[FFmpeg-devel] Make ff_acelp_lsp2lpc() more generic

Måns Rullgård mans
Wed Jun 16 18:11:48 CEST 2010

"Ronald S. Bultje" <rsbultje at gmail.com> writes:

> Hi,
> On Wed, Jun 16, 2010 at 9:30 AM, Mohamed Naufal <naufal11 at gmail.com> wrote:
>> The attached patch allows using the function by both G.729 & G.723.1.
> I might be late here, but is there a reason to not implement 723.1 in
> float? I thought we were trying to implement new codecs in float if
> possible.

It depends.  Anything lossless should be done in fixed-point for
obvious reasons.  If the codec is likely to be used on FPU-less
devices, a fixed-point implementation is probably also preferable.

We use floating-point mostly because it's easier to write, and because
it often is faster on modern CPUs, not because floating-point is
somehow inherently better.

M?ns Rullg?rd
mans at mansr.com

More information about the ffmpeg-devel mailing list