Hello,
Maarten Bosmans wrote:
> Please have another look at your patch. This is setting BLOCK_USEC to
> 10 seconds, not ms. Moreover, I don't think BLOCK_USEC is related in
> any way to the default adjust_time, so if changing BLOCK_USEC, better
> set it to some other hardcoded value, to avoid the suggestion that it
> has anything to do with the adjust_time.
Woops my bad, in my patch I just set it to 0, you're right that patch is wrong, my patch was to simply change
pa_sink_input_set_requested_latency(o->sink_input, BLOCK_USEC);
to:
pa_sink_input_set_requested_latency(o->sink_input, 0);
Which sets it to minimum latency possible, however I can imagine in some
cases you want to allow bigger latencies, so I doubt that is the
complete solution, the total minimum latency should probably be the
maximum of the minimum latency of each individual sink being used, while
the real latency set should be set to minimum tolerable, but that module
confuses me because it's both an output and an input, so I don't see how
it should be done.
Cheers,
Maarten
