Re: [proaudio] Real-time for audio

[Grant <emailgrant@xxxxxxxxx>]

> <SNIP>
>> It sounds like real-time lowers both latency and jitter.  Jitter
>> absolutely affects sound quality.  I wonder if jitter could be the
>> cause of all this.
>
> If a system is exhibiting severe jitter, possibly more typical in
> consumer level sound cards and assuming that's the sort of card you
> are using, then possibly the real-time kernel helps. I wouldn't argue
> that at all.

Do you have a good concept of how latency and jitter are related?  I'd
like to know more about that.

> I'm using a studio quality external DAC (Benchmark) with a low jitter
> external studio clock source (Apogee) driving a good quality digital
> sound interface. (HDSP9652) I stand by my earlier statement that says
> when I use industrial quality components the real-time kernel makes no
> difference that I can hear. On the other hand I'm so old and dusty who
> knows what I can hear anymore? ;-)

I've been making a lot of low-latency tweaks to my system lately.  The
overall change in the sound is somewhat large and very noticeable.
The difference made by any one of those tweaks (such as jackd vs.
jackd -R) is small.

> As for the wave vs FLAC thing one must prove that we're talking apples
> to apples. Wave is lossless by definition. (As far as I know.) It's
> just 16 or 24-bit words bits packed into a file. FLAC may or may not
> be lossless depending on how it's encoded. If it's lossless, and if
> it's encoded from the exact same original data then the two formats
> should produce exactly the bit stream and if the bit streams are
> identical, and if they are delivered to exactly the same reproduction
> hardware at the same settings on the same day at the same time then
> they are going to produce the same sound.

But one could conceivably produce higher latency and/or jitter than
the other, no?  Also FLAC is lossless by definition.  That's what the
L stands for. :)

- Grant