[x265] Severe quantizer difference between {...faster} and {fast...} presets

Deepthi Nandakumar deepthi at multicorewareinc.com
Tue Feb 11 10:44:44 CET 2014


Hi,

Between faster and fast, a number of key default options change. In order
of effect, the options that cause this would be bFrameAdaptive (jumps to 2
in fast), --early-skip, --fast-cbf, and then --max-ref. (Assuming you have
already disabled cutree in fast preset).


On Tue, Feb 11, 2014 at 1:09 PM, Mario *LigH* Rohkrämer <contact at ligh.de>wrote:

> https://bitbucket.org/multicoreware/x265/issue/26
>
> There is a 4..5 dB SSIM drop (filesizes 1:4..1:5) between presets
> {'faster' and faster} and {'fast' and slower}, even if CUTree is disabled.
>
> I verified this gap with several builds I was able to compile with GCC
> 4.8.2, last version was v0.7+103-ce7191f49948 Win64.
>
> I don't believe your desire is the confusion of the user to be unable to
> at least marginally rely on a CRF value, even across different presets. So
> I guess there will be a mistake causing this severe quality difference
> between the two preset groups. But due to the lack of a current Evaluators'
> Guide for v0.7+, I an mot able to guess with any certainty which
> preset-based option default may cause that.
>
> --
> __________
>
> Fun and success!
> Mario *LigH* Rohkrämer
> mailto:contact at ligh.de
>
> _______________________________________________
> x265-devel mailing list
> x265-devel at videolan.org
> https://mailman.videolan.org/listinfo/x265-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20140211/b332946f/attachment.html>


More information about the x265-devel mailing list