Exactly what I have been saying before, James.

According to a bunch of engineers and DSP programmers I spoke to over
the last couple of years, it is not hard at all to develop such a
device. Sampling rate and resolution are not the big issue anymore,
although hi-end AD/DA's are still exponentially more expensive than
their 8- or 16bit counterparts. 96k/24b is what I heard most as a
feasible figure, but higher would still be better. And this is exactly
what Volta solved so brilliantly (in all its simplicity): just use
your DAW's sampling rate. If your system can pull 192kHz @ 24b...
perfect; but you'd still do fine with a more regular setting of 44/24...

What most of the tech guys I spoke with saw as a bigger challenge
(read: problem), is the software. The GUI and such... And again, I
feel that Volta approached this quite well. Simple, but effective, and
apparently also rather flexible (did anyone see its amazing quadrature
LFO yet -- http://motu-volta.blogspot.com/ )

As you already said, James, other hardware is going to follow soon,
most likely as a more dedicated device instead of "just" a DC-capable
audio interface. And I reckon the accompanying software is going to
follow soon too, once people are getting used to this new paradigm in
modular control, and will provide feedback and propose enhancements to
the manufacturers.

Well, at least I hope so.
I also kinda hope Dieter is having sleepless nights over this ;-)


_g

On 27 Jan 2009, at 02:36, James Husted wrote:
> I don't think it would be very hard to modify a typical Audio A/D
> converter design to do the job needed here. Remove the AC coupling,
> change the input scale to handle +/-5 Volts and keep it sampling at
> 96k and 24bits. My math has never been my strong suit but I imagine
> 24bits should cover a 10volt swing pretty easily with the resolution
> needed, and 96kHz should be fine enough timing wise. This could almost
> be done with an adaptor that precision scaled the analog +/-5V down to
> a range that a typical computer A/D could handle (and of course do the
> opposite for the output) and you could then use any interface that
> wasn't AC coupled (hard to find apparently). In any case I can't see
> it being that hard for a company to make a A/D converter that can
> sample modular CV ranges at the speed that typical audio interfaces do
> now and shoot the info to a DAW in a format it could understand. Like
> I've said before on this list - if the Volta take off you will see a
> hardware solution very soon after. It is not like they have to invent
> something from the ground floor.
> -James
>
> On Jan 26, 2009, at 4:17 PM, achtung_999 wrote:
> >
> > > I would say that 22.5 KHZ would already be quite sufficient for
> > > > tracking CV,
> > > > everything above that is luxery.
> > > >
> > > Oh no, certainly not. I'd say 44KHz or thereabouts is the absolute
> > > minimum... I leave the math explanation to the math-oriented
> > people in
> > > here. My coffee has already worn out...
> >
> >
> > Yes I know that. And that still does not answer my question really..
> > Let's say we have an pseudo-ideal CV> digital converter. It would
> > work from
> > 0 volts to anything, so no upper limit. (Just assume this..)
> > I would shift the CV up by the amount it needs to have the most
> > extreme
> > negative amplitude point to be 0 volts.
> > Let's assume this is 5 volts. My device now tracks it nicely and
> > records it
> > in whichever imaginative software format we might think of.
> > I now would like to use this signal to drive my modular.
> > I would take the output from the software via my reverse device and
> > shift
> > the resulting voltages down 5 volts..
> >
> > In my opinion this would not be different than the original signal.
> > would
> > it
> > Sorry if I was unclear about this.
> > And yes, I know it's a hassle ;-)
> >
>