It's an improvement. On actual stations rather than this sample, perhaps
speex_preprocess_estimate_update() could be called when the squelch is
not open, so that its estimation is up to date when transmit begins. I
can hear the denoiser cutting in after about half a second in the sample
you sent.
Can we use the voice activity detector as a VOX?
You might also try the dereverb on some samples, it's easy to get a wall
bounce in a home station.
Thanks
Bruce
On 06/21/2014 08:38 PM, David Rowe wrote:
Hi,
I've coded up a small test program to test the Speex noise suppressor on
(acoustically) noisy input speech.
Here is a sample of the 1300 bit/s coded speech quality with/without the
the noise suppressor applied to the input speech before encoding.
The samples were produced with command line magic like:
./unittest/speexnoisesup ../raw/mmt1.raw - | ./src/c2enc 1300 - -
| ./src/c2dec 1300 - - | play -t raw -r 8000 -s -2 -
Cheers,
David
------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
