This is amazing. It's truly bizarre that 5.1 surround sound, or HDR, or spatial audio, would be proprietary paid formats -- I mean, what if someone told you there was a license for stereo audio?

And sure it's in Google's self-interest so that they can bring these technologies to YouTube without paying anyone else. But it benefits everybody, so this is really fantastic news for everyone if it's something that takes off.

We are seeing a deeper and deeper split between Google (webm/VP9, webp, AV1, strongly pushing their formats) and Apple (HEVC, HEIC, Atmos, completely boycotting Google's formats), with Microsoft caught in the middle, supporting neither that well.

Apple's stance is especially interesting because it's unclear to me what they gain by pushing license fee encumbered formats.

There is already an open media format (edit: for object-based immersive audio), it's called SMPTE 2098. Granted it's basically the mutant stepchild of DTS and Dolby ATMOS, but it does exist.

The real problem isn't the hardware manufacturers but the content producers. Dolby engages in blatant anticompetitive behavior that basically requires hardware manufacturers to support their codecs and make it impossible to innovate on the actual media formats in a way that might compete. For example: paying for content to be released in atmos or giving away the tools to author it for free.

It will be interesting to see how this plays out. Dolby Vision is actually a mess of a standard, with several different not-quite-compatible "profiles." Streaming video is Profile 5, UHD Blu-ray Discs are Profile 7[1], and iPhone Dolby Vision is Profile 8. Profile 7 cannot be converted into Profile 5 [completely incompatible and different algorithms!], devices that implement Profile 5 can't necessarily play Profile 7, but Profile 7 can with difficulty be theoretically converted into Profile 8 which is basically stripped-down Profile 7 with quirks[2]. Basically, Dolby Vision is fragmented within itself. Fun stuff.

[1] And within Profile 7, there is the difference between the MEL (Minimum Enhancement Layer) which just adds HDR data, versus the FEL (Full Enhancement Layer) which actually adapts a 10-bit core video stream into a 12-bit one for FEL compatible players. Not all Profile 7 implementations can handle FEL, but can handle MEL. So even the profiles themselves have fragmentation. FEL and MEL are, within Profile 7, actually HEVC video streams that are 1920x1080 that the player reads simultaneously with the 4K content. So a FEL/MEL player is actually processing 2 HEVC streams simultaneously, so it's not a huge surprise why it isn't used for streaming DV.

[2] Profile 8 comes in 3 different versions, Profiles 8.1 through 8.4. 8.3 is not used. Profile 8.1 is backwards compatible with an HDR10 stream, Profile 8.2 a SDR stream, and Profile 8.4 an HLG stream. Big surprise that iPhone uses 8.4 because HLG can be seamlessly converted into SDR or some other HDR formats when necessary.

As a consumer, I don't care. dolby makes money from licensing, but don't ask that much, they innovate constantly, and they do a lot of public education.

This seems like one corporation flexing on another rather than great sense of mission; it's not like Google doesn't have IP of its own that it prefers to keep locked up. I suspect that this signals a strategic desire to move into the A/V production space, where customers have big demands for storage and computing resources.

Anyone remembers the open format HDR10+ pushed by Fox, Panasonic, and Samsung?

Me neither.

The world at large has settled on Dolby Vision and Atmos and it will be very difficult to change this. Not only from the consumer end but specially in the pro audio/video end.

Google would need first to offer plugins for DAWs, video software, etc, to work with these formats before there's enough content that manufacturers and streamers consider it.

What does Dolby actually do on a day to day basis?

Do they have researchers working on new audio and video formats?

Or is it now all just a self-perpetuating machine for generating licensing revenue, based on existing patents?

Sorry for the ignorant question but I'm clueless about their ongoing contributions to the industry.

I love it when a powerful corporation's self interest happens to align with the public's interests.
Whatever the pirates adopt will be the de facto codec
The EU is randomly investigating the Alliance for Open Media on antitrust concerns --

Can anyone understand how it's in anyone's best interest to investigate / potentially stop an open source standard / royalty-free format that has buy-in from tons of big orgs?

It's super unclear to me why Dolby keeps being the ones to do basic things. What is underneath the marketting gloss? It feels like we are all paying a lot for high bit depth, paying a lot for multi-channel audio.

I have never understood how or why it is that expensive proprietary codecs keep taking over. Maybe there is more value add somewhere, but it's very unclear, esepcially under the gloss of (usually deeply non technical) marketting fluff.

Worth noting that Apple's vice president in charge of the AR/VR team, Mike Rockwell, is a former senior executive at Dolby Laboratories.
For all the evil Google may be guilty of they have done a lot of good work supporting open standards and releasing proprietary formats as open source. Let's not forget how Chrome liberated the web from Microsoft's won't-fix attitude whilst IE remained the dominant browser.
I look at the current reality of VP8/9 with dissatisfaction. Google went all-in with it and made Meet/Hangouts use it. But encoding, decoding, or both end up being done on the CPU usually, since hardware support is way behind. Zoom and FaceTime just used H.264 (and 5?), and it's way more efficient as a result. I don't normally care a ton about efficiency, but it actually matters when your laptop's fans are overpowering the audio in a meeting and draining your battery to 0 within a short time span.

Also, ironically, even Google Chat didn't seem to support webp images until recently. I appreciate the idea of open standards, but compatibility matters way more to the end user.

