Justin Ruggles wrote:
> Michael Niedermayer wrote:
>>>         ff_ac3_bit_alloc_calc_psd((int8_t *)s->dexps[ch], start, end,
>>>                 s->psd[ch], s->bndpsd[ch]);
>> why the int8_t cast?
> 
> That's leftover from Fabrice's original AC3 encoder, which uses int8_t 
> for exponents.  That function is shared between the encoder and decoder. 
>   I see no reason why it can't be changed, but it would have to be 
> across-the-board.

Oh wait, nevermind.  I saw this in my last review.  The variable should 
be changed in the context to be int8_t and the cast would not be needed. 
  Or again we could just change it everywhere to plain int.

-Justin
_______________________________________________
FFmpeg-soc mailing list
[email protected]
http://lists.mplayerhq.hu/mailman/listinfo/ffmpeg-soc

Reply via email to