Tim.Cockram at Tim.Cockram at
Fri Apr 16 10:41:59 CEST 1999

Hi Due to a change job I will be leaving the hallowed ranks of synth-diy for a
while.  I hope that my new job will give me more time to devote diying and
enable me to contribute more to the group.

Just a few comments to end this session:-

Thanks to Joachim for the superb Arp tech site.

I believe that EMS did have a US patent or two relating to a digital synthesis
method involving 256 (?) oscillators.  There is a great example of it's use on
the EMS Synthi flexidisk (doing speech synthesis).

Re the flaming etc.  THERE is NO IDEAL vco, vcf, vca.  It's the different non
ideal characteristics that make them desirable (that's certainly why I'm
here).  From a theoretical point of view the "matched" transistors in the
3046/86 are pretty dubious however there are many great sounding vco's that
use them.  In the mid eighties I experimented with an LM 324 exponential
converter that was temperature controlled using a peltier device with a  3
term control loop (I was using a similar scheme to measure cutaneous blood
flow in humans  at the time).  It worked very well but used quite a lot of
power and SOUNDED rather boring.  The scarcity of tempco resistors does make
this approach a little more difficult  but I (we?) didn't get in to analog
synths without wanting to be challenged a bit.
(Farnell Mercator sell tempco's in the UK)

Please let's just share information and constructive comment.  If it sounds
good to you it doesn't matter about anyone else's opinions.  I have personal
experience of this as a number of years ago a friend and I built a wave table
based voice with a few commercial dreams but we abandoned it because others
didn't like the aliasing and other junk,  when I heard a PPG a couple of years
later I realised that the "junk" was such an important part of it's sound.

Re the sax scream,  there has been a lot of discussion over the years about
the Serge wave shapers, how about the application of a little scientific
method?  If one inputs a linear sawtooth of various amplitudes with different
control voltages it should be possible to establish the devices transfer
function.  Whilst I'm not suggesting that anyone violates Sound
Transform/Serges intellectual property rights it would provide a great
starting point for group members experiments.

Any How




If you'd care to share Ideas / abuse by snail mail my new work address is:
EMAP Radio
Radio House
900 Herries Road
S6 RH1


More information about the Synth-diy mailing list