search

ericmckean / webm

Automatically exported from code.google.com/p/webm
0 stars 0 forks source link

Optimal Chrome Transform Size Decision #751

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
In 420 case, typically chrome transform size is determined by 1/4 of 
corresponding luma block size or the smallest transform size available. This 
may or may not be the optimal decision for non-420 cases. 

Since VP9 profile 0 supports only 420, this was not an issue, but for future 
profile supporting 422 or 444, this decision should be re-evaluated for better 
consistency and/or better compression efficiency

Original issue reported on code.google.com by ya...@google.com on 28 Mar 2014 at 3:11

GoogleCodeExporter commented 9 years ago

Original comment by ya...@google.com on 28 Mar 2014 at 3:11

GoogleCodeExporter commented 9 years ago 
https://gerrit.chromium.org/gerrit/#/c/69383/
will rationalize/generalize the transform size choice.

Original comment by ya...@google.com on 28 Mar 2014 at 3:15

GoogleCodeExporter commented 9 years ago 
With this patch, for a natural clip I tested I see a small increase in Y PSNR 
and a decrease in U and V PSNR. For a screen cast material I see a decrease in 
Y PSNR and an increase in U and V PSNR. This seems a bit strange to me.

Original comment by aconve...@google.com on 28 Mar 2014 at 5:33

GoogleCodeExporter commented 9 years ago 
https://gerrit.chromium.org/gerrit/#/c/69383/ has been merged.

Original comment by ya...@google.com on 3 Oct 2014 at 3:43