[x264-devel] Re: X264 Decoder

[Stephen Henry]

Not really because it's too generic; in software design, being generic
is a good thing. The reason I wanted the x264 decoder to work was so
that I could have a useful, and well written, substitute for the
reference code. If your working in H.264 it's a bit clunky to refer to
x264 for the encoder side of things, and ffmpeg, and entirely different
code base for anything decoder. That and the fact that the majority of
the code necessary to decode and reconstruct an H.264 bitstream is
already present in x264. 

-----Original Message-----
From: x264-devel-bounce at videolan.org
[mailto:x264-devel-bounce at videolan.org] On Behalf Of Hilton Garcia
Fernandes
Sent: 10 November 2005 14:14
To: x264-devel at videolan.org
Subject: [x264-devel] Re: X264 Decoder

Sorry if that's been already discussed, but i'd like to ask why to
replicate the decoder effort of the libavcodec team ? 

Is it because libavcodec is too generic ? 

Thanks,
Hilton

On Thu, Nov 10, 2005, Stephen Henry <Stephen.Henry at elixent.com> said:

> I had a look at the decoder a while back, for the same reasons as 
> Dimosthenis suggested, but I quickly found that the shared code 
> between the encoder and decoder and since diverged quite 
> significantly. It actually took quite a lot of effort just to get the
decoder to compile.
> Many of the problems I was having wasn't the fact that the decoder 
> code was incomplete, in actually the majority of it had already been 
> written, it was the fact that because the "common" structures and 
> functions to the encoder and decoder appeared to change since the 
> decoder was last run. This meant that subtle bugs would creap into the

> decoder and would be quite difficult and lengthy to track down where
it had gone wrong.
> 
> To be honest, you might have better luck if you simply delete the 
> majority of the decoder code and simply build it up from stratch. It 
> would probably take  you less time overall...
> 
> Stephen
> 
> -----Original Message-----
> From: x264-devel-bounce at videolan.org
> [mailto:x264-devel-bounce at videolan.org] On Behalf Of Dimosthenis 
> STAMATIOU
> Sent: 09 November 2005 15:16
> To: x264-devel at videolan.org
> Subject: [x264-devel] Re: X264 Decoder
> 
> Heejune AHN wrote:
>> Thanks, Guillaume.
>>
>> So, you don't have any plan for decoder release anyway.
>>
>> Thanks for reminding me of ffmpeg. I forgot it. I rechecked it today.
> but it sill has many "FIXME" and "incompleted functions" in my view. 
> In fact I searched archive for 5 months before November. I could't 
> find the same topics, just questions on how to build decoder t in
Windows.
>>
>> Anyway, I should review ffmpeg h264.c and update x264 anyway for my
> internal use.
>>
>> Thanks again!
>> Heejune
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> Hi,
>>
>> On 11/8/05, Heejune AHN <heejune at snut.ac.kr> wrote:
>>   
>>> Hi,  I am Heejune, quite new to this project, but I have experience 
>>> with MPEG-2, MPEG-4  codecs.
>>> After I found X264 is one of fast h264 codec in public domain, I am 
>>> doing hard to catch up the progress.
>>> Several question on the Decoder here.
>>>     
>>
>>
>> You should have dig the archives, all if not most of your questions 
>> have been answered before.
>>
>>
>>   
>>>    -  How is going with the decoder part? why Encoder first ?
>>>        Any due date for decoder part? We need decoder before the end

>>> of this year.
>>>     
>>
>> FFmpeg features a state-of-the-art decoder already. There's no plan 
>> that I know of to improve x264 decoder.
>>
>> Guillaume
>> --
>> Reading doesn't hurt, really!
>>   -- Dominik 'Rathann' Mierzejewski
>>
>>   
> Hi.
> 
> The fact that there's no plan of improving the internal x264 decoder 
> doesn't mean you're prohibited from doing such a thing, it's just that

> the effort is mainly driven on improving the encoder first. Also, 
> making the internal decoder work would mean no longer having to 
> download the whole ffmpeg library just to decode x264(at least on
Windows, that is).
> 
> Dimosthenis
> 
> --
> This is the x264-devel mailing-list
> To unsubscribe, go to: http://developers.videolan.org/lists.html
> 
> 
> 
> --
> This is the x264-devel mailing-list
> To unsubscribe, go to: http://developers.videolan.org/lists.html
> 
> 

--
Hilton Garcia Fernandes
Nucleo de Tecnologias sem Fio (NTSF) -- Wireless Technologies Team Lab
de Sistemas Integraveis Tecnologico (LSI) -- Integrable Systems Lab
Escola Politecnica (Poli) -- Engineering School Univ S Paulo (USP)
Tel: (5511)3091-5676 (work)
     (5511)8131-5213 (mobile)
Av. Prof. Luciano Gualberto,158 trav.3 CEP 05508-900 S. Paulo -- SP --
Brazil Pagina inicial: http://www.lsi.usp.br/~hgfernan


-- 
This is the x264-devel mailing-list
To unsubscribe, go to: http://developers.videolan.org/lists.html



-- 
This is the x264-devel mailing-list
To unsubscribe, go to: http://developers.videolan.org/lists.html