similar to: patch for libspeex

Hi, ? Is there a fix for this issue??? ---> http://permalink.gmane.org/gmane.comp.audio.compression.speex.devel/2959 ? I am seeing the same thing when I compile speex in visualdsp ? These are the errors I get from using the assembly version of vq_nbest: ? ..\..\..\..\algorithms\voice\speex\src\vq.c [Error ea5004] "C:\Users\coder\AppData\Local\Temp\acc22e8547f000\acc22e8547f001.s":482

On Sat, Dec 14, 2002 at 01:46:19AM -0500, Jean-Marc Valin wrote: > Thanks for the patch. I applied it and it give me up to 15% in speed. > Doesn't seem to change the results, which is a good thing (though you > originally forgot a "used=0" in vq_nbest_sign). I'll check a thing or > two and I'll apply to CVS. D'oh. My carelessness, sorry! :) > Strange...

Le sam 14/12/2002 à 01:03, Bernard Blackham a écrit : > I have a patch for libspeex, which optimises some of the loops in > vq_nbest and vq_nbest_sign that speeds up encoding - my results: Thanks for the patch. I applied it and it give me up to 15% in speed. Doesn't seem to change the results, which is a good thing (though you originally forgot a "used=0" in vq_nbest_sign).

Most asm statements clobber ASTAT bits (shifts, maxes, etc...) but do declare the register as clobbered. Same thing with CC in a few places. Some places make an attempt at clobbering some hardware loop registers, but it's very incomplete compared with how many asm statements actually use hardware loops. Signed-off-by: Mike Frysinger <vapier at gentoo.org> --- libspeex/bfin.h

On Sat, Dec 14, 2002 at 06:04:08PM -0500, Jean-Marc Valin wrote: > Maybe not everything got re-compiled and some parts were left with -g or > something? With the files you sent, I get (on my PIII 1 GHz, with -O3): > speexenc --comp 3 --quality 3 sample.wav sample.spx > 0,65s user 0,00s system 100% cpu 0,647 total > > strange... I'm recompiling the lot, no debugging, -O3 (or

Hi there, here is a little bug report: # ./configure --with-gnu-ld --enable-sse # make [...] vq.c:99: error: conflicting types for `vq_nbest' vq.h:44: error: previous declaration of `vq_nbest' vq.c:133: error: conflicting types for `vq_nbest_sign' vq.h:46: error: previous declaration of `vq_nbest_sign' The --enable-sse option took this bug to the surface. The header file is

Hi all! Speex can encode voice in eight different bit-rates that can be invoked by setting the "Quality" parameter. Another parameter that is of importance to what I am trying to achieve is "Complexity". My questions deal with "Complexity". I know that the CPU requirements for complexity 10 is about 5 times higher than for complexity 1. I have verified it while

I would like to use speex_preprocess() for voice activity detection. I read the comments in speex_preprocess.h and the relevant paragraphs in the Speex manual. Is there anywhere I can find more information on how to use this function? Thanks for your help, Steve _________________________________________________________________ Express yourself instantly with MSN Messenger! Download today

> -----Message d'origine----- > De : Jean-Marc Valin [mailto:jean-marc.valin at usherbrooke.ca] > Envoy? : dimanche, 14. juin 2009 20:46 > ? : Stephane Lesage > Cc : speex-dev at xiph.org > Objet : Re: [Speex-dev] Resampler saturation > > Just to make sure I understand, the two patches you sent are > two different ways to fix the problem, with the only >

On Wed, Dec 18, 2002 at 04:55:49PM +1100, Conrad Parker wrote: > for a good time read: > http://gcc.gnu.org/onlinedocs/gcc/Inline.html *nod* makes it all that much clearer... tya! > > Australian Linux Technical Conference 2003: http://www.linux.conf.au/ > > oath, see you there :) Where speex will be used to encode/compress the presentation audio streams, and sweep will

I'm trying to get speex to encode a bit faster, mainly by rewriting a few functions in SSE and translating the GCC __asm__ to VC __asm. There's 2 functions I'm targeting, first is vq_nbest which consumes 40% of the time at high complexity and split_cb_search_shape_sign. Which consumes just over 30%. I've split out two functions from: cb_search_precompute_energy - loop at the

Hi Jens, > There is already streaming support in the xmms-plugin. > The 0.6.0 version found at http://jzb.rapanden.dk/speex/ is hopelessly > outdated, but the nightly version should work. D'oh, should've looked at this before hand. Seems you've done more or less what I've done and taken the streaming code from the mpeg/vorbis plugins and meshed it in with yours. I

On Sun, Dec 15, 2002 at 02:15:51PM -0500, Jean-Marc Valin wrote: > I thought that -O3 would inline those functions, but I'll have a closer > look... I'm not sure how C inlining works. Does the inlined function > need to be in a .h? I believe either the .h or the .c or both (both for consistency). It does seem an obvious optimisation for gcc to do though. Maybe because its in a

Hi, Jean-Marc Valin wrote: >> 1. What is the point in multiplying a codebook index with some number >> ant adding a loop variable to it as done in the exc unquant function. >> for (j=0;j<subvect_size;j++) >> exc[subvect_size*i+j]+=s*0.03125*shape_cb[ind[i]*subvect_size+j]; > > That's just how you represent a 2D array in C: the codebook has

Hello, Is it possible to adjust the level of the denoiser ? In an old beta, before you change the aggressiveness of the denoiser, it works very well with the noise of a car, a road etc... but not now. The AEC too works very well in a old beta, but now, I think there is a problem... I have tested it with the same sound card (WB 16000). In any case, your project is the best VOIP library in this

On Tue, Dec 17, 2002 at 11:55:21PM -0800, Greg Herlein wrote: > If such a thing happens, discussion of the RTP profile draft > would be most welcome - please get responses back to the > list! Now, if this were finalised before the conference then we could do a demo and use it for broadcasting the lectures streams around the world... What is currently the best way of doing this? I'm

How about this: the codebook search gives you the N best (in order) combinations of entries for the current sub-frame. You can simply compute a one-bit checksum on all the codebook entries and choose which of the N-best to use based on that and the "message" you have. Of course, there will be cases where all of the N-best match to the same checksum, but that's what error-correcting

> Using zsh's time function (gives real and user time - which are > pretty similar on this unloaded machine). I've put my sample wave > file at http://dagobah.ucc.asn.au/speextest/sample.wav if anybody > wants to compare times. > > I'm curious now why my machine is slower - perhaps it's something > about the way I've compiled it. (Compiles by default with

Hi, I am using speex 1.2beta2 on a narrowband 16-bit, 8khz system that has a severe program space problem and will not fit speex in its normal operation. In an attempt to shrink speex I placed a breakpoint in every function and ran a decode and encode and removed the breakpoints that I hit. in the functions that had a breakpoint that I didn't hit I commented out those functions (as well as

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