<div>I've been playing with this as well. My laptop speakers are quite "tinny" sounding and I get very different real-world results with this laptop than with desktop speakers. Works great with decent speakers and a mic, works terrible with my laptop (Lenovo Thinkpad SL400) built-in mic and speakers. With my Lenovo's built in speakers and mic I eventually get a high pitched "woot woot" creeping in (probably creeps in and out due to auto gain settings I am using?) I'd like to add to these questions if I may:</div>
<div><br></div><div>3. Can choosing a "bad" sample size introduce artifacts due to windowing problems with the Fast Fourier Transform required to do echo cancellation? I mean, if I choose a sample size of 161 samples instead of say 1024, will that affect the echo cancellation in some way.</div>
<div><br><br><div class="gmail_quote">On Wed, Aug 12, 2009 at 4:46 PM, Greger Burman <span dir="ltr"><<a href="mailto:greger@mobile-robotics.com">greger@mobile-robotics.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div>First of all, thank you for your input Tim. That is very helpful.</div><div>I would love to hear from other people with experience of AEC and Speex.</div><div>I guess I have to split my question into to parts now.</div>
<div></div><div>1. </div><div>Is it a fact that using the windows multimedia API (wave audio) for audio capture and playback makes it impossible to do echo cancellation with Speex AEC or other EC method due to inprecise timing?</div>
<div></div><div>I welcome comments from anyone who has got echo cancellation working when recording through WinMM, DirectSound or ASIO.</div><div></div><div>2.</div><div>What about the Speex echo cancellor and the different sampling rates. Will it only perform well in narrowband mode (8kHz)? Does it not perform well in wideband (16kHz) or ultra-wideband (32kHz) mode? </div>
<div></div><div>I read somewhere that you need a longer tail length for higher sample rates. Measured in samples it will of course be longer, but it could be interpreted as longer measured in milliseconds for higher sample rates. Someone please clarify. </div>
<div></div><div>//Greger</div><div></div>
<br>_______________________________________________<br>
Speex-dev mailing list<br>
<a href="mailto:Speex-dev@xiph.org">Speex-dev@xiph.org</a><br>
<a href="http://lists.xiph.org/mailman/listinfo/speex-dev" target="_blank">http://lists.xiph.org/mailman/listinfo/speex-dev</a><br>
<br></blockquote></div><br></div>