similar to: Fixed point modules in floating point

On 11/03/2017 05:48 AM, Vega G wrote: > I was testing floating point  Opus encoder. But it's calling few > functions which are implemented in fixed point . I tried to find > corresponding floating point functions in source. But not available. > Please help me on this. The same code does both fixed-point and floating-point. If you define FIXED_POINT, then the macros and types

I've been browsing around the speex-1.10 source tree. Firstly I see 'float' is still used in many places. Can't we change them all to the appropriate type (spx_wordxx_t or similar) defined in arch.h ? Secondly I see there is a lot of duplication regarding real-valued constants. One way would be to say: #ifndef FIXED_POINT #define CONSTANT(x) (x) #else #define CONSTANT(x)

I've noticed that the nb decoder uses floating point in several places even when using FIXED_POINT. I don't have to deal with lost packets, so I'm mainly interested in innov_gain and pgain in no transmission mode (nb_celp.c around line 1272) and g and exci in vocoder mode (around line 1557). Is there a reason that these places must use floating point, or would it be possible to

Please compare samples: http://www.ivo.pl/ivo2/sound/krwawa_org.wav http://www.ivo.pl/ivo2/sound/krwawa_fixed.wav http://www.ivo.pl/ivo2/sound/krwawa_float.wav krwawa_org.wav is original file. Both output files was created using same program by compressing from wav to speex, then decompressing from speex to wav (using speex_encode_int and speex_decode_int). Only difference was used speex

I see there are still some floating point operations left in the codec init(ialization) code. Changing that code to fixed point is not only difficult (due to the trigonometric functions etc) but may also degrade the precision. Here is an idea whereby we can easily precompute (record) all those values on a powerful processor and then use (replay) them on an embedded processor / DSP. The only

Need to adjust scale for floating input. diff -NaurbB a/libspeex/resample.c b/libspeex/resample.c --- a/libspeex/resample.c 2013-12-20 17:40:27.326576921 +0500 +++ b/libspeex/resample.c 2013-12-20 17:40:16.918577087 +0500 @@ -915,7 +915,7 @@ if (in) { for(j=0;j<ichunk;++j) #ifdef FIXED_POINT - x[j+st->filt_len-1]=WORD2INT(in[j*istride_save]); +

Hello, I've just joined the list. I'm interested in helping on the fixed-point version of Speex. If there is anyone currently directing this work, would they please contact me and let me know how I can help. Thanks, -paul -- "Democracy is two wolves and a lamb voting on what to have for lunch. Liberty is a well-armed lamb contesting the vote ... If you make yourselves sheep,

I need to extract SILK only FIXED POINT code. I have a couple of questions in this regard. 1. Is it enough to enable compile time flag (FIXED_POINT) in the config.h, include silk_fixed library and exclude silk_float in the opus_demo project. I am working in the MSVC framework. Anyone has tried this before? 2. It seems there is no compile time flag to enable SILK only code, the core

Hi Drew, On 10/10/17 02:29 PM, Drew Allen wrote: > 2) Someone on this list might know better than I do on that one, but for > the new _ctl_va_list() calls, I believe the convention is for va_start() > and va_end() to appear in the caller rather than in in the va_list() > function itself. > > *My understanding is that it's impossible to pass ellipsis to another >

Hi Drew, Here's some comments on your patch (in no particular order): 1) I see that it's adding an #include of stdarg.h to opus_multistream.h Is that left over from the previous version? 2) Someone on this list might know better than I do on that one, but for the new _ctl_va_list() calls, I believe the convention is for va_start() and va_end() to appear in the caller rather than in in

Hi Jean-Marc, > Can you confirm I'm understanding everything correctly? You encode > with > the same encoder and then decode with either A) blackfin assembly and > fixed-point or B) fixed-point only on Blackfin. Then A) sounds bad and > B) sounds good. If you do the same in narrowband, it sounds OK. Is > that > correct? If that's the case, it's *probably* some

Hello Jean-Marc, > Could you try mixing the encoders and decoders (encode with one > version, > decode with the other) to see whether the problem is in the > encoder, the > decoder or both. I tested that,and the results are: The fixedpoint+blackfin encoder used together with the fixedpoint-only decoder produces an bad sounding file (for all the combinations for quality

Hello Daniele It would be worthwhile to attach an external serial flash or USB thumb drive, if the intent is store data. This allows for far more flexibility in storage Regards Amit On Fri, Aug 26, 2016 at 11:02 AM, Jean-Marc Valin <jmvalin at jmvalin.ca> wrote: > On 26/08/16 11:40 AM, Daniele Barzotti wrote: > > You're right! I forgot to say that I need only the encoder

I am setting up dovecot 1.0-rc15 as a proxy. When I try to log in, I get errors such as the following: Dec 4 17:39:23 vega dovecot: Dovecot v1.0.rc15 starting up Dec 4 17:39:24 vega dovecot: auth-worker(default): mysql: Connected to localhost (ad_mail) Dec 4 17:39:31 vega dovecot: auth(default): client in: AUTH 1 PLAINservice=POP3 lip=xx.xx.xx.xx rip=xx.xx.xx.xx

Hi, I have a big big problem. I compiled Samba 3.2.6 for Solaris 8 and 10. Environment is Windows 2003 SFU. So far it looked like everything works as expected. Under Solaris 8 i get all groups of a user. Under Solaris 10 I can only see the primary group set under the Unix Attributes Tab in W2k3. I have compiled with the following settings: ./configure --with-ads --with-krb5=/opt/VEGA

Hi Jean-Marc, thanks a lot for your reply. > Well, the first question is whether you want encoding, decoding, or > both. If there's one you don't need then you can remove that > (unfortunately, there's no easy way right now). You're right! I forgot to say that I need only the encoder side (and only for voice). My application have to acquire a 16bit 8KHz PCM stream and

O.k., some more info: I just tested bandwidth widening to fix this. But I need to go to gamma values below 0.9 to become stable -- clearly too much widening, I think. I looked inside the Levinson-Durbin algorithm next. The lines #ifdef FIXED_POINT r = DIV32_16(rr+PSHR32(error,1),ADD16(error,8)); #else r = rr/(error+.003*ac[0]); #endif look interesting. While for floating point,

Hi, I found this message concerning VAD and was wondering whether VAD has been ported to fixed-point in the latest version? Thanks, SingHui ---------- Forwarded message ---------- From: Jean-Marc Valin <Jean-Marc.Valin@usherbrooke.ca> Date: Jul 22, 2005 1:02 AM Subject: Re: [Speex-dev] Fixed-point To: gue baja <gue_baja@yahoo.com> Cc: speex-dev@xiph.org Hi Baja, Here's a quick

Hi, I am working on trying to compile opus to be as small as possible. I have passed the FIXED_POINT and the DISABLE_FLOAT_API but I am looking into making this lib even smaller in size. Where can I find out about how to best strip this lib to only compile in what I need?

Hello All, We are running Asterisk on a linux server as a SIP proxy with Cisco ATA 186's at the subscriber end. For long distance we have iax2 connectivity with a ip carrier. For local calls we are routing out through a commercial VEGA voicestream pots unit to an adtran channel bank and then from there to our class 5 soft switch. The sip to sip calls and the long distance calls work great.