The current code does not specify the metrics of the buffers for the
input device. This makes PulseAudio choose very bad defaults, which
causes input to be unusable: Audio put in gets out 30 seconds later.
This patch fixes that and makes the latency configurable as well.
Signed-off-by: Martin Schrodt <martin@schrodt.org>
Message-id:
20190315084653.120020-4-martin@schrodt.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
{
int error;
pa_sample_spec ss;
+ pa_buffer_attr ba;
struct audsettings obt_as = *as;
PAVoiceIn *pa = (PAVoiceIn *) hw;
paaudio *g = pa->g = drv_opaque;
ss.channels = as->nchannels;
ss.rate = as->freq;
+ ba.fragsize = pa_usec_to_bytes(ppdo->latency, &ss);
+ ba.maxlength = -1;
+ ba.minreq = -1;
+ ba.prebuf = -1;
+
obt_as.fmt = pa_to_audfmt (ss.format, &obt_as.endianness);
pa->stream = qpa_simple_new (
ppdo->has_name ? ppdo->name : NULL,
&ss,
NULL, /* channel map */
- NULL, /* buffering attributes */
+ &ba, /* buffering attributes */
&error
);
if (!pa->stream) {