LiMaoquan2000
2011-Apr-12 06:54 UTC
[Speex-dev] Why most AC97 soundcard has different sample rates of of capturing and rendering?
Hi all, I remember someone (sorry I forgot his name) told me that AC97 capturing codec has its own PLL clock generator to fit for S/PDIF input which has its own clock. So the mismatch between clocks of capturing and rendering is unavoidable. But after reading the Audio Codec 97 Revision 2.3 Revision 1.0 document carefully, I can only find proofs that the mismatch should not exist. In chapter 3.4, page 23, clocking for Multiple Codec Implementations, it is said that "To keep the system synchronous, all Primary and Secondary Codec clocking must be derived from the same clock source, so they are operating on the same time base. In addition, all AC-link protocol timing must be based on the BIT_CLK signal, to ensure that everything on the AC-link will be synchronous.". In Figure 8. Controller to Multiple Codec connections at page 21, you can also find that AC97 controller and all Codecs share the same SYNC and BIT_CLK signal. So anyone could give me more details for the reason of mismatch between sample rates of of capturing and rendering? Which is a HUGE OBSTACLE for all LMS based acoustic echo cancellers, such as speex AEC. Maoquan -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/speex-dev/attachments/20110412/44fd668b/attachment.htm
Apparently Analagous Threads
- Anyone knows how microsoft AEC can deal with mismatches between clocks of capture and render streams?
- Anyone knows how microsoft AEC can deal with mismatches between clocks of capture and render streams?
- problems with alsa (card ac97) in asterisk
- Anyone knows how microsoft AEC can deal with mismatches between clocks of capture and render streams?
- Anyone knows how microsoft AEC can deal with mismatches between clocks of capture and render streams?