[FFmpeg-devel] [PATCH] hook up the atrac1 decoder

Vitor Sessak vitor1001
Tue Sep 22 20:09:12 CEST 2009

Benjamin Larsson wrote:
> Vitor Sessak wrote:
>> Benjamin Larsson wrote:
>>> Vitor Sessak wrote:
>>>>>> I think a good deal of code of both branches of the if() can be
>>>>>> factorized out.
>>>>> What code? IMO not much of the code in the 2 cases can be factored out
>>>>> and still be logical and clear.
>>>> The calls to at1_imdct() and vector_fmul_window() are very similar.
>>>> They are exactly the same if block_size == 32 and nbits == 5 when
>>>> num_blocks == 1. Note also that the for() loop of the else{} branch
>>>> will be run only once when num_blocks == 1, as it should. Of course,
>>>> there will always be a if() for the memcpy().
>>> Ok, I see how it could be factored. But the logic of passing
>>> parameters would be highly obfuscated.
>> Do you consider the following highly obfuscated or am I missing something?
> Now you really have to elaborate what you are meaning.

Ok, my completely untested patch was wrong, but this one (that is still 
missing a little clean up) does not change the output for samples in 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: atrac1.diff
Type: text/x-diff
Size: 2567 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20090922/f99989f2/attachment.diff>

More information about the ffmpeg-devel mailing list