There are a lot of debates and assumptions about the way these USB audio
boxes are clocked.

My understanding is that the operating systems fetch a sample rate in
the file header, then manage to create some clocking at this sample
rate, and eventually pushes out the data bits through the USB to a
"class compliant" audio device.

So I decided to edit the sample rate inside the header of a WAV file.
Easy: just locate the two bytes 44 AC somewhere very close to the
beginning of the file, and edit it with a hexadecimal editor. I use
WinHex for that purpose.
Note: the two bytes are in reversed order, actually 44100 converted to
hex is AC44.
Let's try B333, or 45875. Why? Why not!

For the USB device, I only have cheap devices: here comes a Behringer
U-Control model UCA202. Not really Hi-Fi, yet convenient. Plug it, no
driver to bother with, it works flawlessly and immediately. 

I used the same WAV as for testing jitter, because it's very easy to
check the actual frequency of the Fs/4 signal. It's 11025 Hz when the
sample rate is 44100 Hz.

With 45875 Hz, it should be 11468.75 Hz.
I measured approximately 11476 Hz, corresponding to a sample rate of
45904 Hz.

So, this is a confirmation that, for the clocking purpose, this class
compliant USB relies completely upon an information provided by the
hosting PC, enbedded in the data flow.

More, the PC is not able to provide the exact sample rate, it certainly
allows only some discrete frequencies, and picks one it can elaborate,
as close as possible to the requested value. 
To confirm this, I did another edition: this time with ABCD (simple
enough, uh?), 43981 Hz, quite close to the regular 44100, and this
time, I got 44100 at the output.

The implication is impotant: it means that instead of a stable, high
precision, low jitter clock, we have something derived from software
and/or a clock generator whose specifications are certainly not
inspired by audiophile considerations.

So the signal delivered to the USB device is prone to be affected with
a potentially important jitter.
Assuming that a total immunity of DACs to jitter is questionable, I
feel it's not a good idea to bet on USB. At least this kind of USB. 

JLM


-- 
jlmatrat
------------------------------------------------------------------------
jlmatrat's Profile: http://forums.slimdevices.com/member.php?userid=10656
View this thread: http://forums.slimdevices.com/showthread.php?t=34132

_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/audiophiles

Reply via email to