similar to: No subject

Keith Kyzivat wrote: > Caveat: Take what I'm saying with a grain of salt -- I haven't actually > written any code that decodes speex frames myself, though I have written > code using Jean-Marc's Speexdsp lib. > > Comments below inline > > On Thu, Oct 9, 2008 at 9:28 AM, Darren Longhorn > <darren.longhorn at redembedded.com > <mailto:darren.longhorn at

Jean-Marc Valin wrote: > > > The decoder will return -1 if it can't decode anything else, so you just > do something like: > > while(1) { > err = speex_decode_int(state, bits, pcm); > if (err==-1) > break; > output_data(pcm); > } I think I still didn't fully understand. As a test I encoded some frames (wideband, with quality set to 8),

Oh I overlooked that one.. Thanks Darren! Btw, I was wondering if anyone has tried using the echo canceller function of speex with other codec like g.711? On Tue, Jan 17, 2012 at 8:32 PM, Darren Longhorn < darren.longhorn at redembedded.com> wrote: > On 14/01/12 12:26, Christopher Adoremos wrote: > > What is the highest quality audio data rate and sample rate Speex can >

Ok! If there are indeed cross-compiling targets defined, then vs2005 projects probably should be maintained. In any case, this is all up to Jean-Marc, since he's the one that actually has authority to change it :) Jean-Marc -- what's your opinion? On 11/28/07, Aron Rosenberg <arosenberg@sightspeed.com> wrote: > > I pulled the test directory from 2003, but everything else

On Thu, Oct 2, 2008 at 12:30 AM, Erik de Castro Lopo <mle+la at mega-nerd.com<mle%2Bla at mega-nerd.com> > wrote: > Keith Kyzivat wrote: > > > And Alex has told me that visual studio-built projects cannot link with > > mingw-win32-gcc produced DLLs... > > This is only true when the DLL being called exports C++ name mangled > symbols. Is there a way to get

Hi Keith, Thanks a lot for your reply. Actually I am trying to build an application in linux & C for converting a voice file(that is WAV) to a text message. This is required for one of our voice project. So just wanted to know if there is any existing application already there in linux to do that. Otherwise I have to write my own. Now to right my own "wav to text" converter I

On Fri, Apr 11, 2008 at 6:53 PM, Joshua Gargus <schwa at fastmail.us> wrote: > Fantastic! Thanks Tom. > > (if only it was possible to change the VBR target on the fly for the > H264 codec we've licensed :-) ) Theora might be able to do that. Some other folks more closely tied with Xiph probably can comment more on that issue. -- Keith Kyzivat SIPez LLC. SIP VoIP, IM

Pintu -- Speex is an audio codec that can encode and decode audio for transmission over network (or for storage). Specifically, it's called speex because it is tailored to work well at encoding speech at small bitrates. Speex does not do speech recognition or text to speech. You may wish to look at CMU's speech page: http://www.speech.cs.cmu.edu/(look at Sphynx). And I'll tell you -

I pulled the test directory from 2003, but everything else from 2005. The 2005 files have more ?cross-compiling? targets than the 2003 ones. I fixed the various missing files in the new 2008 projects (misc.c and one other other are gone now). The only thing you could do which would make them more compact would be to get rid of the subdirectories under each vs-version. There is no real need for

Something to look into perhaps is Trolltech's 'qmake' tool. It fills the role of something like autotools or now defunct imake. As far as I understand, you can indeed use it for projects that don't do anything with the Qt libraries. Now, it may not have the support for multiple targets that it sounds like are needed for speex, and I know it doesn't have as comprehensive of

On 02/04/2008, Alexander Chemeris <Alexander.Chemeris at sipez.com> wrote: > On 4/1/08, Keith Kyzivat <kkyzivat at sipez.com> wrote: > > IIRC, under ARM, I believe I was having issues with --without-libogg. > > I resorted to just downloading libogg sources, setting up my > > cross-compilation environment, and cross compiling it per the general build > >

IIRC, under ARM, I believe I was having issues with --without-libogg. I resorted to just downloading libogg sources, setting up my cross-compilation environment, and cross compiling it per the general build directions (./configure, make, make install), then went back to building speex -- everything worked like a charm for me then. So -- Conrad. *Make sure* that you have libogg cross-compiled and

The short: Linking to libspeex.a when generating a .so using libtool results in a non-portability warning. This is due to PIC code and non-PIC code intermingling. How can I go about fixing this whilst still using an installed libspeex present on the user's system? The long: I am using autoconf + libtool to generate a codec plugin for speex (sipXmediaLib), and I'm trying to eliminate

The VS2005 project files are known to not be up-to-date, so I can understand that you'd need to fix some things. Can you verify with an import of the VS2003 project files, that there is no need for any fixes, and just works after importing? Managing multiple project files should be minimized if at all possible. If a straight import of vs2003 project files can be done without any changes

On 11/29/07, Jean-Marc Valin <jean-marc.valin@usherbrooke.ca> wrote: > qmake is just insanely limited in terms of the structure of your code > (can't have an executable if you have a library, ...). On top of that, > I've had way too many cases where it just "forgot" to recompile files > when a header file changed (causing strange segfaults) that I'm never

If you can split your .dcr files into the respective .wav files, you can process each one of those with the 'speexenc' command line encoder app included with the speex codec sources. Just configure, build, and go. speexenc should be self-explanatory -- run without args, and it will provide the usual cmd line opts help that *nix apps generally provide. On Dec 11, 2007 6:26 PM, Michael

This is best answered by the jspeex folks, as it is the interfacing code that you are having problems with. That being said, it looks like, just as it says, you have asked for a conversion it cannot do -- you want it to convert PCM data to speex data at 44.1kHz sample rate -- speex only supports 8kHz in standard operation. It can also support 16kHz (wb) and 32kHz (uwb), but nothing outside of

On 2/18/08, Premkiran Mannava <loverays@gmail.com> wrote: > > Hi, > > *"That's in general not very reliable. You can get PEAQ to say all sorts > of silly things." > > Can you provide me links for any more effective tools other than PEAQ? > Which is more reliable for Speex resampler?* I can already tell you what Jean-Marc will say -- use your ears :)

On Fri, Nov 14, 2008 at 3:57 AM, Vincent Burel <vincent.burel at vb-audio.com>wrote: > > Speech compression algorithms always are tunned to particular freq, > > else they would take tons of time. That's because they use knowledge > > that speech pitch (and other params) lies in well specified regions. > > Thus if you feed algorithm with wrong freq, you'll

On Thu, Oct 2, 2008 at 12:07 AM, Jean-Marc Valin < jean-marc.valin at usherbrooke.ca> wrote: > Tom Grandgent a ?crit : > > If you remove the project files, I doubt it's going to make things easier > > for anyone. You'll probably just increase the number of questions as > > people struggle to compile using less common and less user-friendly > > methods,