[Flac-dev] Re: Reg. FLAC decoding

[Josh Coalson]

--- Joe Steeve <joesteeve at zodiactorp.com> wrote:
> On Fri, 2005-09-30 at 23:14, Josh Coalson wrote:
> > but all samples in buffer[] are 32-bit signed integers in host
> > order, regardless of the bits-per-sample of the frame.  so to get
> > them down to shorts (assuming they'll fit), do like:
> > 
> >   FLAC__int32 * LChannel = buffer[0];
> >   short TempSource = (short)LChannel[LPos];
> > 
> > instead of
> > 
> >   LChannel = (unsigned char *)buffer[0];
> >   short * TempSource = (short *)&LChannel[LPos];
> 
> > let me know if this works or if you have other questions.
> 
> I tried this., but still i dont get the proper output. I'm still
> debugging with the decoder. I'll get back when I get doubts.

one thing that can help: create a stereo WAVE file with one
sample in each channel with a specific bit pattern like
0x55aa 0x5aa5, compress to FLAC, and decode that.  then it is
easy in a debugger to see how the samples are stored.

Josh



		
__________________________________ 
Yahoo! Mail - PC Magazine Editors' Choice 2005 
http://mail.yahoo.com