AVS Forum banner

1 - 12 of 12 Posts

·
Registered
Joined
·
86 Posts
Discussion Starter #1
I'm going to be heavily referencing this white paper from Dolby: https://www.dolby.com/us/en/technologies/dolby-vision/dolby-vision-white-paper.pdf

"The Dolby Vision HDR reference monitor (capable of up to 4,000 nits luminance) is used to make the color and brightness decisions... After the reference color grade is finished, the Dolby Vision color-grading system will analyze and save the dynamic metadata that describes the creative decisions made on the display. The content mapping unit (CMU) maps the content with the metadata to a reference display at a standard brightness (100 nits).

The dynamic metadata generated to create the SDR grade can be used to render the Dolby Vision master on displays... [all] referencing the same metadata and Dolby Vision reference images."

So basically, the DV metadata describes a transform from the HDR10 grade down to an SDR grade for each frame of the film. Dolby's characterization of a manufacturer's display is used to make the DV hardware which uses the transform in the meta-data to perform it "part way"... down from 4000 max nits to 1000 max nits (vs 100) for example.

So what about the 10 bit HD resolution Dolby vision layer that's bundled with the 10 bit base layer 4K video stream? The white paper, on page 11, describes this as the enhancement stream but does not say what it is specifically... does it add the data necessary to bring the base layer 10 bit file to a 12 bit file? If so, why is it at 1/4 the resolution of the base layer? Also, if so, why does a single layer Dolby Vision solution only require a 10 bit file?

Other Dolby lit mentions that the "Dolby Vision dual-layer backward-compatible solution includes an uncompromised SDR Rec. 709 signal." Given that the color information in 4K 4:2:0 video is 1/4 the resolution of the non color information, does this mean that the enhancement layer contains the SDR REC709 color data from the SDR grade? If so, why does does this (along with the meta data) need to be combined with the HDR base layer to create what Dolby describes as the "full-range signal" (in 12 bits of course)?

Would appreciate it if someone in the know could straighten out my naivety and errors in these regards!

On an aside, it really seems like Dolby Vision is the best HDR standard out of any available.
-It requires grading on Dolby Displays, creating a reference standard.
-It allows the colorist to specifically define how the HDR image is to be represented on SDR displays; no leaving it up to display manufacturers.
-Dolby characterizes certified displays and creates custom software to ensure that the conversions are done properly.
-Dolby claims that the 12 bit full range signal drastically reduces the potential for banding and stepping on 10k nit systems.

HDR10+ brings the HDR to SDR metadata to allow the colorist some more control, but because there's no grading display standardization, no standard hardware across display manufacturers... it seems like a free for all that will result in a LOT more variability from display to display, and it also seems like some studios and display manufacturers are opting for it simply because they are cheap and don't want to pay Dolby for the hard work they have done (worth noting that Dolby developed the HDR10 spec, and doesn't see a dime for that).
 

·
Registered
Joined
·
112 Posts
Here is a quote from Display Daily, quoting Patrick Griffs from Dolby Labs. This is about the best easy English explanation I could find.

"Griffis said that for streaming video, Dolby Vision provides a backward-compatible framework for HDR. There is a base layer of SDR content encoded with a Gamma EOTF and decodable by any streaming video decoder for showing on an SDR display. Then there is an enhancement layer that includes all the necessary data to convert the SDR stream into an HDR stream and metadata that tells the Dolby Vision decoder exactly how to use this enhancement layer data to perform this SDR to HDR conversion. Griffis said the enhancement layer and the metadata adds just 15% on average to the size of the bitstream. This allows streaming companies like Netflix to keep just one version of the file and then stream it to either SDR or HDR customers. An SDR (or HDR10-only) system discards the extra data and shows the SDR image. This saves storage space on the server, at the expense of higher streaming bit rates to SDR customers. Of course, in 2016, the vast majority of streaming customers are SDR customers and this streamed enhancement data is mostly discarded."

