Asger Alstrup
Author of the two Examination of SID … waveform pages on Andreas Varga's sid-homepage (1995; HTML by andreas-varga) — the earliest register-level measurements of the sid in this wiki. He read voice 3's output register $D41B on a real commodore-64 once per clock cycle through a REU, and from the data derived how the triangle/sawtooth counter behaves under the test bit and how the noise generator works: a 23-bit shift register fed by bit 22 XOR bit 17, restarted from $7FFFF8 by the test bit — but only $2000–$8000 cycles later — and repeating after about 8 million values (source: s-sidmusic-alstrup-waveform-examinations).
Facts#
- The noise loop length was established with Marko Mäkelä: a first Loopchecker "hacked together over the IRC" gave inconsistent results (16-cycle sampling, the delayed test-bit reset); the corrected program found the loop after 272 seconds at frequency
$8000, hence 23 bits. - The triangle page is dated by its own text: "as of now (23 Mar 1995)"; two questions were left open — whether the analogue output at pin 27 is really 8-bit and whether it is aligned to the Ø2 clock ("I have no access to an oscilloscope").
- His practical advice for using the noise as a random number source in games — reset once, sample after a key press, read at the shift rate, or precompute a table so voice 3 stays free for music — is on randomization-and-probability.
(source: s-sidmusic-alstrup-waveform-examinations). Only this pair of articles is in the wiki; nothing else about him is sourced.
Related#
sid · sid-homepage · andreas-varga · ring-modulation-and-sync · randomization-and-probability · wavetable-programming