatracker.
The editor starts with his editor
page that have all you need to com-
pose into it.
The used colors in the editor are in
gray scale, maybe a combination that
look similar to the Goattracker editor,
and so familiar for the ones that use
GT. However, colors are configurable
by user that can change them and so
make a totally colored editor (look at
the image in right part for an
example).
The editor allow you to create up to
16 sub-tunes, and each sub-tune
have 3 tracks (up left part of the edi-
tor) where you enter the pattern num-
ber to play. Note that you have re-
served pattern numbers for transpose
up of down one patten:
Track Description
00 Loop
01-7F Pattern 01-7F
80-BF Transpose downwards
C0-FF Transpose upwards
In the upper right part there is the pattern editor where you can insert note/duration/commands of
the pattern. His syntax is:
Note / key on (+++) /
key off (---)
Command number (01-7F)
Legato (81-FF)
Note duration
Command name (just
for best looking)
35
===== page 36 =====
You can so indicate the note to play (C1-B7) and, in a notation that look likes tracker, if key stays
on (+++) or off (---). Then there is the space for the command to use (a command is an instrument
setting if one note is specified or effect to apply in the other cases) and the duration of the note. If no
one in inserted, it uses the previous values. In the last part it is reported the name associated with
the command. This is probably one of the best features as it is more simple to look at what you
have done in the pattern.
Note that you could use the editor in a manner very similar to a tracker if you set a fixed notes du-
ration and make use of key on/off.
Each command has this syntax:
ADSR WAVE table PULSE table FILTER table
You can so specify the Attack/Decay/Sustain/Release of instrument (not used if command is in
legato mode), and the pointer (00 means use actual running tables) to use for wave, pulse of filter
table commands.
In each table, there is the command and a parameter for the command:
Wave table Description
00-8F Set waveform, right side is arpeggio (00-7F relative, 8C-DF
absolute notes)
90-BF No waveform, delay arpeggio by 00-2F frames
C0-DF Vibrato with speed 00-1F, right side is depth
E0-FE Slide with speed highbyte 00-1E, right side is speed lowbyte
FF Jump, right side is destination, not to be entered directly from
a command
Pulse table Description
01-7F Modulate pulse for 01-7F frames, right side is signed mod.speed
80-FE Set pulse to right side value
FF Jump, right side is destination, can be entered from a command
Filter table Description
01-7F Modulate cutoff for 01-7F frames, right side is signed mod.speed
80-FE Set passband (left nybble-8), channels to be filtered (right
nybble) and cutoff (right side)
FF Jump, right side is destination, can be entered from a command
Note that Cadaver released many versions of v2.x (you can see the history in the news section).
However they different in some points like hardrestart type, notes duration and other minus effects.
This means that you have to choose the editor you need to use.
Maybe the best solution should be to have runtime loadable players in one common editor, like in
SIDFactory or JCH.
The other things to say about the editor is that you have a function that pack and optimize the pat-
36
===== page 37 =====
tern of the song when it is to release and a relocator that can even save the song without the player
for use as game effects in a game.
The editor use very low rastertime as in version v1.x even if now it has more potential. What is
missing is that you cannot use it for creating multispeed tune inside the editor.
Comparison
Below there is a table that sum some of the features of the two editors, just to try a comparison.
You will see a + or – signal for the features that I like and not like (this is about my taste, don't
take it too seriously. In fact, many features here are not related to quality of music you can create,
but are simple improvements or nice features).
Else the first features (tracker vs duration based editor) is maybe something that let you choose
one editor instead the other at the beginning without look at the other features. I myself prefer the
tracker ones editor as it is more evident the flow of what happens in the tune.
SID Factory Ninjatracker
Tracker based editor Duration based editor (but can be used
even in a manner similar to tracker)
Loadable drivers for having different
music features.
A change to the editor itself is
available to all the player
Different programs for having different
music features. For a programming
point of view this means to have many
versions of the program to
maintenance if you change something
into the editor
Inside the editor, all (patterns,
instruments, tables) are in one screen
Inside the editor, all are in one screen.
This is always a point that for me help
composing: switching to other screens
could cause to lose temporary the
vision of what you are being done
Files to load are selectable by a list.
This is the best way, as you don't have
to type the name of the file
The name of the files to load are to
insert by hand. With some other
editors that add an “ extension” this is
even more complicated
When loading a file does not show
anything until the file is loaded
When loading show in the border an
animated effect (this show you that the
operation is going on)
Source code is not available. Maybe
future additional thirty part
improvements to the editor could not
be realized in full manner
Source code is available. This means
that in future thirty part programmer
could improve/modify the editor.
37
===== page 38 =====
SID Factory Ninjatracker
No online help available inside the
editor. If you don't remember a key
function you have to read the offline
documentation. With today emulator
onto pc, maybe offline documentation
is just at one click distance, but if you
compose onto the C64 it is not so
simple
Online help inside the editor full of
information. Maybe it is always more
simple to look at it that going to search
in offline documentation.
Offline documentation about the editor
is available.
Offline documentation about the editor
is available.
Only one kind of insertion mode for
notes
ProTrack or DMC notes insertion mode
(for sure this help if you come from
other editors)
Instruments are based onto tables.
Even if for beginners composers the
use of table is hard, only with them
you can get the best varieties of sound
from the sid chip
Instruments are based onto tables of
values.
There is a function that pack the tune There is a function for optimize
patterns
I not find a way to relocate the code to
a given address (maybe a features not
present into this alfa version)
Relocator is available (with
Gamemusic option, so you can even
create sound effects to use inside a
game)
Example tunes are available.
Examples are very important for
looking at the power of the editor and
how it works when you try it the first
time
Example tunes are available
Multispeed in a natural way with a
specific player. I can say that this is
the first editor I try that has this
features (in other, if you add
multispeed to a tune, you have to
adjust instruments to the new mode)
Not multispeed tunes can be created
No subtunes availables (this actually is
a point that reduce the possibility to
use it for creating music for
game/demo unless you need only one
tune to play)
Max 16 subtunes
Max 32 instruments (It is a good
number) and max 64 commands
Max 127 commands (a command
inside a note is an instrument)
38
===== page 39 =====
SID Factory Ninjatracker
Fixed colors (even if the color looks
good, maybe someone could be happy
to customize them)
Customization of colors is allowed. If
someone don't like the gray scale color
used, he can change them
No fast forward option found Fast forward when playing
Only raster time and time passed is
showed when playing the tune (no
note being play is showed)
Only rastertime usage is showed (no
note being played is showed)
Instruments cannot be saved for use in
other tunes. This means that you have
to create the instruments form scratch
each time you compose.
Instruments cannot be saved for use in
other tunes
Conclusion
Summing my comparison table, I have that Ninjatraker has 15+ and 4-
, while SIDFactory has 10+
and 9-
. That simple means that all the two editors have space for a little improvements and maybe
SIDFactory is a little behind in my comparison by the fact that it is in alpha state and so no all the
simple features are inserted. Think for example at the online help that it is a powerful tool for using
the editor until you have learn all his keys/features.
In particular only 2 features are missing from the two editors the same time: the ability to show in
real time what it is played for each voices (for a tracker that could be more natural to implement re-
spect to a duration based editor), and the ability to load/save the instrument you have created.
However I hope that this little battle for the two editors give you the idea to test them and produce
some quality sid music and maybe give some suggestion for the editors authors to improve them.
39
===== page 40 =====
40
===== page 41 =====
SIDin 11 end
41