Ah, well, I dont believe that I want each coefficient set seperated, that
is against the iso definition of the mpeg standard, right?  

anyway, I just need to figure out where in the code is the frame ready to
be outputted as one frame.  does the buffer that is written contain only
one frame at a time?  is that what formatbitstream does?  maybe i should
study the code a lil harder b4 coming for help.  
-kali

On Thu, 4 Nov 1999, Mark Taylor wrote:

> 
> > how can I read the mp3 data outputted by lame frame by frame?  I noticed
> > in common.c that all the bits are stuffed into a bit stream and then
> > shoved down the pipe, I couldnt figure out though, before that point, when
> > each frame was formed?  My point in doing so:  I want to serve the
> > microsoft media client.  they require that the frames are all of uniform
> > size and that there is some funky header preceding each mp3 frame. that is
> > why I need the individual frames.  and also, where can I find out the
> > individual frame sizes in bytes?  I emailed mark taylor already regarding
> > individual frame size uniformity, and he told me that I just need to set
> > the padding bit to 0 in order to have uniform frames, will this cause
> > problems of any sort?  thanks a bunch yall.
> > -kali
> > 
> 
> You will probably have to be more specific.  An MP3 frame looks like:
> 
> header
> side info
> data
> 
> The size of the frame varies by at most one byte (at 44.1khz).  AT
> 48khz, all frames will be the same length.
> 
> However, the actual encoded coefficients for a given frame will
> usually be spread over two frames.  For example, 20% of it might be in the
> data block of the previous frame, and 80% is stored in the data
> block of the current frame.  
> 
> This size of the encoded coefficients is variable, it can range
> from 0 to 6000 bits more or less, and it is impossible to make
> every frame use exactly the same amount of data.
> 
> Because of this, a frame is not a selfcontained unit.  You
> usually need at least two frames before you can decode one frame worth
> of data.
> 
> Some decoders work by creating new "meta frames" which contain
> the header, side info and all data associated for that frame.
> If MS wants this kind of information, you are going to have to 
> really muck around with the bitstream.  The mp3asm code does some
> of this.  
> 
> Mark
> 
> 
> 
> 
> 
> 
> --
> MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
> 

--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )

Reply via email to