Google Wants To Take On Dolby With New Open Media Formats (protocol.com) 56
An anonymous reader quotes a report from Protocol: Google is gunning for Dolby Atmos and Dolby Vision: The company is looking to introduce two new media formats to offer HDR video and 3D audio under a new consumer-recognizable brand without the licensing fees hardware manufacturers currently have to pay Dolby. Google shared plans for the media formats, which are internally known as Project Caviar, at a closed-door event with hardware manufacturers earlier this year. In a video of the presentation that was leaked to Protocol, group product manager Roshan Baliga describes the goal of the project as building "a healthier, broader ecosystem" for premium media experiences. The company's primary focus for Project Caviar is YouTube, which does not currently support Dolby Atmos or Dolby Vision. However, Google also aims to bring other industry players on board, including device manufacturers and service providers. This makes Project Caviar one of Google's most ambitious pushes for open media formats since the company began working on royalty-free video codecs over a decade ago.
Google's open media efforts have until now primarily focused on the development of codecs. The company acquired video codec maker On2 in 2009 to open source some of its technology; it has also played a significant role in the foundation of the Alliance for Open Media, an industry consortium that is overseeing the royalty-free AV1 video codec. Project Caviar is different from those efforts in that it is not another codec. Instead, the project focuses on 3D audio and HDR video formats that make use of existing codecs but allow for more rich and immersive media playback experiences, much like Dolby Atmos and Dolby Vision do. Baliga didn't mention Dolby by name during his presentation, but he still made it abundantly clear that the company was looking to establish alternatives to the Atmos and Vision formats. "We realized that there are premium media experiences where there aren't any great royalty-free solutions," he said, adding that the licensing costs for premium HDR video and 3D audio "can hurt manufacturers and consumers."
Dolby makes most of its money through licensing fees from hardware manufacturers. The company charges TV manufacturers $2 to $3 to license Dolby Vision, according to its Cloud Media Solutions SVP Giles Baker. Dolby hasn't publicly disclosed licensing fees for Atmos; it charges consumers who want to add immersive audio to their Xbox consoles $15 per license, but the fee hardware manufacturers have to pay is said to be significantly lower. Still, in an industry that long has struggled with razor-thin margins, every extra dollar matters. That's especially true because Dolby already charges virtually all device makers a licensing fee for its legacy audio codecs. A manufacturer of streaming boxes that wholesale for $50 has to pay around $2 per unit for Dolby Vision and Dolby Digital, according to a document an industry insider shared with Protocol. "For lower-cost living room devices, the cost may be prohibitive," Baliga said during his presentation.
Google's open media efforts have until now primarily focused on the development of codecs. The company acquired video codec maker On2 in 2009 to open source some of its technology; it has also played a significant role in the foundation of the Alliance for Open Media, an industry consortium that is overseeing the royalty-free AV1 video codec. Project Caviar is different from those efforts in that it is not another codec. Instead, the project focuses on 3D audio and HDR video formats that make use of existing codecs but allow for more rich and immersive media playback experiences, much like Dolby Atmos and Dolby Vision do. Baliga didn't mention Dolby by name during his presentation, but he still made it abundantly clear that the company was looking to establish alternatives to the Atmos and Vision formats. "We realized that there are premium media experiences where there aren't any great royalty-free solutions," he said, adding that the licensing costs for premium HDR video and 3D audio "can hurt manufacturers and consumers."
Dolby makes most of its money through licensing fees from hardware manufacturers. The company charges TV manufacturers $2 to $3 to license Dolby Vision, according to its Cloud Media Solutions SVP Giles Baker. Dolby hasn't publicly disclosed licensing fees for Atmos; it charges consumers who want to add immersive audio to their Xbox consoles $15 per license, but the fee hardware manufacturers have to pay is said to be significantly lower. Still, in an industry that long has struggled with razor-thin margins, every extra dollar matters. That's especially true because Dolby already charges virtually all device makers a licensing fee for its legacy audio codecs. A manufacturer of streaming boxes that wholesale for $50 has to pay around $2 per unit for Dolby Vision and Dolby Digital, according to a document an industry insider shared with Protocol. "For lower-cost living room devices, the cost may be prohibitive," Baliga said during his presentation.
what's the endgame for Google? (Score:3, Interesting)
Re: (Score:2)
I assume it's related to what they're trying to turn YouTube into.
