On 03/12/14 14:37, wm4 wrote:
On Tue, 2 Dec 2014 16:48:25 +0000
Vittorio Giovara <[email protected]> wrote:
From: Gonzalo Garramuno <[email protected]>
This is needed to avoid banding artifacts when gammaing the picture.
Currently, if done with a video filter, the process is done on uints
instead of full float.
---
The correct fix is of course to add a float pixel format.
Since you have the quite annoying problem of having to pass the gamma
(for bayer you'd have an arbitrary large matrix) while you can simply do
that in the decoder since it is sort-of codec specific in itself.
Some users do not want to juggle around this kind of data and they also
complained already when I probed them about such topic (Hi jb).
So if possible I wouldn't have hard to deal with pixel formats that only
swscale could somehow manage (by adding additional hacks to fit them in
btw). More and more people do not use swscale and would be less than
happy about it.
lu
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
