A junk test from BBC

[birdie]

The results and an extensive analysis of the formal subjective verification tests of the H.265/HEVC video compression standard are published in the IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), January 2016. BBC R&D video coding research team focused on evaluations of UHD content and definition of analytics as part of standardisation process and presented in this paper. The full text of the paper can be accessed here.

http://www.bbc.co.uk/rd/blog/2016/01/h-dot-265-slash-hevc-vs-h-dot-264-slash-avc-50-percent-bit-rate-savings-verified

the img tag doesn't work apparently, sorry

http://www.bbc.co.uk/rd/images/dynamic/W1siZmYiLCJwdWJsaWMvcmQvc2l0ZXMvNTAzMzVmZjM3MGI1YzI2MmFmMDAwMDA0L2NvbnRlbnRfZW50cnk1MDRlMTgxNTcwYjVjMjBhMGMwMDEzNDQvNTY5NGZlZjIwNmQ2M2VjYjFiMDA0MzFmL2ZpbGVzL2dyYXBoMTE4NHg2NzIuanBnIl0sWyJwIiwidGh1bWIiLCIxMjQ4eDcwMiMiXV0/graph1184x672.jpg?sha=7bf5db3cd3a42710

Oh, yeah, those four guys sitting two meters away from the screen surely didn't notice any loss in details.

[vivan]

> 5 mbit 2160p@50/60fps
> reference encoders
> using PSNR because VQM is too slow and SSIM is not as frequently used.

[littlepox]

PSNR just don't give a d*** about the grain retention. A few years ago there were tons of AVC encoders claiming themselves better than x264, WITH PSNR measurement. (and they ignored the --tune psnr in x264, so while x264 was trying to keep the grain pattern/complexity, those crappy encoders happily wiped out everything and got a higher score. )
SSIM shall be slightly better, slightly.

[foxyshadis]

At least they didn't try to weight time-vs-psnr with the reference encoders, that would have had the complete opposite problem. Since I personally know some of the BBC guys keep up on this stuff, I wouldn't be surprised if there were a few facepalms and quiet mentions of advances in meaningful measures in the last two decades.