similar to: Speex newbie questions

Hello kind developers, I am planning on creating a steganographic extension for speex/CELP. Since speex/CELP is lossy, there should be quite many bits that one can use for hiding data. I'm familiar with the principles of CELP (two filter loops, one for pitch, the other for formants, let the raw speech frame run through those filter loops, take residue signal, map it on a codebook entry,

Hi, I chose speex initially because i had some work in VQ on speex i.e. modifying split VQ to GMM based parametric VQ and I thought If I train the GMM based VQ codebooks with audio signal and then do audio coding with speex, I probably get a better(smaller) residual signal even with speex. But I couldnt get that. I was trying to get a lossless bitstream by MUXing the speex-bitstream and the

Hi, I have some questions regarding Speex. please reply as soon as possible. where are the two filter loops applied to the raw signal? where is the residue (innovation) signal calculated? where is the innovation mapped on a codebook entry (key)? which bits of a transmitted frame are the two filters, which are codebook keys? Is the codebook from the beginning on static (so already known by

Hi, We would like to announce the first release of the Speex project. Speex (http://speex.sourceforge.net) is an open-source (LGPL), patent-free compression format allowing an alternative to expensive proprietary codecs. Unlike Ogg Vorbis which compresses general audio, Speex is designed especially for speech. For that reason, Speex is meant to be a complement to Vorbis. Since it is specialized

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

Hello, Jean-Marc Valin schrieb: >>I am planning on creating a steganographic extension for speex/CELP. >>Since speex/CELP is lossy, there should be quite many bits that one >>can use for hiding data. > > Quite the opposite. Speex uses very few bits, so every bit you corrupt > is likely to be heard. You would need to use "fractions of bits" at a > time.

Hello Jean-Marc and all, I recently had a talk with somebody about CELP. He said, there is this federal standard 1016 (4.8kbps) with a reference implementation of the Department of Defense (only on Sun, unfortunately, if I got this right). This one is noticed in the manual already. He also said, since there is this implementation of the DoD, nobody would voluntarily re-implement CELP. If I read

Hi people Just wondered what your thoughts are on Microsoft's RTAudio codec that they use in their Unified Communications stuff: http://www.microsoft.com/downloads/details.aspx?familyid=5D79B584-79C9-42A8-90C4-4AB3F03D19C4&displaylang=en They let people license it to use in their own products. I couldn't find any sample clips of RTAudio encoded speech. Apart from being $35000 less

You're asking the wrong question. The question is not "why does it would bad with Speex?", but "why does it sound good with LPC10 and MELP?". And the answer is that both are vocoders. Try dropping frames/subframes with anything else (Vorbis, MP3, G.729, u-law, ...) and it'll sound terrible as well. The only reason it sounds good with vocoders is because the

I am having a problem with "record_file" working properly depending on when it is called -- basically if it is called immediately upon a call, it acts like it does not hear anything from the callers phone (yes, my phone is setup properly and functions fine otherwise)... if I do a "background" or "festival" command before calling it, it works fine. Details below:

> I am planning on creating a steganographic extension for speex/CELP. > Since speex/CELP is lossy, there should be quite many bits that one > can use for hiding data. Quite the opposite. Speex uses very few bits, so every bit you corrupt is likely to be heard. You would need to use "fractions of bits" at a time. Even then, it may be hard to hide data efficiently (so it's

Hello I'm working about audio and video codecs. I'm making a research about the state-of-the-art in coding technologies. My question is about the speex codec 1.0. I would like to know the name of the algorithms that it uses.(LPC,CELP..). It uses more or less the same algorithm for narrowband and wideband? Finally i would like to ask about future codec developement projects. Thanks

Hi Jean-Marc, I thought at quality 3 (wideband) - wb_submode1 that the 4-8k band was not using a codebook table. From the code I can see that some sort of "lsp" encoding is performed. What exactly does this encode? (I assume lsp means line-spectral pairs) The reason I am asking is I'm comparing the "effective" spectral bandwidth of Speex against the

-----Original Message----- From: Jean-Marc Valin [mailto:jean-marc.valin@usherbrooke.ca] Sent: Tuesday, June 19, 2007 6:38 PM To: Michael Shatz Cc: speex-dev@xiph.org Subject: Re: [Speex-dev] Blackfin inline assembler and VisualDSP++ toolchain >> Yes, data footprint in the new version is quite manageable. Still I would >> wish better documentation for speex_alloc_scratch(). >

Hi, Jean-Marc Valin wrote: > Not sure I understand your question. Change the order of what within what? In cb_search, we iterate nb_subvect times over a codebook, finding nb_subvect codebook entries to quantize the excitation signal. After finding these nb_subvect codebook entries, they're written into the stream: /*save indices*/ for (i=0;i<nb_subvect;i++) {

hi, thanks for answers to my previous qns. have some more of them. hope it's ok to ask questions on the basics of CELP here. let me know if it's not. 1) synthesized filter stability after interpolation of LSFs I read from some resources that if LSF representation of LPCs is used for interpolation, stability of the synthesis filter is guaranteed. from another source, i read that the

Hi, Jean-Marc Valin wrote: > Well, you could change the order in the encoder as long as you reverse > it in the decoder as well. Ok, I see that in the split_cb_shape_sign_unquant function, that each coefficient is tied to it's position in the nb_subvect exc coefficients. Honestly, I have problems understanding what exactly the codebook search works like. If you have the time to to

Hi, From the vorbis decoder specification, it is clear that the decoder needs to have all the codebooks before decoding can actually begin. I will appreciate if someone can clear the following questions: 1. I guess the codebooks are derived from the actual input data. Probably the encoder may be making two passes through the input. The first pass finds out the frequency of different symbols

Hi all, I'm porting the Tremor code to an ARM platform. Unfortunately, we didn't use the low mem branch code for CPU performance consideration, and now I find the memory use in it is huge for the embedded system. After looking into the code, I found the header parse, especially, the codebook unpack part takes a lot of memory. Tremor decodes the codebooks to a cache to store

Hello, I'm trying to write a Vorbis decoder (just as a hobby) but I've got a problem with reading the setup header. According to the Vorbis documentation the first data that should be read after the initial 'vorbis' string is an eight bit integer which when incremented by one, contains the number of codebooks in the header. My problem is that whenever I read this number it seems