Re: (Score:2)
Youtube and chromecast.
I'm no fan of the Dolby licensing machine, but I don't see Google being able to make a dent or offering any real value since all the gear out there today already supports the legacy Dolby standards. I certainly wouldn't buy a new surround processor or AV receiver just so I could utilize that $50 chromecast dongle (unless it also supported the Dolby standards).
Best,
Re:what's the endgame for Google? (Score:4, Insightful)
Re: (Score:2)
Re: (Score:1)
AV receivers are usually happy with uncompressed multichannel PCM via HDMI, so I don't see that side as being an issue. The need for them to process AC3, DTS, etc is limited to the rapidly disappearing TOSLink/SPDIF standards. So there should be no need for you to buy a new receiver unless I'm missing something. As long as the media player, be that a Chromecast, Roku, games console, etc, has the decoder built in there's no issue. Presumably if we're still using TOSLink/SPDIF in 2025, there'll be hardware to convert SuperGoogleCodec into AC3 anyway, just as there was built into every new player when Blu-ray and HD DVD were introduced.
As far as this goes, we'll see how it goes. AC3 is presumably either out of patent or rapidly approaching it (according to Wikipedia it became patent free in 2017, but I wonder if that's true internationally - and how is it it became patent free at almost the same time as the vastly inferior MP3?) but from my point of view I've never been that impressed with the quality of audio delivered via it.
Maybe Google's codec will be higher quality and/or use less bandwidth. If so, that's a win win even if it's limited to Chromecasts connected via HDMI directly to audio receivers today.
While you're right about converting various audio streams to PCM is very viable, the same can't be said for stuff like HDR. HDR, HDR10, HDR10+, Dolby Vision...there's nothing out there that converts/handles them all. Personally, Dolby's stance on strict licensing and control of Dolby Vision is the main reason I've avoided it. I put all my media on my Plex, played through a Vero 4K+. There is no Dolby Vision support for this setup, hence I stick to the other formats.
Re: (Score:3)
Probably just driving more people to Chrome.
good (Score:2)
frankly the more high quality codecs that are free to implement and ship the better
literally no downside to shipping it in chromium etc
if google pays less and so do others all the better
frankly I sick of low quality stereo because the excuse not everyone can have 3D and we have to pay for it...
demand better sound
Re: good (Score:1)
Re: (Score:2)
Indeed. And that is very likely what they are planning.
Re: what's the endgame for Google? (Score:4, Insightful)
The emdgame is a codec with a tracker.
Re: (Score:3)
Not it isn't. Google already developed the VP9 and AV1 video codes, and they don't have built in tracking.
Re: (Score:2)
Yet.
Re: (Score:2)
Re: what's the endgame for Google? (Score:2)
Okay, FUD.
Re: (Score:2)
what is Google's endgame? Just cost saving?
Probably. Every bit they can avoid sending over Youtube is a $$$ saving.
People aren't going to support these codecs if they're a closed standard so it's in google's interest to get as many people on board as possible.
Re: (Score:2)
Google has been too cheap at this point to add Dolby Atmos or even Dolby Digital 5.1 surround sound to many of their products like YouTube. Coming up with their own surround sound standard will probably help them save on license costs.
Re:what's the endgame for Google? (Score:4, Insightful)
Stuff like this always looks fluffy and nice on first sight - but what is Google's endgame? Just cost saving? There's gotta be more to it than that.
Dolby has a lock on multichannel codecs which is utter bullshit. A 5 or 7 channel surround sound codec based upon flac is completely possible and adding it and syncing it to a video track is also possible. A competing codec for mp4 would be a good thing for the web. Even Microsoft is embracing open codecs with the inclusion of flac in the Edge browser engine to open up sound capabilities and take the idiotic restrictions off multi channel audio codecs. Now that physical DVDS and Blueray disks are dead meat it makes no sense to restrict audio and video codecs to the ones that were used on them. Mp3 is crap for surround sound even at higher bit rates restricted codecs should not become the only option for content creation. Relying on one entity to engineer sound which is distributed over the web in digital format is restrictive bullshit.
