[box4m]

What do we think about this old thread now?

Dolby Licenced Media Players:
Since this thread was started we saw the 1st HDR10 DV capable hardware released in late 2017 in the form of the Apple TV 4K. Free iTunes HD > 4K movie upgrades followed often including DV encoding.
The Infuse and MrMC Apps on the ATV 4K can playback DV Single layer .mp4's

2018 saw the DV capable Amazon FireTV Stick 4K released.
2019 saw the DV capable Amazon Cube released.

---

Amazon was one of the founding members of the HDR10+ Alliance.
Amazon might be using HDR10+ encoding for their 4K HDR original content. There is a lot of uncertainty however.
The only 4K DV content I've seen on Amazon Prime is the Jack Ryan TV series.

Recently we see NVIDIA finally got proper Dolby Licencing for their Tegra X1+ chipset which also allowed the 2019 Shield to internet stream DV content. Plex Beta software on the 2019 Shield is testing DV Single layer .mp4 support

In 2019 Panasonic, also a HDR10+ Alliance founding member finally had to add DV support to their HDR10+ TV range.

DolbyVision vs HDR10+ on 2019 Panasonic OLED's (click)

Virtually all recently produced 4K HDR10 original content from Apple, Netflix and Disney is now available with DV encoding. (And DD+ with Atmos)
So DV has definitely gone mainstream with the Big Boys entering the streaming market.

It is my understanding the DV capable online streaming Media Players just mentioned above are all using the Dolby's Single Layer Profile 5 DolbyVision software solution talked about in Post #1 of this thread.

W.  

and now i want my favorite platform kodi to support it

[ozkhan1]

This will all be dandy, if there was a way to download DV content from the bigger providers (Netflix, Et al) for your library. Honestly I am paying for the service so it’s not piracy. You can technically do it with playon but the quality is crap. If there was a way to download DV/HDR content with atmos/other hd audio then it would be great for Kodi use. I honestly don’t buy ultrahd discs. I am sure I am not the only one in this boat.

If there was such a way within Kodi to do it and Kodi supported DV, it would be the holy grail. No need for other boxes..
Just get a N2 and you are ready to rock n roll.

[box4m]

Someone in another thread wrote "So you're probably going to see single-layer DV MP4 support really soon on Kodi since all they have to do is use the new hardware decoder for DV instead of the hardware decoder for HVEC. Quite easy to change in the software. " im keeping my fingers crossed

[Soli]

Don't think you need particularly many "cpu ponies" to software encode DV. (Not taking into account the decoding of HEVC) But I guesss until somebody reverse engineers it, it won't be supported in Kodi or other open source players. (Except Android boxes with a "black box" GPU path)
I think I used to be able to use Calman and output DV from my Mac Mini 2006 (running Windows) to calibrate DV on a supported TV. As long as the display output is bitperfect it's simple bit shifting/tunneling + metadata really (for intel GPUs you'd need to run in RGB LOW/Full with the infoframe patch that tell the TV to treat the data as HIGH/Limited, and set Kodi to display the interface as HIGH/limited).  Isn't that bascially how they do it on DV UHD BD anways since it's just normal HDR + DVD metadata?  

My first thought regarding Kodi is to be able decode normal HDR then encode it as DV before output to TV. That way I could use Kodi to software decode DV/4K/HEVC with my old HTPC with 4 core i7 4770K,  and output HDR as DV without needing HDMI 2.0 or higher, prolonging it's lifespan but with a higher electricity bill. You'd only get static metadata of course, but I wonder if the metadata benefits tonemapping all that much. I mean, the metadata is kinda embedded in the picture itself, and tonemapping seem to work fine without. Ah well, that's another story...

