--- Manjunath D N <[EMAIL PROTECTED]> wrote:
> Hello!
>       I am trying to develop an MP3 encoder on a 24 bit
> fixed point
> DSP.Due to fixed point degradations, there is a
> mismatch in the
> blocktypes that my assembly code psychoacoustic
> model gives wrt the lame
> floating point code.For example if for a granule the
> floating code gives
> blocktype as 1(START type), my assembly code may
> declare blocktype as
> 3(STOP type)
>       My question to the lame developers is what exactly
> is the effect
> of this mismatch on the final mp3 bitstream? Will
> there very bad
> output(noise) or totally absent output for that
> frame?Because in my
> assembly code I am getting zero output if blocktype
> does not match for
> that frame
>       As far of my understanding goes, the difference
> between various
> blocktypes is the windowing function. The mismatch
> may only reduce the
> quality of the final mp3 stream(ie ODG reduces), but
> it cannot make
> output zero for that frame.
> I would like to know the opinions of lame developers
> regasrding this
> problem.

You are right. Variation in BLOCK_TYPE should only
degrade the quality.
 Also it is very difficult to always match the
BLOCK_TYPE with that we get in float code as there is
lot accuracy loss in Psycho model.
 Please check the asm code.

 Is it a problem in my assembly code
> implementation or is it
> what happens when blocktypes don�t match?
>       Thanks in advance ..
> With warm regards,
> Manjunath D.N. 
> 
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On
> Behalf Of
> [EMAIL PROTECTED]
> Sent: Saturday, December 20, 2003 7:30 AM
> To: [EMAIL PROTECTED]
> Subject: mp3encoder Digest, Vol 7, Issue 8
> 
> 
> Send mp3encoder mailing list submissions to
>       [EMAIL PROTECTED]
> 
> To subscribe or unsubscribe via the World Wide Web,
> visit
>       http://minnie.tuhs.org/mailman/listinfo/mp3encoder
> or, via email, send a message with subject or body
> 'help' to
>       [EMAIL PROTECTED]
> 
> You can reach the person managing the list at
>       [EMAIL PROTECTED]
> 
> When replying, please edit your Subject line so it
> is more specific than
> "Re: Contents of mp3encoder digest..."
> 
> 
> Today's Topics:
> 
>    1.  (Vitaliy Trizna)
>    2. Re:  ([EMAIL PROTECTED])
>    3. Re:  (Vitaliy Trizna)
>    4. Re:  ([EMAIL PROTECTED])
> 
> 
>
----------------------------------------------------------------------
> 
> Message: 1
> Date: Fri, 19 Dec 2003 20:46:44 +0600
> From: "Vitaliy Trizna" <[EMAIL PROTECTED]>
> Subject: [mp3encoder] 
> To: [EMAIL PROTECTED]
> Message-ID: <[EMAIL PROTECTED]>
> Content-Type: text/plain; charset="KOI8-R"
> 
> Hello!
> 
> it seems to me that Lame 3.93.1 has one bug. I use
> lame to decode mp3
> stream. First I initialize decoder:
>     lame_decode_init();
> and then I try to use lame_decode function, but
> always it returns me 0
> (need more data before we can complete the decode).
> I have decided this
> problem, but I haven't confidence that all right. 
> I have changed function InitMP3:
> //mp->fsizeold = -1;
> mp->fsizeold = 0;
> 
> With best regards,
> Vitaliy V. Trizna
>
---------------------------------------------------------
> http://ngs.ru/ny - ����� ��� � ������������!
> 
> 
> 
> 
> 
> 
> 
> 
> 
> ------------------------------
> 
> Message: 2
> Date: Fri, 19 Dec 2003 16:30:34 +0100
> From: <[EMAIL PROTECTED]>
> Subject: Re: [mp3encoder] 
> To: "MP3 encoders development list"
> <[EMAIL PROTECTED]>
> Message-ID: <[EMAIL PROTECTED]>
> Content-Type: text/plain;     charset="KOI8-R"
> 
> Hi
> 
> I didn't know there was a lame decode function
> available. Do you have
> the interface for that fuction!
> 
> Lars
> 
> ----- Original Message ----- 
> From: "Vitaliy Trizna" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Friday, December 19, 2003 3:46 PM
> Subject: [mp3encoder]
> 
> 
> > Hello!
> >
> > it seems to me that Lame 3.93.1 has one bug. I use
> lame to decode mp3 
> > stream. First I initialize decoder:
> >     lame_decode_init();
> > and then I try to use lame_decode function, but
> always it
> > returns me 0 (need more data before
> > we can complete the decode).
> > I have decided this problem, but I haven't
> confidence that
> > all right.
> > I have changed function InitMP3:
> > //mp->fsizeold = -1;
> > mp->fsizeold = 0;
> >
> > With best regards,
> > Vitaliy V. Trizna
> >
>
---------------------------------------------------------
> > http://ngs.ru/ny - ����� ��� � ������������!
> >
> >
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > mp3encoder mailing list
> > [EMAIL PROTECTED] 
> > http://minnie.tuhs.org/mailman/listinfo/mp3encoder
> >
> 
> 
> 
> ------------------------------
> 
> Message: 3
> Date: Fri, 19 Dec 2003 22:02:45 +0600
> From: "Vitaliy Trizna" <[EMAIL PROTECTED]>
> Subject: Re: [mp3encoder] 
> To: MP3 encoders development list
> <[EMAIL PROTECTED]>
> Message-ID: <[EMAIL PROTECTED]>
> Content-Type: text/plain; charset="KOI8-R"
> 
> Hello!
> 
> lame_decode  function is available. It declare in
> lame.h
> file. About what interface do you speek? 
> 
> > I didn't know there was a lame decode function
> available.
> > Do you have the
> > interface for that fuction!
> > 
> > Lars
> > 
> > ----- Original Message -----
> > From: "Vitaliy Trizna" <[EMAIL PROTECTED]>
> 
=== message truncated ===


__________________________________
Do you Yahoo!?
New Yahoo! Photos - easier uploading and sharing.
http://photos.yahoo.com/
_______________________________________________
mp3encoder mailing list
[EMAIL PROTECTED]
http://minnie.tuhs.org/mailman/listinfo/mp3encoder

Reply via email to