Re: what's the endgame for Google? (Score:4, Insightful)
Well, to their credit Dolby didn't just rest on their laurels as their tech aged out. They stayed on top of it, improving it, going lossless, adding channels, bringing "theatre" sound home. I still remember the first time I saw the old DTS logo for the first time in the theatre, during Jurassic park.
https://youtu.be/oaXSa1ZXOVw [youtu.be]
That was a true game changer. They've earned their place. Sure, competition would be good, but in no way do I begrudge them their success. They brought value.
Re: (Score:3)
Well, to their credit Dolby didn't just rest on their laurels as their tech aged out.
Really? I kind of thought that's exactly what they've done.
I still remember the first time I saw the old DTS logo for the first time in the theatre, during Jurassic park. That was a true game changer.
That was 30 years ago!
Re: (Score:2)
[DTS] earned their place.
I do think that DTS earned their place. And they have been very successful within the Blu-Ray market.
A few years ago, I was convinced that the Next Big Thing was going to be object-oriented audio, and I said "Whoever gets there first will be the next Dolby." DTS didn't pull that off. The next Dolby was... Dolby. Atmos [wikipedia.org] is the clear upgrade path after 7.1 audio, and it will be very tough for DTS to compete with.
I'm rooting for a free open-source multichannel solution. I haven't l
Re: (Score:2)
For multi-channel sound, that would be AAC. Support for it is already vast and widespread; multi-channel configurations are just not used very much.
Even Opus, for that matter, could be used for the same purposes. The new 3D audio spec Google is pushing allows for both codecs.
Re:what's the endgame for Google? (Score:5, Interesting)
what is Google's endgame? Just cost saving?
I think so, yes.
Look at Google's history here. They bought the company On2, and then spent a year or so having lawyers check over the source code for On2's video coder, and then released that coder (VP8 [wikipedia.org] as open source. They released On2's patents as well, also for free.
There's a network effect in coders. Users want the decoders for the popular format; then companies license the encoders for the popular format, because so many decoders are out there. It's self-reinforcing. Trying to introduce a new format is like trying to start a competitor to Facebook. (Remember Google+? I had forgotten it so completely I had to do a Google search to remind me what it was called.)
The best chance at getting people to adopt a new alternative is to give it away as the free and open alternative format. Even with it free, when Google released VP8, most people on Slashdot sneered at it. "Only MPEG decoders have hardware support on mobile devices. Why would I burn my phone's battery just because Google wants to save a few pennies on licensing fees?" A few free-software fans were happy but we seemed like the minority voices here.
Now, years later, hardware decoding support [webmproject.org] is common, but MPEG coders still rule the popularity contest.
But IMHO Google doesn't actually care that much if they need to spend a few pennies on MPEG licensing fees. IMHO they wanted a credible threat to make sure that the MPEG licensing authority didn't put the screws to Google's thumbs. Suppose MPEG licensing got greedy and tried to jack the fees... maybe even only jacking the fees for people using the coder at large scale (i.e. targeted fees that only Google had to pay)? Or perhaps introduced obnoxious licensing requirements, like Google needing to submit reports on how many users were watching videos. "Hey, Google, now that you have a hugely profitable business in YouTube, we have decided you need to start paying us a whole lot more." Google needs effective coders; MPEG-1 level technology won't even work for them. But with VP8, Google now could credibly say "gee, your new licensing won't work for us. We're going to have to pass. We have the free software coders; we're using those."
The MPEG licensing guys, in turn, were now incented by this threat to treat Google well, reducing Google's motivation to switch. The current situation is working pretty well for all parties.
Something similar to the above happened in the world of MP3 audio coding. Vorbis, a free software audio coder, is arguably better than MP3, but it never really caught on; MP3 was so popular that everyone was using it (or AAC for the higher-quality streams). Vorbis didn't displace the proprietary formats but IMHO spared us from any major abuses at the hands of the people holding the patents. (I do believe that Vorbis became quite popular for applications like music in video games. The game engines don't have to license any patents and gaming computers have no difficulty decoding Vorbis.)
