2 ms·
Codecs for the 4K Era: HEVC, AV1, VVC and Beyond
- CharlesW 4y ago> We use Netflix’s Emmy-winning Video Multimethod Assessment Fusion (VMAF) to assess visual quality, as it’s the best available tool at the moment. Maybe not? From "Who Has the Best Hardware AV1 Encoder?", discussed in another HN thread (https://news.ycombinator.com/item?id=35587509 https://news.ycombinator.com/item?id=35587509): "Everyone's favorite bad, obsolete, easily tricked metric is VMAF, which was good for a time but simply isn't good anymore."
- ksec 4y agoI think it is fair to say VMAF was never meant to measure still image quality. It is also not perfect and Netflix has plenty of example of areas known to be problematic. But it is still far better than say PSNR or SSIM. The major problem is when encoder optimise for VMAF only.
- Dalewyn 4y agoI like how the anime fansubbing community does it: Take random clips of anime, do various encodes under real world conditions, compare results.
- bick_nyers 4y agoHave any good examples of someone doing this?
- CharlesW 4y agoThe MSU Video Codecs Comparisons do this (see "Subjective Comparison Methodology"): https://www.compression.ru/video/codec_comparison/2021/subjective_report.html https://www.compression.ru/video/codec_comparison/2021/subje... As far as I know, human ABX testing is still the best way to measure the relative perceptual quality of different codecs.
- amluto 4y agoThis compares VVC in VBR mode to the other in CRF mode. I can imagine just that difference accounting for some of the apparent improvement for VVC.
- doodlesdev 4y agoThis table is actually insane. Just from this I would say that HEVC wins if it were to have even comparable quality. It's unfortunate the industry will have to stay with AV1 for now due to HEVC being patent encumbered and because of Google's massive push to force everyone to start using it. Encoding Time | Decoding Time HEVC 2h 34min | 1min 14sec AV1 18h 54min | 1min 42sec VVC 2 days 10h 35min | 2h 42sec
- izacus 4y agoIt's insane because they use an insane codec configuration and then just decide to step back on HEVC for no reason and use a faster preset.
- redeeman 4y agoim sure a lot could be done to make both AV1 and VVC fastrer at encoding (and decoding). x264 has been optimized to extreme levels, i'd imagine x265 has also gotten A LOT of effort, probably way more than AV1
- Gigachad 4y agoNot to say encode time does not matter. But it doesn't matter _that_ much. Companies like netflix will encode once, stream millions of times. Comparatively the encode time is a rounding error. Will matter a lot for things like personal recordings on phones and video calls though. It is possible to get high speed AV1 encodes on phones but I don't know if the quality is acceptable.
- izacus 4y agoThis is one of the worst comparisons I've seen in a long while - using a poor metric and utterly nonsensical incomparable codec settings.
- aaron695 4y ago[dead]
- ksec 4y agoI am sort of surprised this coming from Chip and Cheese didn't talk about the Hardware part, die size of HEVC, AV1, VVC Encoder and Decoder. The complexity of DSP / ASIC. Although it is probably worth mentioning, HEVC manage to hit 90 VMAF on 4K with only 1.5Mbps. You can infer that with 2K / 1080P which is the most common resolution would require less bitrate. So depending on business model a new codec might actually cost more in terms of total storage requirement than Bandwidth Saving as bandwidth cost continues to fall. ( Think YouTube ) i.e The codec to replace AVC / H.264 needs to be so so much better. Otherwise a higher bitrate AVC may actually be better than a newer codec. I also wonder if we could derive a royalty free codec from JPEG-XL.
- enz 4y ago> I also wonder if we could derive a royalty free codec from JPEG-XL. I guess we could imagine a sequence of JXL frames (a keyframe-only codec). Something like Motion JPEG, but better.
- hajile 4y agoThis wasn't written by clamchowder and is this author's first submission. Hopefully there's room for them to grow.