It’s Day 0 of E3 2015. This is the time when all the giants of popular gaming make their big announcements, competing for your attention and future gaming dollars. Today is also a big day for YouTube, which doesn’t make games, but will soon be introducing a dedicated YouTube Gaming service. It too will be competing for the attention of millions.
The goals of YouTube Gaming are as grand as YouTube itself. Google wants its new website and app to become “the biggest community of gamers on the web” and the destination for live-streamed game video, whether it comes from professional tournaments or amateurs playing just for fun. If that sounds exactly like Twitch, that’s because it is. Having lost out to Amazon in the pursuit to acquire Twitch last summer, Google has spent the past year building up its own alternative, and that’s what we have to look forward to in the coming weeks.
YouTube is well-positioned to compete with Twitch, since most streamers upload the VODs to YouTube anyway. Why not have it all in one place?
In any event, yet another case of competition breeding product improvements.
An alternative perspective is that Google sees all activity, all developments of new innovative spaces and connections, that it cannot monitor and mine for data, as a threat. Its response is to either to use its market weight to try to pry open or route round any blocks to its ability to mine data and if that fails to use its financial resources to undermine and ultimately destroy areas of activity it cannot access by creating an alternatives. All large tech companies see some of the activities of other large tech companies and the innovations that bubble up from start ups as a threat. Google is unusual because it see all activities by all other companies and all innovation as a potential threat unless it is given guaranteed access for data mining.
The funny thing is that you can swap out twitch’s owners ( amazon) for google, and its still true.
Yes, that must be why everyone who runs a private intranet constantly receives demands for access from Google… hmm, I’m getting a sense of deja vu, I think I may have already made that crack before. Oh wait, I did – as a reply when you posted that exact same “argument” before. I guess that’s your replacement for the “OMG ADVERTISING COMPANY” talking-points.
I have already commented on this elsewhere, but I cannot find anything official on whether Youtube Gaming (YG) will support input-streams that use either VP9 or HEVC. The advantages from those codecs are obvious: both offer the same quality as H.264 at lower bitrates or much higher-quality at the same bitrate. For live-streaming video-quality definitely matters, but also many people are bandwidth-starved so being able to use lower bitrates would also be very welcome.
Hardware-based encoding of H.265/HEVC is possible already now with GTX 960(m) that can do at least 1080p@60FPS in realtime — both main and high profile — and now that NVIDIA already has hardware that can do it they have zero reason to suddenly drop support in the new “Pascal” architecture they’re releasing next year and as such streaming using HEVC becomes a real, tangible possibility hardware-wise. Software-side? Not so much.
Unless YG implements support for HEVC there’s still not going to be any actual service capable of using the codec; the last time I asked Twitch’s support about it, for example, they said they aren’t even considering adding support. Then there’s the actual capture-software that’s a problem, too: even NVIDIA’s own ShadowPlay is unable to record in HEVC yet, Open Broadcaster Software (OBS) devs have no plans on supporting HEVC whatsoever, I haven’t heard of XSplit supporting it either nor am I aware of anything else.
VP9 has things even worse: while YG probably would support VP9-encoded streams there is no consumer-accessible hardware whatsoever for PCs capable of doing encoding in VP9 in realtime nor is there anyone seemingly even planning to add support for it. I certainly have not heard any plans from e.g. NVIDIA for support for it. Intel’s QuickSync can do VP8 starting with Haswell, so they may add VP9 in the future, but that’s still in the limbo. No H/W-suppport and no software-support kind of creates a chicken-and-egg problem for VP9 :/
The reason mind be because of the bad news: encoding.
It takes at least 3.5 times as much processing power to encoding these videos.
That is a lot. That is where the hardware comes in of course.
Playback/decoding processing power is also higher something 15% – not so great on mobile devices.
Now I don’t know the numbers, but supposedly mobile devices biggest drains on the batteries are the screen and wireless communication.
So half the bandwidth with 15% more processing power needed for decoding. Does that mean these newer codecs are more effienct for playback on mobile devices ? Has anyone done any testing on that ?
If you were doing it in software, sure, but many ARM SoCs already support decoding both HEVC and VP9 in hardware. Even cheap Chinaman-phones do. Some of them can do encoding, too. It’s the PC-landscape and software and any associated services that’re lagging behind.
Sorry, maybe I wasn’t clear. Maybe I should have said:
“not so great for mobile devices.”
Even if you do it in hardware, it will use more power.
Maybe it is even the same increase in power usage.
Obviously with hardware support is much more efficient, maybe even orders of magnitude. So it might not matter as much.
Twitch still can’t make an HTML5 player. Too bad for them.
Twitch uses open protocols like irc and you can stream with vlc or mplayer2 to get the non-flash version and better buffering.
An irc client and a decent mediaplayer beats anything that html5 provides. Especially if the comment system on youtube streams will still be that pathetic garbage excuse of a thing called google+
Multiple times my twitch stream went down, janky and poor quality. Most frustrating my connection went down just as EA was announcing and showing SW Battlefront gameplay footage. I switched to YouTube and it came up perfectly.
(My connection is 76MB)