[sdiy] MIDI VCO Control - but not CV?

[harrybissell at wowway.com]

You need to have a stable amplitude of the core waveform.
Best way is to make it correctly in the first place, or you might
attempt to 'compand' it, but that would take a number of cycles
at best. During that time there would be transient distortion.

I attempt to do this in a guitar systhesizer application... but there
is a two cycle delay before the correct amplitude is reached... I allow
the one wrong cycle to be clipped at what should be the maximum amplitude.
During a guitar attack, you could never notice, but I'm just making a sawtooth
from this and the harmonic difference is not that obvious.

Likewise your idea to produce the CV after the fact, with a filter... will
introduce a huge delay.  If you measure the time directly, you could get to be
as good as one cycle behind... that could still be many milliseconds late.

I once had the idea to use a frequency shifter to raise a guitar's pitch into
the ultrasonic range, do a P/V conversion then to eliminate delay. The 'dome
filter' that you need takes maybe 18ms for the signal to get through it...
worse than the 12ms of the (lowest) string itself.  

Can't always get what you want...?    :^)

H^) harry


On Tue, 15 Jul 2008 17:20:47 +0100, cheater cheater wrote
> What could be done to circumvent the problem of the shapers that
> create the sines?
> Cheers
> 
> On 7/15/08, harrybissell at wowway.com <harrybissell at wowway.com> wrote:
> > OK I get the idea about the pulse generator.
> >
> >  There is a Roland DCO that uses a micro output to generate
> >  an approximately correct current value to charge a capacitor
> >  to full voltage at the desired frequency  (higher frequencies need
> >  more charging current) and then makes a Digitally controlled
> >  reset pulse.
> >
> >  Any inaccuracies in the analog section therefore become errors of
> >  sawtooth amplitude, not of frequency. Your ear is much more sensitive to
> >  frequency than amplituce, so its a useful trade.
> >
> >  You would never hear this in a sawtooth, or a pulse derived from that
> >  saw. In one case its a small error in pulse width (probably being PWM
> >  anyway) or a small error in amplitude.
> >
> >  Try and use that sawtooth in a waveshaper (convert to triangle, or sine)
> >  and the shape errors cause MAJOR changes in the harmonic structure...and
these
> >  waves have very little harmonics so the difference is really obvious
> >  (sticks out like a sore thumb...)
> >
> >  If you got the control current really close, the error could be minimized...
> >  and this would be useful in a poly-synth especially. Trying to retrofir this
> >  to a conventional VCO core is certainly possible, but not with just a precise
> >  reset pulse. You need CV to be close, anyway...
> >
> >  H^) harry
> >
> >
> >
> >
> >  On Tue, 15 Jul 2008 14:57:18 +0100, cheater cheater wrote
> >
> > > Harry,
> >  > this is not about midi sync.
> >  > This is syncing to a pulse generator that has the pitch dictate by
> >  > the incoming midi note.
> >  >
> >  > Also, this is about trying to sync the core only - shouldn't that
> >  > work well? The later shaping etc that goes into making the
> >  > triangle/saw a sine wave would work as usual...
> >  >
> >  > Cheers
> >  > Damian
> >  > On 7/15/08, harrybissell at wowway.com <harrybissell at wowway.com> wrote:
> >  > > Hard-sync of a VCO would be VERY noticible...
> >  > >
> >  > >  A sine wave with even a tiny notch in it is extremely audible !!!
> >  > >
> >  > >  Midi sync would not have the speed to reset often enough...
> >  > >
> >  > >  If you did good MIDI to CV and added some sync it might work, but
for that
> >  > >  effort a better MIDI to CV converter would be the best use of the
time and
> >  > >  money IMHO...
> >  > >
> >  > >  H^) harry
> >  > >
> >  > >
> >  > >
> >  > >  On Mon, 14 Jul 2008 18:44:12 +0100, cheater cheater wrote
> >  > >
> >  > > > Hi guys,
> >  > >  > Looking at the SDIY08 photos, I was wondering about MIDI2CV modules
> >  > >  > and the problems they're (normally) having.
> >  > >  >
> >  > >  > Wouldn't it be *much* easier to make a midi2trigger module, that
> >  > >  > would then hardsync the VCOs?
> >  > >  >
> >  > >  > You wouldn't necessarily get bad jitter, since the MCUs work at 10s
> >  > >  > of MHz nowadays... that's enough resolution even for the best
> >  > >  > 'golden ears' 8^)
> >  > >  >
> >  > >  > This could also mean easier microtunings (?)
> >  > >  >
> >  > >  > Are there any popular/interesting VCO designs that wouldn't work well
> >  > >  > with being synced this way?
> >  > >  >
> >  > >  > How extremely difficult would it be to make a filter track pitch
> >  > >  > like this?
> >  > >  >
> >  > >  > Having asked that, a filter's pitch is nowhere near as important as a
> >  > >  > VCO's, so it could probably use 'traditional' MIDI methods anyways :)
> >  > >  >
> >  > >  > Just wondering what the synth diy gurus here think about this.
> >  > >  >
> >  > >  > Cheers
> >  > >
> >  > > > _______________________________________________
> >  > >  > Synth-diy mailing list
> >  > >  > Synth-diy at dropmix.xs4all.nl
> >  > >  > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >  > >
> >  > >
> >  > >
> >  > > Harry Bissell & Nora Abdullah 4eva
> >  > >
> >  > >
> >  > _______________________________________________
> >  > Synth-diy mailing list
> >  > Synth-diy at dropmix.xs4all.nl
> >  > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >
> >
> >  Harry Bissell & Nora Abdullah 4eva
> >
> >
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy


Harry Bissell & Nora Abdullah 4eva