Also, aside from the issues of control and licensing fees, I believe that many people who work for Google would like to make a positive change in the world. Everyone is better off if the commonly-used media formats are free and open. (Well, everyone except those who profit from the proprietary stuff.) Google can do things for more than one reason and some of the reasons might actually be altruistic.
P.S. Around the time Google released VP8, the MPEG licensing authority was starting to make noises about making end users get patent licenses for "commercial uses" [engadget.com] of H.264, the MPEG4 video coder. If you bought a digital video camera and recorded a video, the owner's manual said you couldn't use it for commercial uses (e.g. shooting video of a wedding for a fee) without g
Re: (Score:2)
Kill a competitor's income is a common capitalism game. Make a free competitor that negates all research they've done. Then reap the customers who think you did it all for them.
Re: (Score:2)
what is Google's endgame?
Commoditize all intellectual property, equipment, etc. required to run a cloud service advertising business... with the exception of IP owned by either Google or Meta (aka Facebook). The only IP that is allowed to be valuable is the algorithm that decides what piece of content shows up at the top of a webpage... that might be a search engine or it could be a Facebook news feed.
Everything other than that must be worth as close to ~$0 as possible because if other types of viable business exists in the technol
Re: (Score:2)
To avoid paying licence fees to Dolby. At the moment if you want Dolby Vision (HDR) and Dolby Atmos (3D sound) you have to pay them royalties on every device sold, and of course buy their over-priced documentation and developer kits. You have to pay them to certify your equipment too.
That limits the availability of those technologies to higher profit margin devices and services.
Google already did this with video, creating a free codec that performs at least as well as the proprietary ones. It's used on YouT
Re: (Score:2)
Re: (Score:2)
Stuff like this always looks fluffy and nice on first sight - but what is Google's endgame? Just cost saving? There's gotta be more to it than that.
In case you were living under a rock for the last 23+ years, there's been a war on general computing and honest text based executables. Beginning with ultima online in 1997 and everquest in 1999.
The public has been robbed of getting local applications as Google, valve, intel, AMD and the rest push us towards a return of mainframe computing (DRM is just the return of client-server model of computing, where you own nothing).
The whole point was to take over the entire input output of the machine and put it un
Re: (Score:2)
Re: (Score:2)
It's a free audio codec dude, not a DRM scheme. Your rant may apply to many things Google does but it has zero relevance to this.
You don't get they want the "free codecs" so they can use it in their DRM implementation idiot, so they don't have to pay money to dolby, aka the reason they want those codecs is you don't grasp once trusted computing and pluton is in every intel and AMD processor, the streams you dl from youtube will be encrypted, the whole point is to get back the "product model" of television so they can sell content to idiots and push unblockable adds.
You don't grasp that google has already been experimenting with DRM o
It Won't Be Open (Score:5, Insightful)
Re: It Won't Be Open (Score:2)
It's more interesting to know if it would be fine if someone else made a codec that would be compatible with the Dolby format.
At least for legacy stuff like 5.1. Patents do expire, but just the format of encoded data would be hard to defend in a patent suit.
Re:It Won't Be Open (Score:5, Informative)
Even as someone who hates Google more than most people, I have to concede that it doesn't look like that will be the case. That is to say, this stuff is open.
HDR10+, the HDR transport format Google is now pushing, is already an industry standard that's controlled by a non-profit consortium [hdr10plus.org]. The technical specifications are accessible for free to software groups, while hardware groups do have to pay a trivial annual membership fee (so it's open in the same way that DisplayPort is open).
In fact Google is a latecommer here, as the consortium and the standard is already several years old. They aren't even listed as a consortium member right now, so it seems they technically have zero control over the standard.
As for the 3D audio standard they're pushing, Immersive Audio Container [github.io], that is a bona fide open standard with the spec fully and freely published. Furthermore it's being overseen by the Alliance for Open Media, the same group that developed AV1.
Now, Google is a contributor to the initial spec, along with Samsung. But as it's now an AOM project, they only have as much control over it as the other members of the AOM.
Both of these are in sharp contrast to AMP, which is very much a Google standard. Specifically, AMP pages were frequently hosted/cached by Google, and there was significant concern there with how Google may have been favoring AMP pages in their search rankings. Neither of these would be applicable to Google, beyond I suppose the fact that the Play Store would be selling content that includes metadata based on these standards.
