The old pa_sink_set_fixed_latency() call didn't take into account that other places use pa_frame_align() on the pa_pipe_buf() result, so the configured latency could be sometimes slightly too high. Adding a buffer_size variable in userdata makes it a bit easier to keep all places that deal with the buffer size in sync. --- src/modules/module-pipe-sink.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/src/modules/module-pipe-sink.c b/src/modules/module-pipe-sink.c index 0eefdf1..01831bc 100644 --- a/src/modules/module-pipe-sink.c +++ b/src/modules/module-pipe-sink.c @@ -75,6 +75,7 @@ struct userdata { char *filename; int fd; + size_t buffer_size; pa_memchunk memchunk; @@ -123,7 +124,7 @@ static int process_render(struct userdata *u) { pa_assert(u); if (u->memchunk.length <= 0) - pa_sink_render(u->sink, pa_frame_align(pa_pipe_buf(u->fd), &u->sink->sample_spec), &u->memchunk); + pa_sink_render(u->sink, u->buffer_size, &u->memchunk); pa_assert(u->memchunk.length > 0); @@ -301,8 +302,10 @@ int pa__init(pa_module *m) { pa_sink_set_asyncmsgq(u->sink, u->thread_mq.inq); pa_sink_set_rtpoll(u->sink, u->rtpoll); - pa_sink_set_max_request(u->sink, pa_frame_align(pa_pipe_buf(u->fd), &u->sink->sample_spec)); - pa_sink_set_fixed_latency(u->sink, pa_bytes_to_usec(pa_pipe_buf(u->fd), &u->sink->sample_spec)); + + u->buffer_size = pa_frame_align(pa_pipe_buf(u->fd), &u->sink->sample_spec); + pa_sink_set_max_request(u->sink, u->buffer_size); + pa_sink_set_fixed_latency(u->sink, pa_bytes_to_usec(u->buffer_size, &u->sink->sample_spec)); u->rtpoll_item = pa_rtpoll_item_new(u->rtpoll, PA_RTPOLL_NEVER, 1); pollfd = pa_rtpoll_item_get_pollfd(u->rtpoll_item, NULL); -- 2.8.1