h265 Support

[Cryonic]

I`m thinking about the h265 just as about the USB3.1 type C connector.
Something that is a core function for the most people working in this field, avaliable as a standart for everyone (that includes direct competitors using the same technology). A universal standart, providing maximum compatibility. And i expected it to be pushed just as the type C connector (and it is not backwards compatible etc). When a single company just makes it avaliable and not only avaliable, but forces the people into it like Apple did by installing it on the macbook without any alternatives.

You know how hard it is to actually talk to Twitch or Youtube, specially about stuff that should be developed. Twitch is still using flash and is crawling behind major VOD platforms in terms of developing and implementing new stuff.
But you guys got a point - without your software (and i mean streaming software in general, not just OBS) livestream-services like Twitch & Hitbox are useless, they dont offer any useable solutions for streaming, they just provide the service.

But fine, if there is nothing planned so far, we will just sit on x264 and enjoy our fine numbers in benchmarks.
Btw i reach 49FPS in the HVEC decode benchmark (Cobra) V1.4 with the Ultra HD 2160p 25Mbps preset. If i push it a bit harder, kill all the background stuff and squeeze a bit more clock out of the CPU, i should be damn close to 60 in 4K and 1080p would be over 120fps. Encoding benchmarks also show some nice results, so i`m excited about it.
If a major platform decides to switch to h265 only, we will see a really fast progress.

 

[Bensam123]

Hmm, just writing a new protocol and making standards wont make anyone switch. I remember the guy that told us twitch would switch to webrtc in 3 months and OBS will be unusable by then. That was a year or two ago.

Probably 99% of computers are not able to encode a video with a useful resolution for streaming with the H265 codec at the moment, at least to what I read about it and saw in tests. Hardware encoders are not available yet, at least not in large numbers and even then it will still take some time. Only the newest graphics chips support hardware decoding of h265 for example and it uses a lot of CPU power. In theory, you might even be able to test h265 using the ffmpeg output of obs-mp, not totally sure.

As soon as we reach a point in which h265 can be encoded and decoded just fine by a big number of people (similar to h264 a few years ago) everyone will switch to it. Back in 2009 I streamed with vp6 at maybe 480p and it nearly ate my whole CPU.
I just checked, work on h264 started in 2001, first version was ready in 2003 and we all use it since, maybe 2008/2009 decoding wise? :D

Yeah, I understand that. I don't expect like 4k or 1080p without more then one computer working together. Do you have a link for these benchmarks? I'm pretty sure 720p isn't that far outside of the scope. It's mainly useful more to people with limited internet connections rather then people that are trying to push super HD content.

I'm not sure how cluster encoding will work, but that's not all that different either. Even if the majority of streamers can't use it, getting things going are part of the problem.

Also agree about OBS trying to push something. It may go no where and maybe this is seen as 'unneeded', but you know we'll never need more then 640k of memory. Also I think the climate for live streaming is quite a bit different compared to 2001 and right now, as are adoption of technologies that may be more future oriented.

Not saying this 'needs support now', but no one has been talking about it for the last year or so on here (or they get deleted).

Well h265 will give better quality at the same bitrate than x264. The cost: raw power needed to encode it as a streamer and the same goes for the playback.
Right now with anything going for multithreading and slapping as many cores as possible on anything, i expected this to have the highest priority here. Specially for twitch, cause they have pretty harsh bandwith restrictions - so everything that can improve the quality with the given bitrate would be welcome.
I know my rig beats 95% of all systems used for streaming, but with this kind of raw power i would love to actually test it in public, not just do some strange benchmarks (thats the only thing possible right now).

Yeah, it will be interesting to see how OCL support changes this. If the multithreading aspect scales as well as they make it out to be, we wont need hardware encoders, rather we will just need to use a GPU to do the encoding (arguably a hardware encoder). All GPUs today have OCL support of various kinds (including Nvidia). It just needs the software to power it.

You're welcome to do that, but the OBS developers will be busy with other improvements to the program that are much more important in the near-term. Developing a protocol for h265 streaming from encoder to distributor is way, way outside the scope of OBS development.

Arguably. I've been around OBS since the beginning, after everyone started working on the rewrite nothing really else has been happening here besides VCE. After reading about the rewrite, I'm still not certain why it's being done except to placate the handful of people operating on nix and OSx. It's been over a year and the rewrite still isn't as good as the original version of the program. The OBS rewrite in my opinion is one huge boondoggle. Multiplatform support always sounds good till you put into context who will actually be using it. It's like nix gaming. Great idea, but never going to go anywhere.

