[x264-devel] Evaluating x264 -- first steps encoding greylevel images

[Jason Garrett-Glaser]

On Mon, Mar 15, 2010 at 1:01 AM, Eeri Kask <Eeri.Kask at inf.tu-dresden.de> wrote:
>>> In order to take x264 further beyond recreational deployments it
>>> would be great if at least yuv444-planar input encoding would be
>>> supported for color images, how much would it be different from
>>> current yuv420 implementation?
>>
>> A lot of work.  It's currently a summer of code project idea:
>> http://wiki.videolan.org/SoC_x264_2010 .  The primary problem is
>> that we don't currently have a _decoder_ for yuv444 (except of
>> course the reference).  Any useful implementation would also have
>> to write support for libavcodec as well.
>
>
> Under the assumption I understand it right, then x264.h contains
>
>    #define X264_CSP_I444           0x0003  /* yuv 4:4:4 planar */
>
> which I was referring to, thinking this is in fact "mathematically"
> essentially the same planar format as X264_CSP_I420 but at full
> spatial resolution.

x264 does not support any format beyond 4:2:0.  All the other CSPs in
x264.h are legacy and will be refused by the encoder.

> Should this appear true, then the decoder should look much like the
> one for I420 which we have already, even marginally simpler and
> faster indeed (due to not having to deal with reduced chroma
> resolution), so one would be tempted to just reuse the already
> existing code (for this special case, for now)...   :-)

If by "marginally simpler and faster" you mean "a 10,000 line patch
affecting all facets of x264", sure.

Dark Shikari