SergeModular

Hi Matt
I'm not sure I understand the fix you suggest, unless you were having a different problem.
Is what you are saying is patch the ACPR to the INV then to the PCO?

What I am experiencing is best demonstrated by this simple patch: With nothing
connected to the ACPR, patch its output to the 1V/oct of a PCO. Then sweep the xfade
control from 1 to 2. I get an arc of pitch change - 0 at either extreme, but about 1/2 step
sharp at 12:00.

When a square wave (gate) is applied to the VC xfade the results are not what I would
expect. With offset on "1", there is a bit of warble in time with the gate. At 12:00 it
becomes a whole step interval, and at "2" declines to a little less than a 1/2 step.

Patching the INV in between only raises and lowers the overall pitch. That was a helpful
tip, though. It seems that certain cv processors come with a certain bias of + or - that
would be corrected more easily this way. Thanks for that. But maybe you had a different
patch in mind.

Let me know what you think of the 2x8 patch. I've had fun setting pitches in pairs,
working from the bottom up, changing the reset point for different lengths. Incorporating
the PDIV to alter the xfade relationship to the clock yields interesting results. (As always in
Serge-Speak: interesting=fun)

Thanks for the feedback
John R.

--- In SergeModular@yahoogroups.com, "matthew carpenter" <matfhew.carpenfer@...>
wrote:

>
> Hi John,
> I had the same experience with all of my ACPRs, actually, and determined
> that it was helpful to 'buffer' the pitch CV by putting it through the INV
> section of the ACPR first. It took me awhile to narrow it down to the ACPR,
> though, as I thought it involved the PDIV and BLOG, but Rex said it's
> basically normal and suggested the 'fix' I just mentioned.
> Thanks for continuing the discussion. I can't recall if I've tried your
> particular technique. I'll patch it up today...
> Matt
>
> On Tue, Jun 17, 2008 at 9:17 AM, johnrichoux <jrichoux@...>
> wrote:
>
> > Hi Matt
> >
> > I was interested in this question and finally got around to testing it
> > myself. I found that I
> > can make a 16 step sequence out of the 8x2 by simply applying the same
> > clock (DTG gate)
> > to both the SEQ and the ACPR. With each pulse, it switches from A to B,
> > then the SEQ
> > advances one stage. The DTG gate has to be as square as you can get it.
> > (Wish I had a VC
> > Clock module)
> >
> > However, using this method to control pitch, I get a significant warble. It
> > is less noticeable
> > using the method suggested earlier, as the crossfade happens only once, but
> > with this
> > method it happens 8 times. I have the ACPR set full left. Even when the SEQ
> > is preset to
> > the last stage, and both A & B are set to zero, there is a distinct pitch
> > shift.
> >
> > The warble is even more pronounced when I run the clock thru the PDIV. The
> > greater the
> > division, more warble.
> >
> > I have run into similar pitch accuracy problems with the ACPR before. I
> > have ACPR's on
> > another panel, I should test with those.
> >
> > Thanks for raising the question
> > John Richoux
> >
> >
> >
>