Hi all,
I'm currently testing out the speex echo canceller to determine if it is
suitable for use in a client's product. I have a few questions regarding
problems that I'm having and I would appreciate it if anyone could help
me out.
1. I provide to .wav files as arguments to testecho and the output
generated always seems to be 44100 Hz even though the sampling rate in
testecho.c is set to 8000 Hz. Why is this? How can I change this so that
the output file is 8000 Hz?
2. The two input files I specified are .wav and the output is .sw,
is this fine? Or do the input files also need to be .sw?
3. When I test the echo canceller on slow, non-continuous speech
(i.e. counting from 1 to 10), the echo is removed and the output sounds
great. However, if I try to remove the echo from an audio file
containing continuous speech (i.e. reading a paragraph of text), the
output is distorted and the speech can barely be heard. Any ideas on how
this can be fixed?
I'm using the speex-1.2rc1 source.
Thanks,
Mani Ramezani
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/speex-dev/attachments/20081028/4893411a/attachment.htm
