FW: multiple-type Digisound VCDO thing?

[David Halliday (Volt Computer)]

> -----Original Message-----
> From:	steve at babcom.u-net.com [SMTP:steve at babcom.u-net.com]
> Sent:	Tuesday, April 29, 1997 11:43 AM
> To:	synth-diy at horus.sara.nl
> Subject:	Re: multiple-type Digisound VCDO thing?
> 
> <previous posts deleted>
> 
> I like this idea! It strikes me that with 2 eproms, dacs, etc it would
> be
> possible to crossfade between the outputs using a pair of vcas.
> 
> You could use a simple triangle/square type lfo, with the triangle
> wave
> controlling the vcas (one would be fed from an inverted version of the
> triangle wave) and the square wave selecting the wave addresses.
> (using the
> rising edge of the square wave for one eprom, the falling edge for the
> other.)
> 
> the idea would be that each eprom would switch to a different waveform
> at
> the point where it's output was crossfaded down to zero volume, so the
> net
> result would be a smooth continuously shifting waveshape a la
> Wavestation/Microwave.
> 
> If a vclfo was used, the crossfade rate could be voltage controlled.
> 
> Of course, I'm probably forgetting something...
> 
> 
> 
> 
> 
> The timing would be hellacious but...  <grin>
> 
> You could use a small microprocessor to generate the waveforms on the
> fly - loading them into static RAM.
> This RAM would be read by your D/A.  
> 
> You could also generate a *huge* number of waveshapes ( 128, 256,
> 512... ??? ) and stash them in a largish EPROM to be retrieved by the
> CPU.  The CPU would then "stitch" them together as you wanted...  
> 
> The CPU is also very good at doing fast integer math so altering them
> on the fly - adding two together, crossfading two, morphing them,
> etc... would be easy.
> 
> You could have libraries of starting forms and middle forms - starting
> forms would go through zero at a constant point and would go to a
> certain amplitude ( maybe ten or so specified values ) where the
> middle form would take over and then hand off to another starting form
> ( played backwords )
> 
> A bit of software for a computer with soundcard would be good for
> designing the waveforms - something with set points and spline grab
> points to adjust the slope of the curve...
> 
> 
> 
> Also, while we are spitting out voltages...    Why not provide for
> some control voltage outputs as well... You could have an
> VCD-oscillator that would have a VCF output which offered dynamic
> tracking and resonance control...
>