[sdiy] VCO jitter, Slop ... (was: Re: Smith's Evolver Desktop's spikey VCO waveforms)
rsdio at sounds.wa.com
rsdio at sounds.wa.com
Mon Jun 10 03:08:09 CEST 2013
One variable that I forgot to mention in my earlier reply is the final output
stage of a synth. Whether pure digital DAC, DCO, or VCO, there is always a
final analog output stage (with the exception of pure digital burned to CD
without running through a "synth" of any kind). That output stage will vary in
quality from one synth model to another, and I suspect that changes happened
in parallel, such that the final output stages in the vintage analog heyday
were probably different than the final output stages in the wave of DCO synths.
I know that proper A/B/X testing requires careful volume matching, otherwise
the louder sound is preferred and that falsely implies that it is better. Is
it possible that earlier analog VCO synths all have higher voltage power rails
and thus produce a "better" sound than later DCOs? I can imagine that a DCO is
possibly more likely to run at 5V than +/-15V, although I haven't checked the
schematics for the famous DCOs. I do know that the Prophet 5 put a lot of
circuitry into boosting the 5V digital control signals to 10V so that the
analog section could run on +/-15V.
I also know that my Pro-One seems beefier than my Matrix-12, but is that all
due to the VCO? Could it be the really hot output stage of the Pro-One?
What I'd like to see is a test circuit that pits three or four kinds of
oscillators against each other - soft-synth DAC, DCO, log and/or linear VCO -
but all through the same final output stage. I guess it goes without saying
that there should be no VCF, or at least the same VCF after each oscillator.
Sorry if I'm re-visiting old tests of VCO versus DCO that may have been done
before, but I'm not aware of anyone comparing isolated components like this.
Instead, I assume most people are comparing the whole synth. Of course,
modulars have the ability to isolate components, but it seems to me that
certain VCO configurations were never available in modular form; same with the
DCO configurations. I suppose that modern modular recreations of classic
analog VCOs and DCOs might be valid.
Anyway, that's a lot of text for a simple idea. Has anyone tested the
contribution of the final output stage to the fatness of classic synths? Is
there any way the modern soft-synth can be beefed up in this way, rather than
trying to focus on the oscillator itself?
More information about the Synth-diy