[FFmpeg-devel] [PATCH] Common ACELP code & G.729 [4/7] - G.729 core

[Diego Biurrun]

On Sat, May 24, 2008 at 04:46:19PM +0700, Vladimir Voroshilov wrote:
> 2008/5/18 The Wanderer <inverseparadox at comcast.net>:
> > Diego Biurrun wrote:
> >
> >> On Sun, May 18, 2008 at 12:15:56AM +0700, Vladimir Voroshilov wrote:
> >>>
> >>> Next update.
> >>>
> >>> --- /dev/null
> >>> +++ b/libavcodec/g729.h
> >>> @@ -0,0 +1,137 @@
> >>> +#define FC_SIGNS_BITS_8K     4  ///< fixed codebook signs, 8k mode (size in bits)
> >>> +#define FC_INDEXES_BITS_8K   13 ///< fixed codebook indexes, 8k mode (size in bits)
> >>> +#define FC_SIGNS_BITS_4K4    4  ///< number of pulses in fixed-codebook vector, 4.4k mode
> >>
> >> Sometimes you write "fixed codebook", sometimes "fixed-codebook".
> >> Settle on one spelling.  Mans or the Wanderer will advise you which
> >> way is preferred.
> >
> > Actually, these two forms mean different things.
> >
> > "fixed codebook signs" means "codebook signs which are fixed".
> >
> > "fixed-codebook vector" means "(the) vector associated with the fixed
> > codebook".
> 
> [...]
> 
> Thsnk you for good explanation.
> I meant "fixed-codebook" everywhere.
> 
> Updated patch attached.
> 
> --- /dev/null
> +++ b/libavcodec/g729.h
> +/**
> + * maximum gain pitch value (3.8, Equation 47)
> + * 0.7945 in Q14
> + * (EE) This does not comply with specification.

with the

> +    uint8_t fc_indexes_bits;  ///< Size (in bits) of fixed-codebook index entry

size

Diego