Re: [LAU] Sample rates

From: Ralf Mardorf <ralf.mardorf@email-addr-hidden-dsl.net>
Date: Fri Feb 15 2013 - 22:04:32 EET

On Fri, 15 Feb 2013 18:30:15 +0100, Bob van der Poel <bob@mellowood.ca>
wrote:
> On Thu, Feb 14, 2013 at 9:03 PM, Ralf Mardorf
> <ralf.mardorf@alice-dsl.net> wrote:
>> On Thu, 2013-02-14 at 20:32 -0700, Bob van der Poel wrote:
>>> Now, if I have an app like audacity set to 48K does it (or jack?) send
>>> a message to the box to enable 48 (or 44.1)? If that is the case, I
>>> assume that using 44.1 or 48 should have the identical latency issues?
>>> 48K should send a bit more data back to the computer?
>>
>> All audio devices I know automatically are set to the sample rate set by
>> alsa, 'jackd -d alsa --help'.
>>
>> No, the latency will differ for different sample rates:
>>
>> 512 samples / 44.1 kHz = 11.609977324 ms
>> 11.609977324 * 2 = 23.219954649 ms ≈ 23.2 ms
>>
>> 512 samples / 48.0 kHz = 10.666666667 ms
>> 10.666666667 * 2 = 21.333333334 ms ≈ 21.3 ms
>
> Is this saying that a higher freq will result in lower latency? Guess
> that makes sense since higher freq contains more data.

Yes, equal settings for frames/periode and periods/buffer will result in a
lower latency at a higher sample rate.

>> Higher sample rate = more data at equal bit depth.
>
> Guess there really is no reason not to use 48K then.

You should prefer 48 KHz, that it does take more data shouldn't matter.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
Received on Sat Feb 16 00:15:02 2013

This archive was generated by hypermail 2.1.8 : Sat Feb 16 2013 - 00:15:03 EET