On Tue, Apr 04, 2006 at 09:14:44AM +0100, Paul Mullen wrote:
>
> One concern is that you will need to build a breakout box for your A/D
> converters or else rely on someone else's converter box and just support
> something like ADAT lightpipe I/O but at that stage you have pushed yourself
> downmarket unless you add something to differentiate yourself like a DSP
> farm.
Well, speaking as an old analog instrumentation engineer, my instinct
would be to push the A/Ds right up to the mike connector when you're trying
to achieve that level of precision. The cable alone is a major error source
when you're playing in that league. You get half a dozen kinds of
electromagnetic coupling right through the shield, triboelectric noise, and
on and on. So it's not a card design problem, it's a total system problem.
And I sure wouldn't try to design the DACs and A/Ds for that performance
level; there are engineers a lot smarter than I am who've spent a whole
career doing that. And it doesn't cost all that much to buy the resulting
chips off the shelf.
I'd say the place to start with something like this is in the market
research. Begin by finding out what the users need and want, and what's
already out there. It might be best to design around remote analog hardware
that already exists, and pick up the signal already digitized to build an
open-ABI hardware platform.
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
