similar to: memory usage

hi, jean-marc: i think we can remove spx_sig_t *orig. but am not sure about exc2Buf. is it for extension? rgds, tk On Mon, 28 Feb 2005 12:42:38 -0500, Jean-Marc Valin <Jean-Marc.Valin@usherbrooke.ca> wrote: > Hi, > > I looked at the code I think there are still places where you can reduce > memory. For example, I think bufSize can be reduced to around 400 > (instead of

Hi I am currently trying to port speex v1.1.6 to a microcontroller with very limited memory (<64Kbyte RAM). what I found when initialising the encoder, a chunk of 32Kb was attempted to be alloced, which failed: src/nb_celp.c: void *nb_encoder_init(const SpeexMode *m) { /* snip */ st = (EncState*)speex_alloc(sizeof(EncState)+8000*sizeof(spx_sig_t)); /* snip */ } same goes for the

Hi, I looked at the code I think there are still places where you can reduce memory. For example, I think bufSize can be reduced to around 400 (instead of 640). Also, in most cases, the stack of the encoder and decoder are not used at the same time. If you use complexity 1, you can save a lot on the stack too. Jean-Marc Le lundi 28 f?vrier 2005 ? 10:59 +0800, tunkeat a ?crit : > hi Alfred,

I noticed that in the narrow band and wide band destroy functions only the main pointer is being freed. I think that it should be: void nb_decoder_destroy(void *state) { DecState *st; st=(DecState*)state; speex_free (st->inBuf); speex_free (st->excBuf); speex_free (st->innov); speex_free (st->interp_qlpc); speex_free (st->qlsp); speex_free

When updating the speex sources from svn tree, I found that the following revision has corrupted the deallocation (segmentation fault): ------------------------------------------------------------------------ r9320 | jm | 2005-05-27 15:05:05 -0300 (Fri, 27 May 2005) | 2 lines Proper de-allocation When compiling with the 9316, everything works fine. but when I update with later

On Mon, 2005-02-28 at 19:42 -0500, Jean-Marc Valin wrote: > > jean-marc: i think we can remove spx_sig_t *orig. > > but am not sure about exc2Buf. is it for extension? > > orig is already removed in SVN (which you should probably use). As for > exc2, it can be removed, but I'm not sure if you can just use exc > instead (maybe yes). > when removing "spx_sig_t

To make it complete, here is the code that I am using to encode a large file: int __stdcall SpxEncode(unsigned char* inBuf, unsigned char* outBuf, unsigned int inlen) { //char *testFile; //FILE *ftest; //testFile = "test"; //ftest = fopen(testFile, "wb"); //fwrite(inBuf,1,inlen,ftest); //take every 320 bytes //copy every short to float

hi all, the stacksize required for encoder and decoder are 32Kbytes and 16 Kbyes respectively. i changed it to 20Kbytess and 8Kbytes and it seems to work. but need to trim further. can someone point me in the right direction to further reducing the stack size? thanks! cheers, tk

Dear all, I have one simple question. I understand that speex_encode and speex_decode takes float * as an arguement to encode and decode the sound. However, when I get the PCM data from the sound card under win32, it is a just 16 bit array. May I know how do I convert this 16 bit value to speex float format and to convert back? Is there got any routine to do this? YueWeng

Hi Jean-Marc, Hi Jim, I have also seen some problems with the 1.1.8 release on the C55x. So far I have boiled down the issues to the following: 1) We need our own "fixed_xx.h" header file. I don't know why, and haven't had time to investigate, but there is a definite improvement when I use the attached fixed_c55x.h file which has turned all the maths into inline functions.

Hi, So 9316 works and 9320 doesn't? How about latest SVN. I just ran everything in valgrind and saw no error at all. Can you give more info on how to reproduce (with speexenc)? Jean-Marc Le mercredi 22 juin 2005 ? 21:19 -0300, Dario Andrade a ?crit : > > > When updating the speex sources from svn tree, I found that the > following revision has corrupted the deallocation

hi all, thanks Jean for the codec. I am new to speech coding. Am interested to implement speex on a TI dsp, say C54. just downloaded speex v 1.1.6. when I removed optimisation flags for win32 in libspeex.dsp, I get a slightly different output for decoding using speexdec.exe (debug version). no prob for speexenc.exe except for some differences in the ogg header. for anyone interested, i

Hi Alfred, First thing I'd like to ask is whether you're trying to achieve real-time performance, as I find it very unlikely that Speex could run in real-time on a micro-controller. Regarding bufSize, no you can't just make it the size you like. Now, it's possible to decrease the amount of memory allocated if e.g. you only use complexity 1. You can also assume that the stack is

Hi, We are currently running Dovecot v1.1.6 on our servers, and are contemplating an upgrade to 1.2, or 2.0 soon. We are wondering how many organizations are still running a 1.1 version of Dovecot, and if anyone has any thoughts on this transition. Have you upgraded from 1.1 to 1.2? What are your experiences with this? Have you upgraded from 1.1 to 2.0 directly? What are your experiences this

I'm confused about the format of the PCM/WAV data for speex_encode & speex_decode speex_encode(enc_state, input_frame, &bits); In the manual, it says "input_frame is a (float *) pointing to the beginning of a speech frame" (for encode) But I've found that in speexenc.c and testenc.c, short* is used instead of float* So, isn't it signed 16 bit samples(if

Hi, i know this question has already been asked, i've checked mailing list logs but ... I didn't figure out the solution for my question. I want to use basic functionnality of speex, only VAD, no pre/post filter, even no VBR. But i *really* want to use less mem/cpu i can. So in which format should i pass data to speex_encode(_int) , float or short ? I quickly looked at the source code

Hi Marian, I took a brief look at RFC 5574 and as far as I understand, you can simply do it like this: SpeexBits b; speex_encode(state, some320bytePCM1, &b); speex_encode(state, some320bytePCM2, &b); After that, get the encoded data with speex_bits_write and put it in your packet. Mark -----Original Message----- From: speex-dev-bounces at xiph.org [mailto:speex-dev-bounces at

Hi, I am writing a program that should be able to compile using speex version 1.0.1 and speex version 1.1.4, but a I am getting some trouble to encode with one version e decode using the other. The program is taking care of the change in both speex_encode and speex_decode. The sample being coded or decoded has changed from float to short. To test it, I have modified the sampleenc.c that

Ok, I'm using 1.1.4, so I should pass signed short(wav data) directly to speex_encode()? But the samplecode that comes with 1.1.4 shows me to put short in float array, and pass the float array to speex_encode() <p>-----Original Message----- From: owner-speex-dev@xiph.org [mailto:owner-speex-dev@xiph.org] On Behalf Of Jean-Marc Valin Sent: Saturday, April 17, 2004 6:25 AM To: speex

Hi there, I enable DTX and VAD in speex_encode to detect whether there is silence in voice data. But the same programme got different result on my computers. The function speex_encode always return 38 on one of my computers although it is silence. On all other computers the function speex_encode return 1 when it is silence. It means DTX and VAD doesn't work on particular OS or PC. Do anyone