[box4m]

Im not asking for software decoding of DV though, im asking for the same thing as the guy said in my post above: hardware decoding of DV with DV compatible hardware, kodi just needs to give the instructions to use the hardware decoder

[noggin]

Don't think you need particularly many "cpu ponies" to software encode DV. (Not taking into account the decoding of HEVC) But I guesss until somebody reverse engineers it, it won't be supported in Kodi or other open source players. (Except Android boxes with a "black box" GPU path)
I think I used to be able to use Calman and output DV from my Mac Mini 2006 (running Windows) to calibrate DV on a supported TV. As long as the display output is bitperfect it's simple bit shifting/tunneling + metadata really (for intel GPUs you'd need to run in RGB LOW/Full with the infoframe patch that tell the TV to treat the data as HIGH/Limited, and set Kodi to display the interface as HIGH/limited).  Isn't that bascially how they do it on DV UHD BD anways since it's just normal HDR + DVD metadata? 

Aren't UHD BDs dual-stream devices - with an HDR10 base 3820x2160p HEVC stream (which HDR10-only UHD BD players happily play) and then a 1920x1080p DV enhancement layer that a DV UHD BD player merges with the HDR10 stream to create a DV output (either player-led low latency, or TV-led tunnelled, metadata). Whether the DV UHD BD player specifically tunnels the DV metadata as a separate process for TV-led DV - or whether it's baked into the HDR10+DV Enhancement layer merged video I don't know, but the metadata can presumably also be separately fed via HDMI (not tunnelled) for player-led DV?

I think the enhancement layer and the metadata are not the same thing in all cases?

[box4m]

Don't think you need particularly many "cpu ponies" to software encode DV. (Not taking into account the decoding of HEVC) But I guesss until somebody reverse engineers it, it won't be supported in Kodi or other open source players. (Except Android boxes with a "black box" GPU path)
I think I used to be able to use Calman and output DV from my Mac Mini 2006 (running Windows) to calibrate DV on a supported TV. As long as the display output is bitperfect it's simple bit shifting/tunneling + metadata really (for intel GPUs you'd need to run in RGB LOW/Full with the infoframe patch that tell the TV to treat the data as HIGH/Limited, and set Kodi to display the interface as HIGH/limited).  Isn't that bascially how they do it on DV UHD BD anways since it's just normal HDR + DVD metadata? 

Aren't UHD BDs dual-stream devices - with an HDR10 base 3820x2160p HEVC stream (which HDR10-only UHD BD players happily play) and then a 1920x1080p DV enhancement layer that a DV UHD BD player merges with the HDR10 stream to create a DV output (either player-led low latency, or TV-led tunnelled, metadata). Whether the DV UHD BD player specifically tunnels the DV metadata as a separate process for TV-led DV - or whether it's baked into the HDR10+DV Enhancement layer merged video I don't know, but the metadata can presumably also be separately fed via HDMI (not tunnelled) for player-led DV?

I think the enhancement layer and the metadata are not the same thing in all cases? 

Yes but there are single layer dolby vision (netflix etc) and there are dual layer dolby vision.
Dual layer dolby vision you can only play with UHD player, single layer you can play with appletv and the new shield with plex beta apparently.

I would love single layer dolby vision support in kodi, so it would call the hardware decoder as it does with HDR10 material but instead dolby vision

[Video Titles]

The only way that you are going to get the real deal is hard copies or the cinema.

If you honestly think that you are getting true 4k dolby atmos and real snow via a streaming service, you really need to think again.

You have more chance of having real snow this Christmas come out of your streaming box than ever having proper 4k and dolby atmos.

A 4K TV will upgrade a video signal anyway, whether it be from a DVD that is 100 years old, dumb and dumber to name but one movie.

You are being sold a lie, you read the manual this TV is that good it will even make you coffee in the morning.

My God I really want a TV that makes coffee, and you wake up in the morning, well I am still waiting for my coffee, but I still believe that it makes coffee.

Although the internet is huge, but if you was getting real 4K plus all those cat videos, and even worse there is always one person on your faceyawn account that posts 1,000 I am depressed pictures an hour, you would have a melt down on the servers and the phone lines.

Plus when you add to that all the claims of multiple room no slow downs on internet speed rubbish, I mean come on people.

I have had buffering just from watching youtube videos by myself at 720p a very rare occasion, but it has happened. Do you honestly believe everything that you are told?

[box4m]

The only way that you are going to get the real deal is hard copies or the cinema.