Not saying that to be mean, just the way it is and people can't use the rewrite for a production machine right now.

Curiously what sort of near-term developments are happening in OBS besides the rewrite, which is probably more of a six month outlook in it's current form? I haven't heard really anything...

 

[dodgepong]

The original code base was a mess and had a lot of design flaws that prevented several desired improvements. When Jim first wrote it, he had no idea how much people would use it, and the ways in which they would want to use it. Building it again from the ground up helped to fix a lot of those mistakes, having a much better idea of how people want to use the program.

Curiously what sort of near-term developments are happening in OBS besides the rewrite, which is probably more of a six month outlook in it's current form? I haven't heard really anything...

Development of the original OBS has halted, as all work is being done on the rewrite. Off the top of my head, these are things that the OBS team has been working on for the rewrite, and will be working on.

Features that have been added to the rewrite that were never in OBS1:

* Mac and Linux support
* Encoded capture device support (HD PVR, Roxio, etc.)
* Filters (color correction, mask, delay, scroll, etc.)
* FFmpeg output
* Mutli-track audio recording
* Native .MOV, .MKV, and .TS output in addition to .MP4 and .FLV
* Visual Theme support
* OpenGL renderer
* Different settings for streaming vs. local recording
* AVerMedia hardware encoder support
* Native BlackMagic device support

Things that still need to be added to the rewrite for parity with OBS1:

* Deinterlacing
* Quicksync and NVENC support
* Image slideshow plugin
* Ability to delay stream
* Ability to select network interface
* FPS indicator on status bar

A few things that we want to add to OBS Multiplatform eventually, in no particular order:

* Ability to preview and edit non-live scenes
* Plugin manager
* Transitions
* Multiple stream outputs
* Adjustable UI
* Better source management tools
* VCE support

If you think the rewrite is a "boondoggle", then you haven't been paying attention.

Multiplatform support always sounds good till you put into context who will actually be using it. It's like nix gaming. Great idea, but never going to go anywhere.

Is it wrong of me to find this utterly hilarious in a thread where you're asking for support of what is currently an extremely niche and arguably non-existent feature to be implemented?

 

[Jack0r]

(or they get deleted)

The only things we delete are: spam, advertisement, videos moaning at the admins.
If people start to act badly we normally lock the threads and if necessary block them from posting further. But why should we delete any general discussions?
https://obsproject.com/forum/threads/x-265.1231/
https://obsproject.com/forum/threads/will-obs-look-into-using-x265.25158/ (this one is from february this year btw)
https://obsproject.com/forum/threads/hvec-and-vp9-when-will-it-be-implemented.22325/
https://obsproject.com/forum/threads/h-265-support.21965/

Thats what a quick search showed in less than 10 seconds :)

 

[dodgepong]

And let me be clear, to keep the discussion on-topic: it is very possible (and probably likely) that support for an h265 encoder will be added to OBS Multiplatform, whether that's native x265 or a simple FFmpeg update. In fact, if there's an FFmpeg build floating around out there with h265 encoder support already, you could build OBS with that version of FFmpeg yourself and then you have it.

I've just been saying that h265 livestreaming will not be possible for a while.

 

[Cryonic]

h265 is already possible. And it should look amazing with low bitrates.
The other thing is - massive, mainstream livestreaming for people with mid-end hardware, will be almost impossible to achieve. And it will also be unwatchable for people with low-end hardware.
I personally dont care, having the i7 5820K OC made me ignore so much stuff that will waste a lot of time of other people.
And yes i`m a graphic snob. I want my stream to look as good as possible with the given bandwith limit.
h265 is almost ready to take over anything, just a small push from a big company and everyone will jump on the bandwagon.
Sadly we dont see that impressive rush for innovations in out department, we are still stuck with outdated stuff, using low bitrates and low resolutions.

 

[Krazy]

You guys aren't getting it. Support/availability in OBS is not the problem. Support by streaming platforms is the problem. As with everything else tech/internet related these days, that support will ultimately be driven by what the majority of mobile devices are capable of.

So, if you *really* want to push for h265 livestreaming to be widely adopted, then beg smartphone and tablet manufacturers to implement h265 hardware decoding in their devices.

The other facet of this entire discussion is the assumption that h265 will even be the method of choice for the future of livestreaming. Dodgepong briefly mentioned it, but Google is working on their own implementation of a next generation codec. It's very likely they won't use h265 at all. If that's the case, we have the potential for a codec war, as I highly doubt most device manufacturers will want to support both.

