Lets not worry too much about silicon - it's ridiculously cheap these days,
so long as you are not going for the top end. If we are to use cheap, ready
built, USB units and not our own purpose built kit (and that's not
unthinkable, this is the age of the maker, right?) I think it all hinges on
two things, synchronisation and bit rate. For 24 bits and 48 Khz, which is,
I guess, about the lowest we'd all be happy with, that's 1152 kilobits per
second, which means for USB 2, a practical limit of 8/10 channels per plug,
all things being equal. This is why there are 7.1 units out there. For USB
3 this goes up by a factor of 20, at least theoretically. Whether having
more than 1 USB socket on your computer helps is implementation dependent.
Synchronisation is the real killer for cheap units as they frequently don't
have clock input so the the clocks would have to be free running. This
could easily run into 360mS of error (at the typical error of 0.01% on
xtal clocks) at an hour into a piece. Having some of your speakers
effectively 100 m further away is _not_ good [?] . That's what you could
get, tho' some systems will deal with this by designating one interface as
the master, and sending data via a sample rate converter to the other(s).
This keeps the audio in sync tho' quality might suffer because of the src.
I have to say that my experience of the src in the Mac works really well
and I've run 24 outputs of this Mac with one set of 8 outputs (on an FA101)
going through the internal src without any noticeable problems. (albeit
that was Firewire and not USB).
My own personal feeling is that using a dedicated processor, preferably ARM
based, and the TDI interface, possibly using an RJ45 plug and ethernet
cabling, would be the best approach- on the other hand, we saw a plethora
of cheap multichannel boards talked about a few days ago, why not go for
them?
Dave
On 6 April 2014 03:38, Sampo Syreeni <[email protected]> wrote:
> On 2014-04-06, Augustine Leudar wrote:
>
> Can you do this with any 7.1 usb card ? Wich did you use ?
>>
>
> Well now you opened a can of worms even on my behalf. What really is the
> minimal, asymptotic per-channel cost of well-synchronized multichannel
> audio D/A? Let's say, presuming you can only draw whatever power you can
> from an USB3 port, or a PoE enabled Gigabit Ethernet one? Then presupposing
> at least CD level 98dB (linear unfiltered SPL) S/N for any outbound signal
> is required? How many outbound analog, balanced connections could you
> honestly claim to be able to do, at the standard levels? And once you got
> that, what's the cost-against-number-of-chn curve? What's the curve against
> diminishing voltage, assuming as little output current as you can achieve
> over our outputs in average? How would you achieve the optimum scaling
> using extant mass-produced, mutually syncronizable to
> sampling-time-accurate D/A-chips at your disposal? Which were the chips
> precisely which you would utilize? Are they really the cheapest out there
> even taking into account the external components they require in order to
> function properly, and how/why do they scale out/in-parallel?
>
> Seriously, guys, this is like the Grail of low cost ambisonic on the
> hardware side. Right now we can do first order at four hifi outputs, but
> quite likely we can't even do eight for second order. Simply forget about
> third or fourth order using any current standard interconnect, either
> because you can't get the data across, or because nobody likes an
> extraneous transformer for something like this.
>
> So if we do it as compatibly and harmlessly and most efficiently as we
> can, all round, where're connector count, power and datarate curves per
> dollar, here?
> --
> Sampo Syreeni, aka decoy - [email protected], http://decoy.iki.fi/front
> +358-40-3255353, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
> _______________________________________________
> Sursound mailing list
> [email protected]
> https://mail.music.vt.edu/mailman/listinfo/sursound
>
--
As of 1st October 2012, I have retired from the University.
These are my own views and may or may not be shared by the University
Dave Malham
Honorary Fellow, Department of Music
The University of York
York YO10 5DD
UK
'Ambisonics - Component Imaging for Audio'
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20140406/9ed8c10b/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 104 bytes
Desc: not available
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20140406/9ed8c10b/attachment.gif>
_______________________________________________
Sursound mailing list
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound
