[sdiy] modern CEMs?

[René Schmitz]

Hi Antti and all,

> Excepting oscillator drift/jitter/warmth, everything else in a synth can 
> be modelled without much trouble. 

:-P

Once you understand it, yes. But there are many attributes that we may 
not understand and thus can model. And the question if a model can be 
computed efficiently (within device constraints) remains still, so that 
for an implementation some corners might have to be cut. If I'm copying 
a schematic I'll get even those things right which I don't understand.

> And nothing stops you from mixing and 
> matching "components" or even synthesis models. Ever wanted to see how a 
> Moog filter would fit into DX7? Now you can.

And I may ask why is this always being so retro oriented, should this 
not open some totally new soundscapes instead? I mean synthesized sound 
when it was new gave rise to something unheared before. Shouldn't there 
be a new quantum leap?

The question isn't really that of how much horsepower you have, but what 
to do with it. Speed gets you nowhere if you're headed in the wrong 
direction. And I think that you don't have to wait for a fancy new FPGA 
to appear, you could even do a lot more on a regular PC than you might 
think, if one were to do away with the ubiquitous desktop (or server) 
OSes that are usually run on these, and use just a realtime OS, do away 
with virtual memory and run your code practically on the bare metal. Yet 
no one seems to do that? Its just a little coding exercise. Not. :-P

Cheers,
  René

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159