xiphmont: (Default)
[personal profile] xiphmont

Had this happened next week, I'd have thought it was an April Fools' joke.

Out of nowhere, a new patent licensing group just announced it has formed a second, competing patent pool for HEVC that is independent of MPEG LA. And they apparently haven't decided what their pricing will be... maybe they'll have a fee structure ready in a few months.

Video on the Net (and let's be clear-- video's future is the Net) already suffers endless technology licensing problems. And the industry's solution is apparently even more licensing.

In case you've been living in a cave, Google has been trying to establish VP9 as a royalty- and strings-free alternative (new version release candidate just out this week!), and NetVC, our own next-next-generation royalty-free video codec, was just conditionally approved as an IETF working group on Tuesday and we'll be submitting our Daala codec as an input to the standardization process. The biggest practical question surrounding both efforts is 'how can you possibly keep up with the MPEG behemoth'?

Apparently all we have to do is stand back and let the dominant players commit suicide while they dance around Schroedinger's Cash Box.

Date: 2015-03-27 08:34 pm (UTC)
jered: (Default)
From: [personal profile] jered
Thankfully this patent insanity didn't exist around the time of K&R, or we'd all be arguing about patent pools and licensing requirements for C compilers and the software industry would be stuck in the dark ages.

Best of luck, Monty; you've been working on this for a loooooooong time!

AT&T

Date: 2015-03-28 01:53 am (UTC)
From: (Anonymous)
If you didn't study history of AT&T
AT&T was forced by US Government decree to license
all its patents as a condition of Monopoly.

This is why Transistor was so quickly taken up by Japanese
companies like Sony to make a portable radio.

Unix and C is in the same situation because AT&T created it.
Not some hippies.

Re: AT&T

Date: 2015-03-29 02:23 pm (UTC)
From: [identity profile] sauergeek.livejournal.com
While you seem to have read something of the history of AT&T, your reading of it appears to have been cursory, and thus your analysis is badly flawed.

Transistor invented: 1947. Licensing for transistors opened in general: 1952. AT&T, as part of a settlement, required to license all its patents to all interested parties: 1956. Japanese companies -- and who knows how many others around the world -- had licensed the transistor four years before the court settlement mandated licensing.

Unix: first release, 1973. Licensed, and still under ownership and licensing, which is at least in part claimed by the Company That Will Not Die No Matter How Much We Want It To, SCO. That license does not extend to rewritten Unix-alikes, so the hippies have been instrumental in spreading Unix in the form of BSD, Linux, and other license-free non-AT&T Unix relatives.

C: first release, 1978. So far as I am able to tell, the language itself was never patented or placed under license, which permitted anyone -- including those hippies at GNU -- to write a compiler.

WTF? No! WAFO!

Date: 2015-03-27 11:59 pm (UTC)
From: (Anonymous)
I don't see this as a WTF moment, I see it as a WAFO (What A Fantastic Opportunity) moment. This is the case study in the benefits of royalty free codecs. I don't think standing back will be enough. Now's the ideal time for VP9 to gain ground over HEVC. I just hope Google cares enough to help it do so.

The WebM project website seems to be suffering quite a lot of bit rot. The most recent blog post is dated December 15, 2014 but the post talks about VP8, not VP9, and it begins with the line "Since the WebM project was open-sourced just a week ago.." Something has gone wrong there.

This sort of neglect undermines the perception of VP9 because it makes it look as though Google doesn't take VP9 seriously. Maybe it really is the case that they don't take it seriously.

Re: WTF? No! WAFO!

Date: 2015-03-28 04:25 pm (UTC)
From: (Anonymous)
I don't think that there is any doubt that the "it's magic" perception of video codecs has worn off. Video is video - the price that people are going to expect to pay is zero.

VP9 support is mandatory in Android 4.4 (although sadly so is HEVC) and it is used by default on YouTube for Firefox and Chrome, so I think that Google's commitment to the codec is pretty clear. They could have just threatened to support VP9 to get a discount on HEVC licensing, but they have completely followed through.

The best the HEVC crowd can say is "you have no choice but to have HEVC on your device". That is their only survival strategy, because HEVC offers the consumer **NOTHING** over VP9.

It'll be interesting if Microsoft and Apple decide that VP9 is worth fighting against (TBH - I'm not sure that they actually care much at this point).

Re: WTF? No! WAFO!

