[LAU] Sample rate vs. SNR

Joe Hartley jh at brainiac.com
Wed Jan 23 20:20:47 UTC 2013

On Wed, 23 Jan 2013 17:43:25 +0100
"Ralf Mardorf" <ralf.mardorf at alice-dsl.net> wrote:

> IIRC it was Fons who some times ago has written that "most devices perform  
> better at 48 KHz". This is what I experienced too, while I indeed own two  
> devices that did sound better at 96 KHz on my computer.

I don't have the original thread any more, but I'm pretty sure it was Fons
as well who explained why this mattered.

It has a lot to do with the native rate of the device and the math involved
at having to achieve the non-native rate.  Some ADCs (like the Delta 1010)
allow you to set their native rate, so running my 1010s at 44.1k was no
problem.  I recently got an Alesis MultiMix16, which gives me 18 channels
of input to the computer via firewire.  I found that if I ran JACK at
anything other than the Alesis' native 48k, performance was awful.  Running
at 48k has been smooth sailing though.

       Joe Hartley - UNIX/network Consultant - jh at brainiac.com
 Without deviation from the norm, "progress" is not possible. - FZappa

More information about the Linux-audio-user mailing list