11 Jul
2023
11 Jul
'23
1:19 p.m.
On Jul 11, 2023, at 7:43 AM, steven(a)malikoff.com
steven--- via cctalk <cctalk(a)classiccmp.org> wrote:
Speaking of old computerized music playing technology, there are two from the PLATO system
at the University of Illinois that are perhaps the earliest of all in their category and
not all that well known.
The first of the two is the GSW (Gooch Synthetic Woodwind), which is a four voice, 7
levels per voice, square wave synthesizer. It's fully documented in Sherwin's US
patent 4,206,675. That one was attached to the auxiliary device port of a PLATO terminal
and driven from the host computer, at 1200 bps. It worked quite well for playing music
and was widely used for music education. It's a very simple device as you can see
from the full schematics (which are, surprisingly, given in the patent). That patent was
filed in 1977 but the invention is somewhat older, perhaps 1975 or 1976.
The followon to that is the GCS (Gooch Cybernetic Synthesizer), unfortunately not well
documented. That's a 16 voice programmable waveform (256 words by 16 bits per voice),
more levels (256?), driven as a peripheral off the 8080-based "programmable PLATO
terminal" from a program running in that terminal. So the musical score level
definition of what to play still came from the host, still at 1200 bps, but the
attack/decay etc. shaping would happen in the terminal. That one was a bit of an
electrical muddle, with memory, logic, and D/A per voice followed by a 16 input combiner.
Getting the analog parts to work right was a hairy task with far too many trimpots.
Sherwin vowed that any followup would be digital all the way to one final D/A, which of
course later became the answer in the PC sound cards, but if he did that it was after I
left. The GCS was built around 1977. There were some interesting related pieces of work,
such as a speed-sensing piano keyboard (so unlike an electronic organ you could have
dynamics, exactly as on a piano), a music editing system with a score printing program to
print on a dot matric (electrostatic) printer, and some other stuff.
I'm not sure if the GCS is the earliest fully programmable waveform digital music
synthesizer, but if not it is close.
paul
On 07/10/2023 11:31 PM AEST Mike Katz via cctalk
<cctalk(a)classiccmp.org> wrote:
Way back in the 80's I was able to do stereo 4 part harmony on a 2 MHZ
6809 using two 8-bit D/A converters.
Much the same here. I recounted this on VCFed a few months ago about building a simple
2-chip 8-bit ladder DAC with one-transistor amplifier for my Applied Technology DG680 S100
machine back in the early 80s from this absolutely excellent BYTE article on how to do
polyphonic synthesis on a microcomputer (KIM-1):
https://archive.org/details/byte-magazine-1977-09/page/n63/mode/2up
A schoolfriend who had an Apple ][ and had not done any Z80 machine code before asked for
me to hand him my Zaks book, upon which he wrote out one attempt in Z80, crossed it out
and wrote a second version. Which worked perfectly. For the music piece I got it to play
four-voice polyphony after painstakingly encoding Bach's Praeludium in C Major from my
mothers' collection of piano music scores.
A few years ago I had thoughts about porting the 6502 code to the PDP-11 and use the same
sort of ladder DAC. Not sure if the slimline 11/05 would be fast enough for anything too
high frequency, but if it was, the slimline 05's power supply could then temporarily
come out and be perhaps be powered off some beefy batteries in that space, along with a
small 1970s transistor amp and 1970s headphones topped off with a leather shoulder strap
to lug it around like a giant Walkman.