Date: 2015-03-29 12:30 am (UTC)
From: (Anonymous)
VP9 support is mandatory in Android 4.4 (although sadly so is HEVC) and it is used by default on YouTube for Firefox and Chrome

This doesn't seem to be the case for Firefox 37 beta on OS X. I tried a few videos on YouTube. The best available resolution for all was 720p and in the "Stats for Nerds" window (right click on the video to choose it from the context menu) it said AVC (H.264) was used. The YouTube HTML5 (https://www.youtube.com/html5) status page tells me MSE isn't available, which would explain the lack of 1080p resolution. Maybe MSE and/or VP9 is available for HTML5 video in Firefox on some platforms, but it isn't seem to be there yet for Firefox 37 beta on OS X.

Re: WTF? No! WAFO!

Date: 2015-03-30 10:53 pm (UTC)
From: (Anonymous)
The Firefox devs are still working the kinks out of the Media Source Extensions. If you don't mind running beta code, you can enable them for YouTube via about:config.

Re: WTF? No! WAFO!

Date: 2015-03-31 01:48 am (UTC)
From: (Anonymous)
Firefox doesn't support MSE with WebM yet, even with all prefs on, so no VP9 yet.

Re: WTF? No! WAFO!

Date: 2015-04-03 03:44 am (UTC)
From: (Anonymous)
I haven't changed anything in about:config, but I updated to Firefox 38 beta on OS X and the YouTube HTML5 status page now indicates that MSE with H.264 is available. However, MSE with VP9 still isn't available by default.

Re: WTF? No! WAFO!

Date: 2015-03-29 10:34 am (UTC)
From: (Anonymous)
They offer a good encoder (libx265), good tools (ffmpeg), good standard way of encoding stuff: mpeg ts and mp4.
There is _NO_ way to do live video with vp9 and the encoder is too fucking slow!
I would have implemented a trial live ecoding long ago if it worked.

Re: WTF? No! WAFO!

Date: 2015-04-05 05:00 am (UTC)
From: (Anonymous)
You're not supposed to be using libx265 or ffmpeg without getting a patent license from the MPEG group, if you build any significant commercial thing without one they will sue you to oblivion.

Re: WTF? No! WAFO!

Date: 2015-03-29 02:38 pm (UTC)
From: (Anonymous)
The other part of the equation is the lack of a publicly available hardware implementation of either VP8 or VP9- and an unwillingness to talk with honestly interested parties that would've been WILLING to pay a royalty on the proprietary implementation of the same that Google had done.

"Fail" doesn't really begin to describe this whole thing right at the moment. They could have the keys to the kingdom if they could find a means to have DRM that even if it's Open Source, it wouldn't matter. Providing the superlative open answers means that they would have won- even with Chrome.

Re: WTF? No! WAFO!

Date: 2015-03-30 06:28 am (UTC)
From: (Anonymous)
I'm not sure what you mean. The WebM project website has hardware designs for VP9 encoding and decoding available for semiconductor companies to use. The WebM wiki lists a number of manufacturers who already support hardware VP9 decoding on their SoCs.

Encrypted Media Extensions is probably as close as its every going to get to open source DRM. I suppose people like Netflix will (or are) use EME for their DRM, maybe in combination with the WebCrypto API.

Re: WTF? No! WAFO!

Date: 2015-03-29 04:33 pm (UTC)
From: [identity profile] mzso.livejournal.com
For what? To popularize handicapped-by-default codecs because all the good stuff's patented?
This idiocy won't stop until software patents are abolished.
Which won't be anytime soon, because FOSS zealots don't have spending money to lobby (aka bribe) for it. Apparently Google doesn't have enough pull for this either. (Or isn't actually interested)

PS:
It's really irritating how much time I have to waste just to comment with this livejournal nonsense. This time tha captcha wouldn't load in my FF profile...(domain / ckey mismatch )

Biggest threat to VP9 adoption is not h265

Date: 2015-03-28 06:31 pm (UTC)
From: (Anonymous)
It is a good, easy to use GUI encoder such as Handbrake for h265's.

Re: Biggest threat to VP9 adoption is not h265

Date: 2015-03-28 07:14 pm (UTC)
From: (Anonymous)
It seems ffmpeg already supports vp9 : http://superuser.com/questions/705579/convert-video-with-vp9-codec-using-ffmpeg

Meanwhile Daala (NetVC) already has OSX support and ffmpeg can handle the binary stream : https://wiki.xiph.org/Daala_Quickstart

Your GUIs just need to update.