In my view, it will be 2 years minimum before we start to see true widespread adoption of the next generation codecs. Probably another 2 years after that before you can confidently say that the majority of people are using h265/VP9 instead of h264.

Just remember to temper expectations, and realize that codec/internet protocol changes are *never* fast. I mean, good lord, html5 is still kind of a mess and that was the cutting edge latest and greatest thing that was going to transform the internet...several years ago now.

 

[Cryonic]

Well most of us know what it is to be an early adopter, specially in the hardware world. Things are not easy there, but they dont take years - weeks, in worst cases a couple of months to actually implement and fix the new driver etc.
And i hate the idea of working around mobile devices, consoles, handhelds, smart TVs and whatever else in a PC-related field where PC is dominating (streaming, not watching it!). But i understand that the time where PC was the leading plattform for many things and was not forced to deal with ports, low end hardware and poor optimizaton for multiplatform-software...
And a codec war is the worst that can happen to us.
And why the heck would anyone implement a h265 hardware encoder in smartphones and tablets right now? Everyone is using the x264 now etc, why the heck would they push it by adding something that is not cheap, when others can push it for free?
But i see Google here with good chances to force people to move on to a new standart. They have manpower, money and everything else to make it.
But still i want the best encoder to be implemented just like new hardware standart - they are here, everyone agrees to use it and everyone is happy. And it happens really fast compared to the codec and internet protocol upgrades.

 

[Osiris]

And let me be clear, to keep the discussion on-topic: it is very possible (and probably likely) that support for an h265 encoder will be added to OBS Multiplatform, whether that's native x265 or a simple FFmpeg update. In fact, if there's an FFmpeg build floating around out there with h265 encoder support already, you could build OBS with that version of FFmpeg yourself and then you have it.

I've just been saying that h265 livestreaming will not be possible for a while.

Yep, using the shared build from http://ffmpeg.zeranoe.com/builds/ and replacing the ffmpeg dlls included with obs with the ones from the site. Doing this will enable you to record in h265 with the custom ffmpeg output. Ofcourse not streaming it, since no service supports it.

Not saying this 'needs support now', but no one has been talking about it for the last year or so on here (or they get deleted).

Please don't accuse us of deleting discussion about h265, since that is simply not true.

 

[Bensam123]

Typo as it should've been 'got' instead of 'get', as in I assume there were more but they were deleted.

You guys aren't getting it. Support/availability in OBS is not the problem. Support by streaming platforms is the problem. As with everything else tech/internet related these days, that support will ultimately be driven by what the majority of mobile devices are capable of.

So, if you *really* want to push for h265 livestreaming to be widely adopted, then beg smartphone and tablet manufacturers to implement h265 hardware decoding in their devices.

I disagree. Hardware decoders didn't come before 1080p or 720p on h264. Matter of a fact, h264 was out quite a few years before h264 decoders came around for laptops to help reduce utilization.

If you don't talk about it and start some buzz, people don't actually know how much interest there is for something. Twitch generally only caters to the big casters and doesn't even know the little guys exist. A lot of people are stuck with shitty internet though and could really benefit from such a adoption in whatever form it takes. Even if it's not a complete solution.

This isn't the same thing as suggesting a complete replacement of h264 either. Often times multiple standards exist at the same time as things are phased out.

Features that have been added to the rewrite that were never in OBS1:

Is it wrong of me to find this utterly hilarious in a thread where you're asking for support of what is currently an extremely niche and arguably non-existent feature to be implemented?

I don't think there are many people that find h265 hilarious. How many of the aforementioned features are there to support nix/osx? How many could be supported in the current version too? Seems like some of those features are simply there because development on OBS1 was halted.

I read the posts as well. I never said anything because Jim can do what he wants. Though Nix and OSx support are very much just something a handful of very vocal people want.

About the best thing I could see from that is mobile support for OBS to stream from your smartphone, but you guys aren't even designing or talking about that.

 

[Krazy]

I'm going to pass over the rest of your post because it's already been hashed out several times, but this stood out to me:

Though Nix and OSx support are very much just something a handful of very vocal people want.

You are seriously underestimating the amount of people that have actually been installing and using OSX and Linux versions of OBS MP. It's more than just a "handful".

 

[Osiris]

Typo as it should've been 'got' instead of 'get', as in I assume there were more but they were deleted.

I disagree. Hardware decoders didn't come before 1080p or 720p on h264. Matter of a fact, h264 was out quite a few years before h264 decoders came around for laptops to help reduce utilization.

