measuring audio latency?

[Dan Kegel]

On Wed, Mar 10, 2010 at 8:32 AM, Francois Gouget <fgouget at free.fr> wrote:
>> Yes, absolutely, but it'd be good to make this measurement easy
>> to repeat by anybody interested.  To do that, let's just loop the
>> audio output back into the audio input.  The user will have to
>> provide a loopback cable
>
> Won't the DirectSound capture latency come into play too?
>
> Unless we are able to put a precise timestamp on each capture sample
> independently of when it gets to the Winelib application it seems to me
> that all we can mesure is the latency of the sound output + capture
> loop.

That's right, but it's the best one can do without special hardware, I think.

> If the total latency is low then we're in great shape. If it's too high
> then we should remember this and try to figure out how much each side
> contributes to the latency.

Right, that's the point at which we reach for the special hardware
and/or more manual testing somehow.

But the simple, automated, loopback test of total latency is
definitely worth doing, IMHO.
- Dan