[FFmpeg-devel] [libav-devel] [RFC] M-bit to N-bit YUV conversions

[Michael Niedermayer]

On Thu, Aug 25, 2011 at 08:47:23AM +0200, madshi wrote:
> 2011/8/25 Michael Niedermayer <michaelni at gmx.at>
> > for 8->10 full range its a compromise between loosing
> > the 3 brightst levels vs. some slight loss of smoothness.
> 
> There's a solution which has neither of these two
> disadvantages:
> 
> 10bit = dither(8bit * 1023.0 / 255.0)
> 
> Of course that's strictly only for full range content, it
> shouldn't be used for limited range content.

true, i had somehow missed that the low frequency can be made more
fine grained and accurate at the cost of poorer accuracy at the
high frequency.
I agree this likely is the ideal solution from a psychovissual point
of view


> 
> >> b) Should we distinguish between limited and full range YUV?
> > for 8->10 it likely doesnt make a vissible difference
> 
> I'm not sure about that. At least the current way to use
> unsigned byte duplication for chroma upscaling is probably
> visible, because gray isn't gray after upscaling, anymore
> (128 becomes 514).

fixed in ffmpegs swscale :)
took less time than the reg tests 

more bugreports welcome!

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

The greatest way to live with honor in this world is to be what we pretend
to be. -- Socrates
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20110825/b90b6097/attachment.asc>