🚀 go-pugleaf

In article <8901121633.AA12910@TIS.COM> dmb@TIS.COM (David M. Baggett) writes:
>
>   Has anyone experimented with this?  (We've all heard Starglider...)  Looking
>in ST Internals, I found that the PSG sound chip has an 8 bit D/A converter.  
>Unfortunately, it doesn't look like you can change the value directly; you have
>to go through the waveform synthesis stuff (or the frequency/volume registers).
>   8 bits of resolution produces decent (listenable) quality sound; does anyone
>know how/if it is possible to change the 8 bit digital value directly?  The only
>alternative I can see is using the noise generator and the volume control, but this
>only gives four bits of resolution (awful).  The Stargilder intro sounds to be about
>6 bit resolution -- anyone know how this was done?
>   Maintaining timing for the samples is straightforward with Xbtimer, so that isn't
>a problem.  44KHz is "correct"; 22KHz should be ok...

An 8-bit D/A converter?  From the specs I've seen, I thought it had a trio of
4-bit non-linear D/A converters.

You can get a little more than 4 bits of resolution by setting the different
sound channels to different volumes; e.g. by setting channels A and B to, say,
level 6 and channel C to level 7, you get something between levels 6 & 7.

Also, you wouldn't need nearly that many samples.  A sampling rate of 6KHz
at 6 bits per sample sounds like a typical telephone.  A sampling rate of
10KHz should push the Atari's PSG and the monitor's speaker to their practical
limits.

(Sampling like this was popular on the Radio Shack Color Computer a few years
ago.  That machine has a 6-bit D/A converter hooked directly up to its sound
output, and a 4-channel 6-bit A/D converter that could take samples at about
12KHz, but most people dropped down to 6KHz to conserve memory.)

-- 
"If I seem insensitive to what you're going through,
	keep in mind that's the way I am." -- Mr. Spock