If you honestly think that you are getting true 4k dolby atmos and real snow via a streaming service, you really need to think again.

You have more chance of having real snow this Christmas come out of your streaming box than ever having proper 4k and dolby atmos.

A 4K TV will upgrade a video signal anyway, whether it be from a DVD that is 100 years old, dumb and dumber to name but one movie.

You are being sold a lie, you read the manual this TV is that good it will even make you coffee in the morning.

My God I really want a TV that makes coffee, and you wake up in the morning, well I am still waiting for my coffee, but I still believe that it makes coffee.

Although the internet is huge, but if you was getting real 4K plus all those cat videos, and even worse there is always one person on your faceyawn account that posts 1,000 I am depressed pictures an hour, you would have a melt down on the servers and the phone lines.

Plus when you add to that all the claims of multiple room no slow downs on internet speed rubbish, I mean come on people.

I have had buffering just from watching youtube videos by myself at 720p a very rare occasion, but it has happened. Do you honestly believe everything that you are told?

Im not sure what to reply to this but.
Im sorry youre this mad from a DV question
Single layer DV is just metadata from the DV stream merged with the hdr10 stream, Im not sure why you say its not DV.
All it is is metadata

[noggin]

Single layer DV is just metadata from the DV stream merged with the hdr10 stream, Im not sure why you say its not DV.
All it is is metadata 

That's not how single stream DV works is it?

AIUI Dual stream DV uses a base stream (for UHD BDs with DV this is a 2160p HEVC HDR10 stream, but DV allows for AVC or HEVC and SDR Rec 709, SDR Rec 2020, HDR10 Rec 2020 or HLG Rec 2020 base streams) and an enhancement layer (which can be lower resolution than the base stream) which allows a DV Composer in the UHD BD player or other dual stream box to create an HDMI output accompanied with metadata (*). (The DV Display Management processing that takes this video and metadata is in the display)

AIUI Single stream DV uses just one stream - but AIUI this is not HDR10, it's a DV native stream (OTT platforms have the luxury of being able to send a player a stream that is suitable for it - so DV players request and receive a DV stream, HDR10 players request and receive an HDR10 stream etc.)? The metadata is extracted from the stream and carried over HDMI (*).  The single stream is a 10-bit HEVC video stream, so can be decoded by a standard HEVC decoder (the metadata is carried within the stream using additional data that can be extracted separately).

AIUI Single Stream DV isn't usually / always just an HDR10, HLG or SDR compatible stream with metadata (HDR10, HLG and SDR compatible streams need a DV enhancement layer and dual stream as well as metadata)? Or am I missing something ?

https://www.dolby.com/us/en/technologies...t-v2.0.pdf

AIUI single stream DV stuff used by OTT operators is profile 5 ?

4.3.2 Profile 5 Dolby Vision stream
This MPD example describes a media presentation that consists of a Dolby Vision profile 5 video component with a Dolby Digital Plus audio component. The essence of the Dolby Vision track is a profile 5 Dolby Vision stream encoded as 10-bit HEVC video with a resolution of 3840 × 2160 at 30 fps, where the base layer is not backward compatible.

(*) Metadata is either 'buried' into the active video (i.e. tunnelled) or carried separately over HDMI in the standard HDMI dynamic metadata channels?

HOWEVER - I think it is also possible to carry a base and enhancement layer (i.e. a dual stream DV signal) multiplexed into a single video stream (which may be how some devices are doing things?)

My understanding is that a DV signal either contains a base layer (backwards compatible for non-DV devices), an enhancement layer and a reference-picture stream - which the DV composer uses to create a DV video signal to carry over HDMI with accompanying metadata - or a single layer that is just DV (and not backwards compatible) with a reference-picture stream that the DV composer uses to generate a DV video signal to carry over HDMI with metadata.

Or am I totally misunderstanding this?

[ozkhan1]

Single layer DV is just metadata from the DV stream merged with the hdr10 stream, Im not sure why you say its not DV.
All it is is metadata 

That's not how single stream DV works is it?

