[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: PC sound cards

On 30 Jan 03, at 9:59, Pallier Christophe wrote:

> On Thu, 30 Jan 2003, David Isherwood wrote:
> > The main problem with such soundcards is that ALL internal processing
> > is done at a sampling rate of 48kHz, which means that any audio stream
> > having a different sampling rate than this must be upsampled at the i/p
> > buss and downsampled at the o/p. The quality of this onboard sample rate
> > conversion can vary greatly from card to card with some soundcards
> > producing very noticeable artifacts.
> >
> Am I correct in believing that this implies that I should better
> work with sampling rates that are integer dividers of 48 Khz (e.g. 16
> Khz rather than, say, 22050 Hz)? I hope all sample rate
> conversion algorithms should be able to correctly oversample by an
> integer (?)

Unfortunately, I belive it is not that simple.  The problem is that (as
far as I can tell... and I'd love to be proven wrong) is that there is no
mechanism for software to determine what Windows is doing.  You can only
specify an intended sample rate, and Windows decides about sample rate
conversion.  There is no way to tell what rates Windows will or will not
apply conversion to, nor even what rates are natively supported by the
card.  (There is a Windows API function that is supposed to report on
capabilities, but it gives bogus results.)

I have heard complaints of people getting slightly different record
and playback rates when using both simultaneously (full-duplex mode), even
when selecting standard rates like 48 kHz or 44.1 kHz.  Since the same
clock is certainly used by the hardware, I can only assume that this is
some screw-up in the Windows rate conversion.

Robert Masta

         D A Q A R T A
Data AcQuisition And Real-Time Analysis
 Shareware from Interstellar Research