H.266/Versatile Video Coding (VVC) • Hacker News Board

Re: H.266/Versatile Video Coding (VVC)

#121

Post by pault » Mon, Jul 06, 2020, 2:44 PM UTC

sly010 wrote: post #64

Unklejoe wrote: post #14

It's interesting that they are able to continue improving video compression. You'd think that it would have all been figured out by now. Is this continued improvement related to the improvement of technology? Or just coincidental? Like, why couldn't have H.266 been invented 30 years ago? Is it because the computers back in the day wouldn't have been fast enough to realistically use it? Do we have algorithms today tha…

It's about the assumptions made during the standardization. Compared to 30 years ago, we now have better knowledge and statistics about what low level primitives are useful in a codec. E.g. jpeg operates on fixed 8×8 blocks independently, which makes it less efficient for very large images than a codec with variable block size. But variable block size adds overhead for very small images. An other reason can be common…

Something that I learned about 10 years ago when bandwidth was still expensive is that you can make a very large version of an image, set the jpeg compression to a ridiculously high value, then scale the image down in the browser. The artifacts aren't as noticeable when the image is scaled down, and the file size is actually smaller than what it would be if you encoded a smaller image with less compression.

Re: H.266/Versatile Video Coding (VVC)

#122

Post by donatj » Mon, Jul 06, 2020, 2:46 PM UTC

ctdeneen wrote: post #86

donatj wrote: post #3

Huh. I wonder how encoding speeds compare. I rarely chose h265 over h264 because similar levels of visual quality took massively more time.

But for most situation you encode once and play multiple times. Wouldn't it be better to reduce storage and bandwidth costs with a smaller file (assuming the same quality)?

It's a trade off. When I have a batch of 40 videos and encoding h264 takes 20 minutes per video and h265 takes 4 hours means the difference between 13 days and 160 days.

The latter isn't practical, I'll eat the couple hundred MB in order to save a lot of time.

Re: H.266/Versatile Video Coding (VVC)

#123

Post by striking » Mon, Jul 06, 2020, 2:48 PM UTC

ChrisMarshallNY wrote: post #117

cyphar wrote: post #45

Earlier quoted context omitted.

> The dirty secret of video codecs is that you can't make a modern video codec that isn't patent encumbered The existence of Theora, VP8, VP9, and now AV1 seems to contradict that theory. You could argue that they infringe on some unknown patents, but that is also arguably true of patent cabals like MPEG (you just hope that the cabal is big enough that there aren't any patentholders lurking outside). The only differe…

Yeah. It's a mess. The video patents aren't just "patent troll" patents, either. They are highly enforceable, and were registered by corporations like Ampex. I have been trying to write a simple app to stream RTSP (security cameras), and that has been a pain. I need to basically use either proprietary (paid) or GPL software to do it. Video software is not for the faint of heart. Much as I grouse about the licensing,…

You can license x264 to not have to worry about the GPL. Though that costs money and you may find yourself signing MPEG LA contracts.

edit: Your project reminds me, https://github.com/arut/nginx-rtmp-module is super worth checking out and might be helpful to you.

Re: H.266/Versatile Video Coding (VVC)

#124

Post by throwaway8941 » Mon, Jul 06, 2020, 2:48 PM UTC

pault wrote: post #104

speedgoose wrote: post #77

Earlier quoted context omitted.

They use ffmpeg, which is developed by people who do not care about software patents because it doesn't apply to them.

Why don't the patents apply to them?

Because the developers of x265 (which is the real codec and ffmpeg is a wrapper around it) are located in France.

http://www.videolan.org/developers/x265.html

https://en.wikipedia.org/wiki/VideoLAN

Re: H.266/Versatile Video Coding (VVC)

#125

Post by pault » Mon, Jul 06, 2020, 2:49 PM UTC

realusername wrote: post #44

hirako2000 wrote: post #4

Earlier quoted context omitted.

How does x264 etc get away with it?

Software patents are not valid everywhere, x264 is brought up by VideoLAN in France, where software patents don't apply like in the rest of the EU.

So if a company in france writes open source software that infringes on a US software patent, and a company in the US bundles that source code in a product, is the US company liable for damages?

