[FFmpeg-devel] [PATCH] QCELP decoder postfilter & final gain control

[Reynaldo H. Verdejo Pinochet]

Hi Kenan

Kenan Gillet wrote:
> 
> - patches 1, 2 and 3 are splitting and cleanup the current code.

I'd reject patch 2 out of it not providing any real advantage. The
first mult (0*40) (all even?) should get optimized out anyway and
we would end up adding a final 'rewind' that it's not needed with
the current code.

Maybe someone can correct me if I'm wrong?

Bests
--
Reynaldo