There is difference between hearing vs listening vs feeling. I hope more creative stuff comes out this. My first experience on 5.1 was Top Gun.. The next one to top that was in Audium with sound and vibrations from everywhere. Nothing to top the sound of silence.
W-We already have AV1 and Opus. Google backed AV1, and they already use both on YouTube.

I mean, they could make a better open video codec, give me AV2 any day. But why not push the pre-existing standards as "premium offerings"?

And btw [Opinion Incoming!] I believe Opus is as good as lossy audio formats will ever get. I'd love to be proven wrong...

Wonder how well this will compare to Ambisonics?

That's supposed to be a "full sphere" surround sound format (developed ~50 years ago), but hasn't been picked up widely:

Good idea. Perpetual patents on audio and video are ridiculous. Shouldn't they all expire?
Given Google's history of killing projects I'm not jumping on this ship just yet.
Subtext: Google doesn't want to pay licensing fees.
Any suggested readings about codecs/encodings/formats/algorithms used or whatnot? I’m afraid it’s the thing I lack most as a dev
Surprised Dolby has survived for so long.
I’d like to think that “don’t be evil” applies here, but that might be wishful thinking.

What are some of the risks of media formats being centralized by a mega corp like Google who works with nation states?

Can we truly expect something free… or can we expect all of the content we create to be steganographically watermarked in surveillance states that appear to be fully cracking down on encryption?

Would be nice if they started from opensourcing chromecast
Through Google Fonts, they can track page views under the guise of beautiful fonts for your site. I wonder if they could do the same with these media formats.
'Google wants to take on...'

My immediate reaction to reading these few words is - "another tool for the Google graveyard"

What's in it for Google? Less friction allows users to consume more which gives Google more data?
(Someone who works in streaming)

While I see both sides, I don't agree with this strategy. In fact, I think the public should be more aware of just how damaging Google/YouTube is to the streaming ecosystem and if you really stretch this argument, the planet.

It is true - HEVC's original licensing structure was a nightmare, but it seems to have been resolved and we now have hardware decoders in nearly all modern consumer devices.

This is also becoming true of Dolby's formats. maybe I am biased or not as informed as I could be but they did the R&D, worked with some of the brightest (pun intended) in the industry and created a production-to-distribution pipeline. Of course there are fees, but vendors are on board and content creators know how to work with these standards.

Now here comes one of the largest companies in the world. HEVC? Nope - they don't want to pay anyone any fees so instead they're going to develop the VP9 codec. Should they use HLS or DASH? Nope, they are going to spin DASH off into our own proprietary HTTP deliverable and only deliver AVC HLS for compatibility reasons. Apple customers complain and after years they cave and support VP9 as a software decoder starting with iOS14. This means millions of users eat significant battery cycles just to watch anything, including HDR video.

Then we get to Chrome. HEVC? Nope. Dolby? Nope. HLS? Nope. The most popular browser in the world doesn't support any of the broadcast standards. It's their way or fallback to SDR and the less efficient AVC codec.

So now anyone else in the streaming industry trying to deliver the best streaming experience has to encode/transcode everything three times. AVC for compatibility (and spec) reasons, HEVC for set-top boxes and iOS, and VP9 for Google's ecosystem. If it wasn't for CMAF the world would also have to store all of this twice.

In the end, to save YouTube licensing and bandwidth costs, the rest of the industry has to consume 2-3x more compute to generate video and hundreds of millions of devices now consume an order of magnitude more power to software decode VP9.

If and when Project Caviar becomes reality, it'll be another fragmented HDR deliverable. Dolby isn't going away and Chrome won't support it, so the rest of the industry will have to add even more compute and storage to accommodate. In the name of 'open' and saving manufacturers a couple dollars, the rest of the industry is now fragmented and consumers are hurt the most.

YouTube weirdly admitted this fragmentation is becoming a problem. They can't keep up with compute and had to create custom hardware to solve. Of course, these chips are not available to anyone else and gives them a competitive edge:

Pono Player checking in.
The last thing I want to do is upgrade all my TVs and audio equipment for new standards
kill em, google :)
obligatory relevant XKCD:
yes Fuck Dolby.
I have an entire speaker setup that runs on the chromecast protocol(s?)

They've been repeatedly bricked (features rolled back, support changed, can't set up complete groups, etc) by Google in the last few years, to the point where I don't even think I have them connected right now.

I don't trust consumer products from Google at all.

AKA we will crush your smaller company focused on high quality standards with our half assed support that lasts until you’re dead, maybe people will confuse this for “open”
Month M + 0: Google to take on $INDUSTRY_STANDARD with $GOOGLE_THING standard

Month M + 4: Google shows off $GOOGLE_THING and announces $PARTNER devices

Month M + 9: $PARTNER releases first devices with $GOOGLE_THING support (also supports $INDUSTRY_STANDARD, of course)

Month M + 18: Google disappointed with lack of adoption of $GOOGLE_THING announces first-party products with $GOOGLE_THING support

Month M + 24: Google's internal team working on first-party $GOOGLE_THING products dissolved

Month M + 36: $PARTNER announces future products will no longer support $GOOGLE_THING due to lack of demand

Month M + 48: Google removes all mentions of $GOOGLE_THING from their websites, docs, etc.

While I don't have sympathy for proprietary formats that come with an added use charge, alternative Google formats forced upon the world in the name of a “healthier, broader ecosystem” tend to create friction and unwanted overhead. Thinking of AMP and webP in particular.

And uncertainty ... how long will such efforts last before Google loses interest or is forced to abandon them?