<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><DIV>Hi </DIV>
<DIV> </DIV>
<DIV>I used this "sample counting " method to resample and put my audio signals in synch.</DIV>
<DIV>It worked perfectly in XP machines using a SoundMax audio card, but it failed in other XPs using Realtek cards. </DIV>
<DIV>As seen on <A href="http://lists.xiph.org/pipermail/speex-dev/2008-September/006889.html">http://lists.xiph.org/pipermail/speex-dev/2008-September/006889.html</A> my application continously checked my AEC level to slighly modify resample frequency, but convergence was impossible with that audio cards.</DIV>
<DIV>Then, I used ASIO (through port-audio library) in those machines, and AEC worked again. But this solution wasn´t practical (user must install a special driver) so i can´t say if this method will work or not on *every* machine. </DIV>
<DIV>I think it will not work on Vista, no mather what card you use.<BR></DIV>
<DIV>Just my 2c.</DIV>
<DIV><BR>--- El <B>mar, 7/7/09, John Ridges <I><jridges@masque.com></I></B> escribió:<BR></DIV>
<BLOCKQUOTE style="BORDER-LEFT: rgb(16,16,255) 2px solid; PADDING-LEFT: 5px; MARGIN-LEFT: 5px"><BR>De: John Ridges <jridges@masque.com><BR>Asunto: Re: [Speex-dev] AEC with different soundcards<BR>Para: "Alexander Chemeris" <Alexander.Chemeris@sipez.com><BR>CC: "speex-dev@xiph.org" <speex-dev@xiph.org><BR>Fecha: martes, 7 julio, 2009 9:36<BR><BR>
<DIV class=plainMail>Measuring clock drift is a thorny problem that depends greatly on your <BR>hardware setup and OS. In my case (Windows machines) I simply measure <BR>how many samples are played and captured in a given time period, and use <BR>that ratio to calculate the clock drift. Unfortunately these <BR>measurements are usually very noisy and require some high-powered <BR>filtering to get stable results that converge in a reasonable time. I'm <BR>afraid I'm not at liberty to get any more detailed than that. Sorry.<BR><BR>John Ridges<BR><BR><BR>Alexander Chemeris wrote:<BR>> AFAIK, that's a common point for all AECs. But some of them<BR>> solve the problem by resampling on of the end to keep it in sync<BR>> with the other.<BR>><BR>> On Tue, Jul 7, 2009 at 5:14 PM, ggb<<A href="http://es.mc242.mail.yahoo.com/mc/compose?to=ggb@tid.es" ymailto="mailto:ggb@tid.es">ggb@tid.es</A>> wrote:<BR>> <BR>>> Thank
you John.<BR>>><BR>>> On 07/06/2009 11:03 PM, John Ridges wrote:<BR>>><BR>>> ly synchronized, and therefore the clock drift adds a non-linear<BR>>> factor to the audio path. The AEC can only cancel linear changes to the<BR>>> audio path, and so the AEC never converges.<BR>>><BR>>> Is this a limitation common to all the AEC implementations/algorithms, or<BR>>> just the speex one?<BR>>><BR>>> One solution is to measure<BR>>> the clock drift and resample either the input or output signal so that<BR>>> they *are* synchronized, and then the AEC can converge. It's not fun,<BR>>> but it works.<BR>>><BR>>><BR>>> Could you detail me how to measure the clock drift? Any pointer to doc or<BR>>> example?<BR>>><BR>>> I apologize for my lack of knowledge in this topic,<BR>>> G.<BR>>><BR>>>
_______________________________________________<BR>>> Speex-dev mailing list<BR>>> <A href="http://es.mc242.mail.yahoo.com/mc/compose?to=Speex-dev@xiph.org" ymailto="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>>><BR>>> <BR>><BR>><BR>><BR>> <BR><BR>_______________________________________________<BR>Speex-dev mailing list<BR><A href="http://es.mc242.mail.yahoo.com/mc/compose?to=Speex-dev@xiph.org" ymailto="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></DIV></BLOCKQUOTE></td></tr></table><br>