Dolby Vision 2 Announced!

[hajj_3]

https://news.dolby.com/en-WW/253671-dolby-unveils-dolby-vision-2-a-new-era-for-tv-picture-quality/

Dolby Vision only has ~9yrs of patent life left so now they have Dolby Vision 2!

[kurkosdr]

https://news.dolby.com/en-WW/253671-dolby-unveils-dolby-vision-2-a-new-era-for-tv-picture-quality/

Dolby Vision only has ~9yrs of patent life left so now they have Dolby Vision 2!

Dolby is quickly becoming "the enhancement layer company" and "the signal processing company": Dolby Vision is an enhancement layer over HDR10 (in most implementations), Dolby Atmos is an enhancement layer over TRUEHD (whose patents should expire within the decade), and Dolby Digital Atmos (E-AC3+JOC) is an enhancement layer over E-AC3 (whose patents are either already expired or will expire within the decade, haven't found a clear answer to that). Plus the appropriate signal processing to make the enhancement layers work on consumer hardware of course (or kind-of-work if we are talking about IPS displays and crappy soundbars).

And I kind of like that, if you want enhancements, pay patent royalties for it, much better than those days when you had to pay patent royalties for an AC3 decoder even if you only wanted basic stereo playback of DVDs or TV broadcasts in some countries.

The only exception is AC-4, but in practice it's used as an enhancement layer (with E-AC3 or AAC being transmitted alongside it to maintain backwards compatibility).

That said, we have zero information what this Dolby Vision 2 thing even is. It's probably going to be something over HDR10, but what is it? Is it an enhancement layer? Is it new signal processing techniques over the existing Dolby Vision enhancement layer? The announcement offers zero technical details.

[ShortKatz]

I also just read about Dolby Vision 2. I would be interested to hear the community's opinion on Dolby Vision 2. Will it be a real hit or more of a flop? Is it useful and helpful, or not really?

[kurkosdr]

I also just read about Dolby Vision 2. I would be interested to hear the community's opinion on Dolby Vision 2. Will it be a real hit or more of a flop? Is it useful and helpful, or not really?
It's going to be something every high-dollar TV will have for sure (nothing wrong with that, premium products are expected to include premium technologies).

For anything more, we'll need to know what it even is.

[Blue_MiSfit]

From talking with them at NAB in April it looks like a backwards-compatible set of new metadata options that creatives can use to optimize presentation in different environments.

DoVi has classically solved for taking a large color volume and mapping it down into a smaller one - for example to take a 2000 nit peak with colors at P3 limits (i.e. a very big volume) down to a display with a 700 nit peak and maybe not-so-good P3 coverage. It gives creatives the tools to define various "trims" on a shot by shot basis to optimize this mapping to their liking for multiple types of displays.

Inverting this idea (i.e. mapping volumes UP) is I think part of DoVi 2. For example, maybe you have content mastered to a peak of 1000 nits, but you acknowledge that some people have very bright ambient environments and they might have a display capable of a 4000 nit peak (hypothetically). Those people want a bright image so they can actually see it, and they have hardware that's capable of it. Classic DoVi would not expand the content to fill the capabilities of the display, but I think DoVi 2 will allow this. I don't have specifics, but at a minimum creatives can pick from a handful of preset modes, presumably on a shot by shot basis. There's also likely "automatic, based on shot by shot analysis" capabilities when creatives don't want to spend the time on this.

They also demonstrated similar "mode based" signaling for various types of post-processing. I think motion interpolation is called out in the publicly released info and I wouldn't be surprised if other image enhancement things like noise reduction / grain management / sharpening end up in there too.

In typical DoVi fashion this should all be backwards compatible so older encoders / displays can take the new metadata and just ignore the elements of it they don't understand.

[kurkosdr]

From talking with them at NAB in April it looks like a backwards-compatible set of new metadata options that creatives can use to optimize presentation in different environments.

DoVi has classically solved for taking a large color volume and mapping it down into a smaller one - for example to take a 2000 nit peak with colors at P3 limits (i.e. a very big volume) down to a display with a 700 nit peak and maybe not-so-good P3 coverage. It gives creatives the tools to define various "trims" on a shot by shot basis to optimize this mapping to their liking for multiple types of displays.

Inverting this idea (i.e. mapping volumes UP) is I think part of DoVi 2. For example, maybe you have content mastered to a peak of 1000 nits, but you acknowledge that some people have very bright ambient environments and they might have a display capable of a 4000 nit peak (hypothetically). Those people want a bright image so they can actually see it, and they have hardware that's capable of it. Classic DoVi would not expand the content to fill the capabilities of the display, but I think DoVi 2 will allow this. I don't have specifics, but at a minimum creatives can pick from a handful of preset modes, presumably on a shot by shot basis. There's also likely "automatic, based on shot by shot analysis" capabilities when creatives don't want to spend the time on this.

They also demonstrated similar "mode based" signaling for various types of post-processing. I think motion interpolation is called out in the publicly released info and I wouldn't be surprised if other image enhancement things like noise reduction / grain management / sharpening end up in there too.

In typical DoVi fashion this should all be backwards compatible so older encoders / displays can take the new metadata and just ignore the elements of it they don't understand.

So, it does define with new metadata, which means you need new software to export Dolby Vision 2 content, right?

Also, is backwards compatibility something they told you explicitly, or something that you assume should be true? Asking because if it's not backwards compatible, existing Dolby Vision TVs will drop to plain HDR10 when playing Dolby Vision 2 content.

[Blue_MiSfit]

So, it does define with new metadata, which means you need new software to export Dolby Vision 2 content, right?

Also, is backwards compatibility something they told you explicitly, or something that you assume should be true? Asking because if it's not backwards compatible, existing Dolby Vision TVs will drop to plain HDR10 when playing Dolby Vision 2 content.

I'm _fairly certain_ The product manager I was speaking with (Max Denton) did say that this would all be new metadata that would be backwards compatible. Feel free to ask him, he's a great guy!

In other words, a legacy encoder would still work with new metadata (ignoring it, making regular DoVi), and a legacy display would still work with a new stream (ignoring the bits it doesn't understand).

That's how things worked with Content Mapping v4 came out (kind of DoVi 1.5) We shall see though :)

[benwaggoner]

I'm _fairly certain_ The product manager I was speaking with (Max Denton) did say that this would all be new metadata that would be backwards compatible. Feel free to ask him, he's a great guy!
Max is indeed a font of awesomesauce.

In other words, a legacy encoder would still work with new metadata (ignoring it, making regular DoVi), and a legacy display would still work with a new stream (ignoring the bits it doesn't understand).

That's how things worked with Content Mapping v4 came out (kind of DoVi 1.5) We shall see though :)
That is my understanding, and really the only way this could work in practice without splitting the ecosystem.