Hacker News Reader

Comment by mastax | Hacker News Reader

I had regarded the lossless re-compression as an interesting parlor trick, but I get it now.

Coincidentally today I was testing how much we can shrink our archived videos. We have lots of videos recorded with the Intel hardware encoder in H264, so not even the state of the art of H264. I figured: pop it into handbrake, change the codec to H265 (X.265), and boom 30%+ reduction. No, the H265 file got 3X larger. Well the default handbrake quality might be higher than the original video, and now it's going to some spend bits trying to encode the artifacts in the source video, while adding artifacts of its own. Sure you can find a happy medium where the quality won't be too much worse and you reduce the file size in most cases. But it sure would be nice to be able to losslessly recompress those videos!

View on HN

> But it sure would be nice to be able to losslessly recompress those videos!

And you highlight a crucial thing here: being able to do so without fiddling with parameters for hours to fine-tune it or even just finding something that works at all.

Most online video uses crazily overspecified h.264 bit rates for low complexity content. It's often possible to get 1080p well under 2 Mbit/s with little to no quality loss, and even lower by using 2 pass encoding where that's available. I'm not sure how things are with h.265 in a production setting, but at least for home use it seems to have much of the same flexibility

I get nervous when I see motion artefacts on netflix. This means I'm pretty much always nervous when watching netflix. I use a 1 gbps home connection and pay for the most expensive option netflix has. I'd like to think that the whole concept of motion frames based prediction will disappear in future videos.

Motion JPEG XL would be like in the movies where they have Motion JPEG 2000, but ~35-40 % more dense. We could add usual delta frames without motion without compromising quality criteria. This would get us in the 0.3 BPP range.

4k at 30 Hz would be 3840 x 2160 x 24 x 0.3 should need about 60 mbps (~ 7.5 MB/s), still doable for home internet speeds and would be visually lossless, a better experience than home movie streaming is today. (free startup idea) :-)

60 mbps is easily beyond the limits of many home wifi network setups, and leaves the serving end with capacity for something like 166 users per 10 Gbit port. There are many additional costs to consider beyond the size of the home pipe

I believe this capacity roughly triples every five years -- so when building something for the far future things can look different in bandwidth/quality perspective.

Do you work for Amazon Video or Netflix?

Reply on news.ycombinator.com