Re: H.266/Versatile Video Coding (VVC)

#126

Post by pedrovhb » Mon, Jul 06, 2020, 2:50 PM UTC

sp332 wrote: post #26

Unklejoe wrote: post #14

It's interesting that they are able to continue improving video compression. You'd think that it would have all been figured out by now. Is this continued improvement related to the improvement of technology? Or just coincidental? Like, why couldn't have H.266 been invented 30 years ago? Is it because the computers back in the day wouldn't have been fast enough to realistically use it? Do we have algorithms today tha…

I don't know much about H.266, but some of the advances in H.265 depended on players having enough RAM to hold a bunch of previous decoded frames, so they could be referred to by later compressed data. Newer codecs tend to have a lot more options for the encoder to tune, so they need a combination of faster CPUs and smarter heuristics to explore the space of possible encodings quickly.

I wonder if instead of heuristics, machine learning could be used to figure out the best parameters.

Re: H.266/Versatile Video Coding (VVC)

#127

Post by masklinn » Mon, Jul 06, 2020, 2:53 PM UTC

Unklejoe wrote: post #14

It's interesting that they are able to continue improving video compression. You'd think that it would have all been figured out by now. Is this continued improvement related to the improvement of technology? Or just coincidental? Like, why couldn't have H.266 been invented 30 years ago? Is it because the computers back in the day wouldn't have been fast enough to realistically use it? Do we have algorithms today tha…

> It's interesting that they are able to continue improving video compression. You'd think that it would have all been figured out by now.

Would you? AV1 was only officially released 2 years ago, h.265 7, h.264 14, …

Re: H.266/Versatile Video Coding (VVC)

#128

Post by ChrisMarshallNY » Mon, Jul 06, 2020, 2:53 PM UTC

striking wrote: post #123

ChrisMarshallNY wrote: post #117

Earlier quoted context omitted.

Yeah. It's a mess. The video patents aren't just "patent troll" patents, either. They are highly enforceable, and were registered by corporations like Ampex. I have been trying to write a simple app to stream RTSP (security cameras), and that has been a pain. I need to basically use either proprietary (paid) or GPL software to do it. Video software is not for the faint of heart. Much as I grouse about the licensing,…

You can license x264 to not have to worry about the GPL. Though that costs money and you may find yourself signing MPEG LA contracts. edit: Your project reminds me, https://github.com/arut/nginx-rtmp-module is super worth checking out and might be helpful to you.

Yeah, but it was a free app.

One of the frustrating things about implementing video software, even licensing it, is pretty much everything out there is an expression of ffmpeg, which is a really, really good system, but does have some baggage.

Re: H.266/Versatile Video Coding (VVC)

#129

Post by throwaway8941 » Mon, Jul 06, 2020, 2:53 PM UTC

mtgx wrote: post #53

DrBazza wrote: post #9

Naively hoped I'd read 'this will be released to the community under a GPL license' or similar. Instead found the words 'patent' and 'transparent licensing model'. I appreciate that it costs money and time to develop these algorithms, but when you're backed by multi-billion dollar "partners from industry including Apple, Ericsson, Intel, Huawei, Microsoft, Qualcomm, and Sony" perhaps they could swallow the costs? It…

What you're looking for is the AV1 codec, which finalized a year and a half ago and will likely see much wider adoption, simply because none of the members want to pay royalties. https://aomedia.org/av1/ https://aomedia.org/membership/members/ AV1 decoding has already been in Chrome and Firefox since at least a year ago. We're just waiting for hardware decoding and encoding support now, which should start appearing t…

>The next version of Chrome will also support the AV1-based AVIF image format this month:

FWIW, Firefox supports it already (behind the image.avif.enabled knob).

Re: H.266/Versatile Video Coding (VVC)

#130

Post by ChuckMcM » Mon, Jul 06, 2020, 2:53 PM UTC

DrBazza wrote: post #9

Naively hoped I'd read 'this will be released to the community under a GPL license' or similar. Instead found the words 'patent' and 'transparent licensing model'. I appreciate that it costs money and time to develop these algorithms, but when you're backed by multi-billion dollar "partners from industry including Apple, Ericsson, Intel, Huawei, Microsoft, Qualcomm, and Sony" perhaps they could swallow the costs? It…

Notice that the missing party here was 'Google'. These are the folks who are really competing with VP9, the royalty free codec that limited the uptake of H.265.