Re: Biggest threat to VP9 adoption is not h265

Date: 2015-03-29 05:36 pm (UTC)
From: (Anonymous)
I've read about encoding speed issues with vp9 on the handbrake & doom9 forums. I hope devs are keeping a close eye on these forums for feedback sake. If these crowds are convinced, there's a chance in the consumer market for vp9 to thrive. Handbrake is ready to pull the trigger on adding vp9 as soon as the speed issues are sorted https://forum.handbrake.fr/viewtopic.php?f=26&t=27343

Re: Biggest threat to VP9 adoption is not h265

Date: 2015-03-30 08:32 pm (UTC)
From: (Anonymous)
Google doesn't give a damn. It is how many months/years since VP9 release? And just few weeks ago, they implemented the single most basic (and crap) method of multithreading to the encoders. Speaking about tile threads, which really, could have been bolted on the encoder in one day all teh way back when it was first released. No, they simple don't care, because youtube can encode everything in 1-thread but with many instances in paralel. Rest of users can go forth and multiply, it seems :)

Side note: this might be a blow to HEVC popularisation, but it also serves as FUD to those free and independent codecs too. If rogue patent pool can happen for standardised codec, why couldn't it happen for a "me too" one from a single party that had less resources to check all the potential IP risks?

Re: Biggest threat to VP9 adoption is not h265

Date: 2015-03-30 09:31 pm (UTC)
From: (Anonymous)
If rogue patent pool can happen for standardised codec, why couldn't it happen for a "me too" one from a single party that had less resources to check all the potential IP risks?

But if it can happen to all codecs, then what's the difference? You might as well go with the codec without any obvious licensing problems today (VP9) as opposed to the codec which two competing patent pools and an uncertain licensing future (HEVC).

Re: Biggest threat to VP9 adoption is not h265

Date: 2015-03-31 10:25 pm (UTC)
From: (Anonymous)
Well, there are the minor details like HEVC having multiple encoder vendors to choose from (including x265 which should be the best choice ATM). It also seems like being an overally better designed and better performing format :)

The situation with encoder speed, quality and choice may of course improve over time, but libvpx/VP9 has a lot of catching up before it even is on par with HEVC camp... and based on the experience with VP8, it will probably only keep getting more and more behind.

Looks like it is "technical merits" versus "no royalty to use", as before.

Ogg

Date: 2015-03-28 07:47 pm (UTC)
From: (Anonymous)
So what's wrong with Ogg?

Re: Ogg

Date: 2015-03-28 11:45 pm (UTC)
From: (Anonymous)
Ogg is a container--i.e. framing--format. It's not a codec.

Re: Ogg

Date: 2015-03-29 01:29 am (UTC)
From: [identity profile] chrisslicks.livejournal.com
Ogg is a container format, not a video codec.

Re: Ogg

Date: 2015-03-29 04:51 am (UTC)
From: (Anonymous)
There are a few Oggs to consider.

Yes, as others have pointed out Ogg itself is a container.

Ogg Theora is a video codec, but it is old, very old by codec standards, and it really isn't up to handling today's demands.

NetVC/Daala, which is too new, and isn't widely supported yet, but over the next couple of years looks to be exceptional.

In the relatively near term, Google has indicated that VP10 should arrive rather soon for a standard.

Basically, HEVC is already approaching End-Of-Life, and it never really got anywhere.

VP10

Date: 2015-03-29 01:40 pm (UTC)
From: (Anonymous)

Google has indicated that VP10 should arrive rather soon for a standard.

Ouch - I hadn't spotted that news. They could at least wait for the hardware manufacturers to catch up a bit... There are already people complaining that CPU usage for YouTube is significantly higher for VP9.

If my Googling is correct, on desktop AMD and NVidia don't have any VP9 acceleration, and even Intel only does selected bits of it in silicon.

Re: Ogg

Date: 2015-03-29 05:21 pm (UTC)
From: (Anonymous)
I remember the VP10 is coming soon thing to be confuted.
They're focusing on VP9 and only will make a VP10 if they can achieve meaningful improvement.

Re: Ogg

Date: 2015-03-29 05:22 pm (UTC)
From: [identity profile] mzso.livejournal.com
I remember the VP10 is coming soon thing to be confuted.
They're focusing on VP9 and only will make a VP10 if they can achieve meaningful improvement.

Date: 2015-03-29 06:32 pm (UTC)
From: (Anonymous)
Good thing in my country (Mexico) software patents are explicitly illegal.

How are people already shipping HEVC?

Date: 2015-03-30 03:45 am (UTC)
From: (Anonymous)
What I don't understand is how there are already shipping products claiming HEVC support. How is this possible given the murky licensing situation?

Profile

xiphmont: (Default)
xiphmont

Most Popular Tags