If you don't talk about it and start some buzz, people don't actually know how much interest there is for something. Twitch generally only caters to the big casters and doesn't even know the little guys exist. A lot of people are stuck with shitty internet though and could really benefit from such a adoption in whatever form it takes. Even if it's not a complete solution.

This isn't the same thing as suggesting a complete replacement of h264 either. Often times multiple standards exist at the same time as things are phased out.

Talking about and "creating buzz" by us is not gonna change a thing, creating buzz for something very hard to implement at the moment is a bit useless. But you don't seem to be the getting the technical problems that still exist to stream using h265.

I don't think there are many people that find h265 hilarious. How many of the aforementioned features are there to support nix/osx? How many could be supported in the current version too? Seems like some of those features are simply there because development on OBS1 was halted.

As has been stated before, the OBS1 code was basically a mess preventing certain new features, so it was not an option to continue developing it.

I read the posts as well. I never said anything because Jim can do what he wants. Though Nix and OSx support are very much just something a handful of very vocal people want.

You are underestimating it, it is not a handful, certainly not for OSX

 

[Cryonic]

There is no technical problem. Just the need for powerful hardware under the hood to drive it.
A lot of people, specially in rich countries with high income, can afford a stupid fast PC or a dual PC setup, but are limited by shitty assymetrical internet connections.
h265 is the best solution so far to work around this issue, and many people think when the support for it will be implemented in the software, streaming services will start using it. Hardware decoders are nice to have, but they are still far away.
Still waiting for someone to push h265 hard, forcing it everywhere it goes.

 

[Osiris]

There are technical problems, read what was said earlier.
Just software supporting it is not gonna push it, when even rtmp doesn't support h265. Switching to something else is gonna take a while.

 

[Cryonic]

Yeah some small problems are usual for new tech, there is no way around it.
But somebody has to start implementing it and i would really love to see the client development jumping in first and pushing a new technology. I expect the same like in the hardware world, bring out something new (like HBM, what AMD did), even if it is far from perfect and cant beat the old one in most cases - pushing it out will help in the near future and giving a small advantage over other products on the market.
You guys push nix/OSX support, multichannel audio recording etc, but the most important part, the codec behind all this stuff, there is a great option producing amazing results if powered by a highend rig (and they are common, just like OSX!).
Streaming was always limited by the bandwith, this is why i expect to see the h265 here in the first place - this is where it is needed the most. With anything else reaching amazing speed (100mbit/s and higher mobile connections, gigabit internet, SSD becomes cheap, flash drives like USB sticks and SD-cards pushing the limits etc), streaming is falling behind in quality compared to other media, and changing the codec is a great way to maintain the low bandwith limits (which by the way are stupid, but its twitch only twitching about it, instead giving us something like 10mbit/s limit) and produce better quality.

 

[Osiris]

nix/OSX support and multichannel audio recording are more important then h265 support at the moment. No streaming service is making any effort to support h265 streaming, mainly because of the hardware requirements for encoding and decoding. Sure, theoretically obs can be made to use the h265 codec for streaming, but it's pretty useless, since you can't stream it anywhere. So it's better to focus on other things then h265 right now, until streaming services are making an effort to support h265
Oh and OBS adding h265 support is not going to make streaming services try harder.

 

[Cryonic]

I expected OBS to have higher influence, specially because its so popular and without OBS and Xsplit we would still live in the stone age. Usually its a great thing when the software covers over 50% of the current market - people start to listen and deal with it. But whatever, i have huge problems right now with the old OBS and its not getting better with updates.

 

[Osiris]

If only the problem was just software support, but the fact is that there are more problems that need to be solved before streaming services can really adopt h265. When those problems become less of an issue then adding h265 support would make sense, until then the development focus is on feature parity with the old OBS and new features (and of course bugfixing).

 

[dodgepong]

This conversation is kind of circling itself at this point, so here's the bottom line:

* h265 encoding will likely be added to OBS MP, either by x265 or FFmpeg
* If you want h265 encoding right now, you can compile OBS MP with a fork of FFmpeg that already has h265 encoding implemented
* Streaming with h265 is currently an impossibility
* OBS developers currently have much higher priorities related to OBS development over creating a new streaming protocol that can use h265

I understand that h265 is "the next big thing", but OBS MP needs to have many more "fundamental" features covered before anyone can even think of inventing a way to stream h265 video.

There's not really any more to discuss about it. I don't think there's anything anyone can say that will convince the developers that h265 streaming should be a priority right now.

 

[AllanDavidson]

Just bumping this thread since it's Feb 2016 and we already have real-time H265 ASIC encoders on consumer hardware like Nvidia Maxwell and Intel Skylake and AMD GCN.