Re: (Score:1)
Re: (Score:2)
It'll be as 'open' as AMP, which Google touted as an open format yet they completely control it and all updates to it are to benefit themselves and prevent competitors from benefitting from it.
Why would you compare it to amp rather than say ... any other one of Google's open media formats. VP8, VP9, webm, webp, etc. Does it not fit your narrative that Google controls everything and it benefits only themselves?
The fact that Google actually has been in this business and gone down this path before, and that you had to pick one of their non-media related side businesses to make your point is actually quite a glowing review for Google. And the fact you're +5 insightfully is a not so glowing demonstrat
Expect it to be dropped in 3-6 years... (Score:5, Insightful)
Like all Google projects that do not massively increase their pile of cash. Relying on anything Google longer term is foolish.
Re:Expect it to be dropped in 3-6 years... (Score:5, Insightful)
Re: (Score:3)
These standards will be used on YouTube, so will certainly generate plenty of revenue for Google. Just like their previous open standards, VP9 and AV1, did.
Re: (Score:2)
Re: (Score:2)
Maybe. Linus Tech Tips used to produce HDR content, but they stopped a few months ago because the number of people actually watching in HDR was so tiny it wasn't worth the effort.
I suppose that might change as more people get HDR displays, although most devices that claim to be HDR aren't really.
Immersive audio seems even more niche. You need to do a lot of work in post to produce it, it's not a case of having a few extra microphones on your rig.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
My next question would be how this affects productions, then. As I understand it, they have to choose whether they want to encode in AC3, Atmos, DTS:X, etc. These formats require different mixing techniques and licenses paid by the production. Is Google selling them on yet another format they need to mix for, or will Google take the Atmos signal and reencode it for transmission to clients. While I understand the cost savings aspe
Re: (Score:2)
Like all Google projects that do not massively increase their pile of cash. Relying on anything Google longer term is foolish.
Like the 14 year old VP8 codec actively in use on Youtube? Google hasn't abandoned or depreciated a single media codec to date. The only truly foolish thing is to generically paint everything with the same brush.
How does this solve the Dolby problem? (Score:2, Interesting)
Customers still want their hardware to support Dolby formats so as to be compatible with every other non-YouTube source, so how does this actually solve the Dolby problem for manufacturers? It's adding another (maybe free) option to their feature matrix, but the manufacturers will still have to pay Dolby licensing costs for everything else their devices support.
And Google is not altruistic: what gain is Google getting out of this arrangement? More access to customer data?
Re: (Score:2)
Re: (Score:2)
Most low end TV's don't support neither Dolby Vision nor Atmos
Except Hisense, which supports both in their shitty models. $300ish for a 55" 4k with google tv and both of those features. The picture quality is mediocre, but my living room is dark enough that it doesn't matter. And somehow it has SOME specs (few, admittedly) that are better than much more expensive sets.
Re: (Score:2)
Don't Know If This Will Help or Hurt (Score:2)
Did they mean Dolby Digital Plus or TrueHD? (Score:2)
A manufacturer of streaming boxes that wholesale for $50 has to pay around $2 per unit for Dolby Vision and Dolby Digital,
Isn't AC3 out of patent protection now because of age?
Re: (Score:2)
Yes. But not Dolby Digital Plus (E-AC3), Dolby Atmos, and Dolby TrueHD. E-AC3 in particular is heavily used by the major streaming services, so as a hardware manufacturer you can't (currently) avoid it.
what else? (Score:2)
"Dolby makes most of its money through licensing fees "
Most?
What else does it make money from? genuine question.
Re: (Score:2)
Well, there's the twice-weekly late-night gambling ring they hold in their attic, but they don't really like to talk about that.
Less likely to happen (Score:2)
AV1 is easy - you just get a new hardware box that supports it and it outputs HDMI - you don't care what format the video Netflix or Youtube or other site sends you - as long as it shows up on the screen.
But HDR is different - there are literally dozens for ways to do it. The lowest common denominator supported by all TVs is known as HDR10. But you also have Dolby Vision, HDR10+, HLG, among others. And no screen supports more than one or two extra - usually Dolby Vision. But Samsung (and Panasonic) support
social-media (Score:1)
Good! (Score:2)