10 Apr
2013
10 Apr
'13
4:29 p.m.
return (nsec > codec_nsecs) ? nsec - codec_nsecs : 0;
Is this correct? I would think for capture you need to add the delay, and subtract it for playback?
Should the capture timestamp represents the wall time when the sample is converted by the A-to-D? If so, would be the correct time be the codec delay nsec before what is reported here?
For playback audio_timestamp = wallclk - codec_delay_in_nsec for capture audio_timestamp = wallclk + codec_delay_in_nsec