Hi,
I was wondering if it is possible to convert mel-frequency cepstral coeff to sound? I am sure, if it is possible, a lot of information will be lost. But i am perplexed by the thought that after juicing out the information out of a sound wave, is it possible to get it back and if yes to what extent.
If you could direct me to any resources which explain this inverse conversion from mfcc to sound waves, I'd be grateful. I have spent quite some time searching and I end up with a deluge of corpora on wav to mfcc.
Thanks in advance
Aneesh
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks Nikolay,
I will look into sptk, I briefly went through the paper you suggested, seems interesting.
I have found more information as well... I will add in this msg, when I have read them properly.
btw, I really appreciate quick response.
Regards
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I was wondering if it is possible to convert mel-frequency cepstral coeff to sound? I am sure, if it is possible, a lot of information will be lost. But i am perplexed by the thought that after juicing out the information out of a sound wave, is it possible to get it back and if yes to what extent.
If you could direct me to any resources which explain this inverse conversion from mfcc to sound waves, I'd be grateful. I have spent quite some time searching and I end up with a deluge of corpora on wav to mfcc.
Thanks in advance
Aneesh
Thanks Nikolay,
I will look into sptk, I briefly went through the paper you suggested, seems interesting.
I have found more information as well... I will add in this msg, when I have read them properly.
btw, I really appreciate quick response.
Regards
MLSA filter does this
http://www.sp.nitech.ac.jp/~tokuda/selected_pub/pdf/conference/tokuda_icassp1998.pdf
the sources are available in sptk.