AIUI Dual stream DV uses a base stream (for UHD BDs with DV this is a 2160p HEVC HDR10 stream, but DV allows for AVC or HEVC and SDR Rec 709, SDR Rec 2020, HDR10 Rec 2020 or HLG Rec 2020 base streams) and an enhancement layer (which can be lower resolution than the base stream) which allows a DV Composer in the UHD BD player or other dual stream box to create an HDMI output accompanied with metadata (*). (The DV Display Management processing that takes this video and metadata is in the display)

AIUI Single stream DV uses just one stream - but AIUI this is not HDR10, it's a DV native stream (OTT platforms have the luxury of being able to send a player a stream that is suitable for it - so DV players request and receive a DV stream, HDR10 players request and receive an HDR10 stream etc.)? The metadata is extracted from the stream and carried over HDMI (*).  The single stream is a 10-bit HEVC video stream, so can be decoded by a standard HEVC decoder (the metadata is carried within the stream using additional data that can be extracted separately).

AIUI Single Stream DV isn't usually / always just an HDR10, HLG or SDR compatible stream with metadata (HDR10, HLG and SDR compatible streams need a DV enhancement layer and dual stream as well as metadata)? Or am I missing something ?

https://www.dolby.com/us/en/technologies...t-v2.0.pdf

AIUI single stream DV stuff used by OTT operators is profile 5 ?

4.3.2 Profile 5 Dolby Vision stream
This MPD example describes a media presentation that consists of a Dolby Vision profile 5 video component with a Dolby Digital Plus audio component. The essence of the Dolby Vision track is a profile 5 Dolby Vision stream encoded as 10-bit HEVC video with a resolution of 3840 × 2160 at 30 fps, where the base layer is not backward compatible.

(*) Metadata is either 'buried' into the active video (i.e. tunnelled) or carried separately over HDMI in the standard HDMI dynamic metadata channels?

HOWEVER - I think it is also possible to carry a base and enhancement layer (i.e. a dual stream DV signal) multiplexed into a single video stream (which may be how some devices are doing things?)

My understanding is that a DV signal either contains a base layer (backwards compatible for non-DV devices), an enhancement layer and a reference-picture stream - which the DV composer uses to create a DV video signal to carry over HDMI with accompanying metadata - or a single layer that is just DV (and not backwards compatible) with a reference-picture stream that the DV composer uses to generate a DV video signal to carry over HDMI with metadata.

Or am I totally misunderstanding this?

Sheesh, fellas!

I just want something that looks nice! If I wanted to read and understand this much, I could have been a rocket scientist

[poofyhairguy]

Single layer DV is just metadata from the DV stream merged with the hdr10 stream, Im not sure why you say its not DV.
All it is is metadata 

That's not how single stream DV works is it?

AIUI Dual stream DV uses a base stream (for UHD BDs with DV this is a 2160p HEVC HDR10 stream, but DV allows for AVC or HEVC and SDR Rec 709, SDR Rec 2020, HDR10 Rec 2020 or HLG Rec 2020 base streams) and an enhancement layer (which can be lower resolution than the base stream) which allows a DV Composer in the UHD BD player or other dual stream box to create an HDMI output accompanied with metadata (*). (The DV Display Management processing that takes this video and metadata is in the display)

AIUI Single stream DV uses just one stream - but AIUI this is not HDR10, it's a DV native stream (OTT platforms have the luxury of being able to send a player a stream that is suitable for it - so DV players request and receive a DV stream, HDR10 players request and receive an HDR10 stream etc.)? The metadata is extracted from the stream and carried over HDMI (*).  The single stream is a 10-bit HEVC video stream, so can be decoded by a standard HEVC decoder (the metadata is carried within the stream using additional data that can be extracted separately).

AIUI Single Stream DV isn't usually / always just an HDR10, HLG or SDR compatible stream with metadata (HDR10, HLG and SDR compatible streams need a DV enhancement layer and dual stream as well as metadata)? Or am I missing something ?

https://www.dolby.com/us/en/technologies...t-v2.0.pdf

AIUI single stream DV stuff used by OTT operators is profile 5 ?

