Hi, I was just trying to use speex for sampling frequency >48KHz. In the original Speex-1.0.5 its restricted only upto 48KHz. I tired to modify it by changing the boundary conditions( the error conditions, i.e. if sampling freq >48KHz, it gives error) in /src/speexenc.c and then it atleast doesnt give the error, there is flow in decoding or encoding(i think). I suspect there are other constraints in the encoder and decoder to modify other than this to make it work properly. At this point, I am not getting the proper decoded(reconstructed) bitstream. Can you please suggest me the places where I should make changes to incorporate this requirement. thanks, Devilal *************************************************************************** Devi Lal Sharma Final Yr. Dual Degree(Communications) IIT MADRAS Mobile: +91-9986423985(I am currently in bangalore) email: devilal_sharma@yahoo.com, devilal@gmail.com *************************************************************************** --------------------------------- Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/speex-dev/attachments/20060326/6a281a6f/attachment.html
> I was just trying to use speex for sampling frequency >48KHz. In the > original Speex-1.0.5 its restricted only upto 48KHz. I tired to modify > it by changing the boundary conditions( the error conditions, i.e. if > sampling freq >48KHz, it gives error) in /src/speexenc.c and then it > atleast doesnt give the error, there is flow in decoding or encoding(i > think). > I suspect there are other constraints in the encoder and decoder to > modify other than this to make it work properly. > At this point, I am not getting the proper decoded(reconstructed) > bitstream. > Can you please suggest me the places where I should make changes to > incorporate this requirement.OK, I'm not sure what exactly you're trying to achieve, but I can't see any sane reason to use Speex at > 48 kHz sampling. I already think 44.1 kHz is a bad idea. If you want high-fidelity, use Vorbis, not Speex. If you want decent quality speech, use Speex at 16 kHz or something like that. Jean-Marc
Hi Jean-Marc Valin, Thank you for your suggestion. I am shifting to using Vorbis for Audio. I was using speex because I was using speex for previous work and it was easy to continue on speex than shifting to vorbis, not only that but speex is a waveform coder and I get quite good waveform matching when I reconstruct back the waveform using Speex even for music upto 44.1KHz. My motive is to get a lossless coder out of this(I am aware of FLAC). I want to compress and decompress the .wav file and then get the residual signal which I want to entropy encode using some program i have. I just want to use Vorbis(I was using speex till now) to get the smallest possible residual signal. The question is this that whether Vorbis will get me the waveform matching and the residual signal small enough? So please tell me how to go about it. I have downloaded libvorbis-1.1.2 and trying to compile it. Please tell me how should I use it to compress and decompress files for above use. Thanks, Regards, Devilal Jean-Marc Valin <jean-marc.valin@usherbrooke.ca> wrote: > I was just trying to use speex for sampling frequency >48KHz. In the> original Speex-1.0.5 its restricted only upto 48KHz. I tired to modify > it by changing the boundary conditions( the error conditions, i.e. if > sampling freq >48KHz, it gives error) in /src/speexenc.c and then it > atleast doesnt give the error, there is flow in decoding or encoding(i > think). > I suspect there are other constraints in the encoder and decoder to > modify other than this to make it work properly. > At this point, I am not getting the proper decoded(reconstructed) > bitstream. > Can you please suggest me the places where I should make changes to > incorporate this requirement.OK, I'm not sure what exactly you're trying to achieve, but I can't see any sane reason to use Speex at > 48 kHz sampling. I already think 44.1 kHz is a bad idea. If you want high-fidelity, use Vorbis, not Speex. If you want decent quality speech, use Speex at 16 kHz or something like that. Jean-Marc --------------------------------- Blab-away for as little as 1?/min. Make PC-to-Phone Calls using Yahoo! Messenger with Voice. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/speex-dev/attachments/20060327/51e6fc16/attachment.htm
Hi, I have one doubt again, that is Vorbis use DCT/MDCT based algorithm and also use psychoacoustic model so this is lossy codec. And I dont think it ca regenerate a better matching waveform than speex. Then there comes FLAC which is the perfect answer to my question, I suppose. But my concern is this that FLAC use simple prediction algorithm and doesnt use any CELP based algo which could have model the waveform coding by having a large codebook and comparing the residual signal and selecting the codebook index. For this, shall I start understanding and modifying FLAC itself in case I need to do something for lossless coding or I can try on Speex and than apply entropy coding. I am getting quite good(comparable) results for audio signal(44.1KHz) if use speex and separate entropy coding. Please suggest me clearly as I have very small time left to wrap up my work to submit. thanking you, Regards, Devilal Jean-Marc Valin <jean-marc.valin@usherbrooke.ca> wrote: > I was just trying to use speex for sampling frequency >48KHz. In the> original Speex-1.0.5 its restricted only upto 48KHz. I tired to modify > it by changing the boundary conditions( the error conditions, i.e. if > sampling freq >48KHz, it gives error) in /src/speexenc.c and then it > atleast doesnt give the error, there is flow in decoding or encoding(i > think). > I suspect there are other constraints in the encoder and decoder to > modify other than this to make it work properly. > At this point, I am not getting the proper decoded(reconstructed) > bitstream. > Can you please suggest me the places where I should make changes to > incorporate this requirement.OK, I'm not sure what exactly you're trying to achieve, but I can't see any sane reason to use Speex at > 48 kHz sampling. I already think 44.1 kHz is a bad idea. If you want high-fidelity, use Vorbis, not Speex. If you want decent quality speech, use Speex at 16 kHz or something like that. Jean-Marc --------------------------------- Yahoo! Messenger with Voice. Make PC-to-Phone Calls to the US (and 30+ countries) for 2?/min or less. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/speex-dev/attachments/20060327/951326cf/attachment.htm