[Ffmpeg-devel] [PATCH] 24-bits flac support (with 16-bits output)

Dave Chapman dave
Wed Jun 21 16:29:32 CEST 2006


Luca Barbato wrote:
> Dave Chapman wrote:
> 
>>Luca Barbato wrote:
>>
>>>Dave Chapman wrote:
>>>
>>>
>>>>i.e. it might be more efficient to have an if statement to choose
>>>>between two loops - one performing 32x32->32 and one performing
>>>>32x32->64, depending on the precision required.
>>>>
>>>
>>>feel free to try this patch, if nobody complains I'll add it.
>>>
>>
>>I'm assuming this means that for some frames of > 16-bit samples, it's
>>possible to use the 32-bit version of the loop without fear of overflow
>>- further optimising for CPUs where the 32x32->64 multiplication is
>>expensive.
> 
> 
> The patch got committed with an unoptimized check then, could you check
> with samples how much we lose?

I've just done a quick check with the only 24-bit file I've got to hand,
and the total number of frames using the 32x32->32 version of the loop
(with the "optimised" check) was zero...

So it seems that at least for my test file (which I tried encoding with
both flac -5 and flac -8) the libFLAC way of doing things gives no
improvement at all.

Dave.




More information about the ffmpeg-devel mailing list