[LAU] Tracking at 44.1 vs. 48k vs 96k?

[Brent Busby]

On Thu, 7 Feb 2013, Len Ovens wrote:

> On Thu, February 7, 2013 6:23 am, James Stone wrote:
>
>> Yep - thanks Len - that was certainly my experience with SBLive and
>> onboard sound.
>>
>> With the Focusrite Scarlett 2i4, it actually seems to run with 
>> (slightly) less CPU load at 44.1k than 48k (much less than 96k), and 
>> I think it might make 44.1k a better choice to avoid the extra 
>> downsampling etc. for the final mix in that instance. Not sure if it 
>> actually runs natively at 48k or whatever tho - not sure how I would 
>> find that out.
>
> The focusrite will clock the ADCs at whatever rate you select, it will 
> not resample. However, the low pass filter at 44.1k will have more 
> effect on the in band audio than at 48k. Do a listening test, if you 
> can't tell the difference with cymbals and such, I guess it doesn't 
> matter. If your mics roll off at 18k (or less) anyway it may not be an 
> issue. But honestly my engineering is not the best here.

That's the experience I've had here also with my RME Multiface II:

Despite that it's been regarded as a nice interface (and it is!), I've 
found that it's *much* darker sounding at 44.1/48kHz, and only seems to 
have full clarity in the treble at 96kHz.  Since I really doubt that my 
ears need the ultrasonic frequency response that desperately to get nice 
treble, I'd say it's probably the filters, owing the effect you're 
describing and others have mentioned in other threads in the past.

Some hardware isn't just 96kHz -- it's designed for it, and the filters 
won't let it sound good any other way.  If you have such an interface, 
the sample rate question is already answered for you.  You just have to 
listen to it at the slower sample rates and see if it still sounds good. 
If not, you may end up choosing 96kHz just because that's what your 
audio interface's filters like.

-- 
+ Brent A. Busby	 + "We've all heard that a million monkeys
+ Sr. UNIX Systems Admin +  banging on a million typewriters will
+ University of Chicago	 +  eventually reproduce the entire works of
+ James Franck Institute +  Shakespeare.  Now, thanks to the Internet,
+ Materials Research Ctr +  we know this is not true." -Robert Wilensky