similar to: SPEEX AEC makes voice volume very small.

Yes. I guessed that too, however I am not sure why it keeps repeating every time the user stops / pauses and starts speaking again in a single session. I am using a laptop with standalone speakers. For echo cancellation to work one has to make sure that the ref and echo buffers are synchronized. I guess this is the most common problem. -Elston -----Original Message----- From: Anton A.

I usually use the cross-correlation between the speaker signal and the mic signal. It works pretty well. The problem is the you need enough data in order to get a good S/N. I compute the cross-correlation between 1.5 second data of each signals. In this way the application detection of any anomalies is quite slow. I am looking for something faster, but I fear it will be something very application

2010/5/11 Elston Sa <jose at rebaca.com>: > Is there an API in speex or have you implemented your own? I don't know if there is an API in speex for this, probably not. There is a test for AEC which does something very similar. I implemented it starting from a FFT library. It is quite easy to do. > > I am currently trying to find one. But if you could point me to one it will

Hello. I have the following situation. AEC is used in network chat software over DirectSound API. Echo and reference signals are almost aligned (delay is no more than 30ms). When echo is emulated in notebook (built-in speakers + mic) everything goes fine and echo is cancelled. But when configuration includes stand-alone speakers and mic no echo is removed. Audio is in 22050 hz at 16 bit

1) Everytime a participant speaks there is a echo for a short duration (maybe a word or two) but as the participant continues to speak without a any break the echo is 95% cancelled (i.e there is a feeble echo still present if observed very carefully). 2) The moment the participant stops / pauses speaking and start talking again, scenario 1 is repeated as if the echo state has been re-initialized

Hi Jean-Marc, thanks for the response. First, I will clarify again that floating-point solves this - so isn't that a bug in fixed-point? Also, I understand that algorithmically the AEC won't cancel echo properly on a non-linear signal, but why completely distort the output? If the echo just won't get cancelled it would be acceptable, but in the current state it disables the ability to

If more than one speaker receives the *same* signal, it doesn't matter the number of speakers. It only gets tricky when the speakers are playing slightly different signals (e.g. from a stereo song). Jean-Marc Quoting Greger Burman <greger at mobile-robotics.com>: > One thing I can think of is if you are using two or more speakers. If the > speakers are not at the exact same

Hi Guys, I am trying to incorporate the speex AEC into my RTP stack. Trying to diagnose why the echo is not being cancelled I enabled DUMP_ECHO_CANCEL_DATA as per the docs. The problem is that the dump files the AEC generates are of different size. As I understand this and looking at echo_diagnostic.m, the files are supposed to be of same size. Can someone give me a hint as to why this could

Sorry for not being accurate in the original question. My testing environment is set to 20ms frames (PCM 16 8 1 to be precise). I have verified that all the frames that I feed to the AEC have the fixed size of 320 bytes. So this is why I am presuming that the the dumped files should be of equal size. The output signal is good and not distorted. It has echo, though. In a try to diagnose that I

On 2010-03-16 14:22, Josh Gargus wrote: >> If more than one speaker receives the *same* signal, it doesn't >> matter the number of speakers. It only gets tricky when the >> speakers are playing slightly different signals (e.g. from a stereo >> song). > > Does "tricky" mean that the Speex AEC won't handle such situations > well? Or just that you

Hi all, I have a query. I developed NLMS acoustic echo cancellation algorithm for real time voip application and using general purpose processor. Used more filter length (4096/2048) for NLMS and faced CPU consumption.With less filter length unable to estimate the exact echo signal and the quality of the speech also getting degraded. Could any one please suggest me the filter length

Hi, Arun and me are working on a echo-canceler module for pulseaudio build around speex. When using the preprocessor's echo filter, specifically in combination with denoise and AGC, it's unclear whether these have to run before or after the AEC. Right now we've got one instance of speex_preprocessor doing AGC and denoise running before the echo canceler; and a second for echo

Hi, I've set up speex AEC in our application. The echo's seem to be canceling, but the captured voice is distorted somehow. It sounds to me like low and/or high frequencies are removed. This happens even if I set my playback(echo) data to only zero's. Is this "normal" for the AEC? As far as I understand, if I send 0's as playback/echo data, the resulting data should be

Hi I am using speex 1.2-beta-3. I got AEC working on windows CE platform by synchronizing MIC and speaker frames. The AEC is working perfectly fine for default volume but fails miserably for max volume level. Are there any more changes i need to do if we increase volume of device? -- SHAHEER AHMED - -------------- next part -------------- An HTML attachment was scrubbed... URL:

Hello, We've integrated the speex echo canceller into PulseAudio, and coupled with the AGC/denoiser, it works reasonably well and has been a breeze to plug in, so thank you! One thing we're seeing, though, is that the canceller seems to take some time (a few seconds) to "learn", so initially the echo is clearly audible and it slowly fades out. This sometimes occurs in the middle

Hi all, I am attempting to test AEC behavior at various sample rates. I ran a little experiment: I recorded a 10 seconds voice clip and the resampled at 8000, 11025, 16000, 22050, 24000, 32000, 44100 and 48000. I have a small applications that plays a wave file, records whatever comes in from the microphone and applies the Speex AEC and preprocessor on the input. It then saves the raw

Hi Shridhar, Sample rate conversion is not enough to solve this problem. I have tried this method several months ago. The first step is to measure the difference between sample rate of capturing and rendering. Then resampling (by what you said "sinc interpolation") one signal to eliminate the difference. The frequency step in my experiment is less than 0.1Hz. I have tried speex AEC

Hi all, We all know that mismatch between clocks of ADCs of far-end voice and near-end voice is not allowed in a time-domain or frequency-domain LMS based AEC system. It means that capture and render audio streams must be synchronized to a same sample rate. However, I found that this restriction is removed in microsoft AEC from Windows XP SP1. Anyone knows how microsoft AEC do it? This technology

Hi, is it somehow possible to use the internal data of the echo canceller to get some kind of a double talk detector? This would allow to perform echo suppression in scenarios where the cancellation fails, like with webcams, where AEC doesn't work at all. Best regards Andreas

Hi, I couldn't find a discussion that specifically addresses this, so here it is. I'm using Speex AEC in my mobile VoIP application to cancel speaker echo. The used version is 1.2rc1 from the website, and I'm compiling with fixed-point. On most occasions, the AEC works very well and cancels most of the echo (combined with the preprocessor). On some devices, where the microphone signal