[Speex-dev] Re: bug or somthing like it :)
Jean-Marc.Valin at USherbrooke.ca
Tue Jun 6 03:26:57 PDT 2006
I would definitely appreciate if you could send me an example of that.
Please send 1) the input file, 2) the speex-encoded file and 3) the
decoded file (that becomes unstable).
Le mardi 06 juin 2006 à 12:13 +0200, julio preciado a écrit :
> From: "Paul Bryson" <paul at msn.com>
> To: "julio preciado"
> <jepreciado at hotmail.com>,<webmaster at speex.org>
> Subject: Re: bug or somthing like it :)
> Date: Thu, 1 Jun 2006 18:57:43 -0500
> This address is for website issues. For technical questions
> about Speex, please contact the Speex mailing lists.
> ----- Original Message -----
> From: julio preciado
> To: webmaster at speex.org
> Sent: Thursday, June 01, 2006 10:04 AM
> Subject: bug or somthing like it :)
> I am using Speex v.1.1.12 in a videoconference
> software application, using WinXP and
> Visual Studio .NET 2003. I am compiling Speex in
> floating-point, 44kbps and UWB mode. Audio input
> is 32Khz 16bit mono.
> the "problem" is that some times there are input
> signals produced by some mixers(hardware) in the
> system (mic-mixer-encoder-network-decoder-speakers)
> that make Speex decoder unusable, it
> begins to produce a decoded output signal entirly
> zero(like a silenze).
> When I was debugging I dicovered that there are a lot
> of multiplications that produce floating
> point overflow exceptions(Not A Number).
> functions like compute_rms(const spx_sig_t *x, int
> len) in file filters.c produce this kind of
> I don't know if you alredy know this "problem" or not,
> but, if you want I can send you in a raw file
> the bitstream encoded by Speex encoder.
> I've temporarily resolved it compiling Speex in fixed
> point mode.
> p.s. Congratulations for your work, Speex is the best
> WWW audio encoder.
> webmaster mailing list
> webmaster at xiph.org
More information about the Speex-dev