CompuServe Messages

Art: It's probably a lot like my programs "Wink" and "8-pus" which play 8 voices on amiga in real time: I'm using additive synthesis. Note that 16 voices makes the samples effectively 4-bits, so the fidelity is not so hot (unless some time-consuming compansion is also going on). On a 7Mhz machine, the best I coud manage was about 7Khz of sound before the double buffers couldn't flip over. That was with 8 voices and not much processing overhead, in assembler using some unrolled, register intensive code. Though on my 25-Mhz machine, there's no problem getting good fidelity (which is after all related more to the sample rate than the number of bits in the sample , the output filter acting as an integrator…), and in fact I've bumped wink and 8-pus up to 20MHz or so , 16 voices on a 7Mhz machine may not sound too hot. (especially with all 16 going at once) On the other hand, B&P doesn't need to monitor the incoming MIDI stream like 8-pus and Wink do, and another trick, involving alternating samples from two banks of sound instead of adding them can seem to allow more accurate sample lengths (at half the sample rate, if used, at full sample rate if not used…the trick is to only use it when all other bandwidth is filled up…). Ah, sweet music! As simple as a rustic pan-pipe or the burbling of a daydreaming child! —> the amusing JHHL <—- a message not processed by WHAP!