4.3.2 Profile 5 Dolby Vision stream
This MPD example describes a media presentation that consists of a Dolby Vision profile 5 video component with a Dolby Digital Plus audio component. The essence of the Dolby Vision track is a profile 5 Dolby Vision stream encoded as 10-bit HEVC video with a resolution of 3840 × 2160 at 30 fps, where the base layer is not backward compatible.

(*) Metadata is either 'buried' into the active video (i.e. tunnelled) or carried separately over HDMI in the standard HDMI dynamic metadata channels?

HOWEVER - I think it is also possible to carry a base and enhancement layer (i.e. a dual stream DV signal) multiplexed into a single video stream (which may be how some devices are doing things?)

My understanding is that a DV signal either contains a base layer (backwards compatible for non-DV devices), an enhancement layer and a reference-picture stream - which the DV composer uses to create a DV video signal to carry over HDMI with accompanying metadata - or a single layer that is just DV (and not backwards compatible) with a reference-picture stream that the DV composer uses to generate a DV video signal to carry over HDMI with metadata.

Or am I totally misunderstanding this? 

When I have looked into this it seems there isn't dedicated silicon to recombine the two layers and instead it is done via proprietary software on 4K Blu Ray players (the only mainstream source of true 12 bit dual layer DV content).  So that means it could technically be reverse engineered one day as it isn't a situation where you need a certain SoC to decode dual layer content. 

But as you said the whole point of dual layer DV is backwards compatibility and so the motivation to reverse engineer it might never exist. In dual layer DV the first stream is SDR, and the second layer has both the HDR10 and DV metadata to enhance the first steam. It would be cool to be able to have a single file in your library that can scale from SDR devices to HDR10 to DV devices, but efforts into HDR->SDR tone mapping that eventually also works for DV content is probably an easier way to get there (that then works with single layer files too) than figuring out how to handle an actual dual layer DV file. 

And then of course there are other big questions: how to support Dolby Vision away from proprietary devices or kernels, or how to accommodate it in a mkv container, how to handle the licencing on the device side so open source software can leverage the Dolby Vision bitstreams, etc. It feels like HD Audio bitstreaming back in 2008 all over again except with even more challenges. 

What sucks it I personally think Dolby Vision is a big upgrade over regular HDR10, and I had that proven to me again this week as I go back and forth between my devices that only support HDR10 and support Dolby Vision with Disney+. Luckily its not THE standard yet (Samsung will eventually wave the white flag as it has way more support than HDR10+) and so a lot of content has already been released in a HDR10 format that will probably never be upgraded to Dolby Vision. So getting good HDR10 support is enough for now, at least in my opinion. I am much more worried about Nvidia no longer supporting Linux decoding options like they used to, or the fact disks (the main source of good content) are dying and so a lot of great HDR and Dolby Vision content might never even get a chance to be played in Kodi (like those excellent Dolby Vision copies of the original Star Wars movies locked in Disney+).

[box4m]

You guys clearly know alot more about this then me, all i know if Infuse on appletv and Plex on shield can play single layer dolby vision, so i just thought maybe kodi can get it aswell, maybe naive

[noggin]

But as you said the whole point of dual layer DV is backwards compatibility and so the motivation to reverse engineer it might never exist. In dual layer DV the first stream is SDR, and the second layer has both the HDR10 and DV metadata to enhance the first steam.

That's not the case for UHD Blu-rays with DV is it? The base layer is HDR10 (and so when you play a DV UHD BD on a non-DV UHD player you get HDR10, not SDR) - and the enhancement layer is just used for DV purposes. The HDR10 backwards compatibility in non-DV aware players is a key reason that DV makes sense for disc publishers - the disc is HDR10 on a non-DV UHD BD player and DV on a DV-aware one.

There ARE Dual Layer DV profiles for SDR, HDR10 and HLG base layers - but UHD DV Blu-rays use the HDR10 base level variant?

[box4m]

So whats the result of these discussions? do we think kodi will get single layer dv mp4 support if the hardware can play DV?