And here is a more in depth Q and A with Patrick Griffs.
https://hdguru.com/griffis-dolby-visions-asuper-set-of-all-other-4k-hdr-approaches/

These are both over a year old, but this is a good read for anyone interested in HDR
 

·
Registered
Joined
·
259 Posts
Just FYI Dolby Vision is never carried in an actual 12-bit stream. On disc it uses a 10-bit HDR10 3840x2160 base layer and a 10-bit DV 1920x1080 enhancement layer containing the difference data and dynamic metadata, these two layers are combined to produce the 12-bit 3840x2160 Dolby Vision output.
:)
 

·
Registered
Joined
·
87 Posts
Above quote from Geoff D, is that a marketing bull****?

First of all, H.265 HEVC Main 10 only supports up to 10-bit.

Dolby Vision Profile 7 (dual layer) which is used on 4k blu rays has two layers of which:
1. base layer in H.265 HEVC Main 10 YUV420 10-bit
2. DV enhancement layer in H.265 HEVC Main 10 YUV420 10-bit

How 10-bit data in two different streams in a lossy codec can be combined to construct a 12-bit data. Lost data is not retrievable, only interpolatable.
Also, 12-bit of data is awful lot to distribute. The difference in file size (HDR10 vs DV) is too small to accommodate this.

Most likely he meant padding 10-bit into a 12-bit container. This is very different from combining data from two different streams to construct a higher resolution data.
 

·
Registered
Joined
·
992 Posts

·
Registered
Joined
·
87 Posts
At the encoder stage level (i.e. post-production), the 12-bit Dolby Vision signal is splitted into a base layer and an enhancement layer.

At the decoder stage level (i.e. UHD Blu-ray player), the 12-bit Dolby Vision is recombined.

https://www.avsforum.com/forum/150-...f-2017-players-later-year-5.html#post52706065
I see. Are you implying that 12-bit reconstruction technique is Dolby's secret sauce (patent) ?

Further questions I have are:

1. Official Dolby Vision video demo files available Dolby site and 4kmedia.org usually are in .ts and/or .mp4 containers.
Using MediaInfo tool, one can see that those files are in DV Profile 5 (single layer) and in 10-bit color depth.
Since they are in a single layer and there's no additional/enhancement layer to be combined with, are those official DV video demo files are in 10-bit?

2. Apple iTunes 4K DV streaming as well ad Netflix DV contents streaming use DV Profile 5.
According to Apple, they are a single layer (Profile 5) DV encoded in HEVC Main 10 YUV420 10-bit. So, 10-bit (without reconstruction) for streaming services?

3. Are 4K UHD Blu-ray discs with DV Profile 7 (dual layers) the only available 12-bit source in consumer market?

4. It's my understanding that in early days of DV, it required a hardware module. However, these days, DV is capable of running on software (provided that silicon has enough additional headroom to run additional post-processing job required by DV).
Is Dolby Vision's secret sauce (12-bit reconstruction technique) in any way affected by the presence (or absence) of dedicated DV hardware?
In other words, any difference in hardware-based vs. software-based implementation on display (TV) and/or source player (AppleTV and/or 4k blu-ray player) in terms of 12-bit reconstruction technique?
 

·
Registered
Joined
·
992 Posts
I see. Are you implying that 12-bit reconstruction technique is Dolby's secret sauce (patent) ?
Yes, I do.

It is not anymore secret.
https://www.etsi.org/deliver/etsi_gs/CCM/001_099/001/01.01.01_60/gs_CCM001v010101p.pdf


Further questions I have are:

1. Official Dolby Vision video demo files available Dolby site and 4kmedia.org usually are in .ts and/or .mp4 containers.
Using MediaInfo tool, one can see that those files are in DV Profile 5 (single layer) and in 10-bit color depth.
Since they are in a single layer and there's no additional/enhancement layer to be combined with, are those official DV video demo files are in 10-bit?
Yes, they are.


2. Apple iTunes 4K DV streaming as well ad Netflix DV contents streaming use DV Profile 5.
According to Apple, they are a single layer (Profile 5) DV encoded in HEVC Main 10 YUV420 10-bit. So, 10-bit (without reconstruction) for streaming services?
Yes, a Profile 5 bitstream is only 10-bit, but it uses IPT / ICtCp constant luminance color space necessary for a higher quality color volume mapping to serve the diversity of HDR displays than the legacy non-constant luminance YCbCr used by other Profiles.
https://www.jonolick.com/uploads/7/9/2/1/7921194/p020160311294995790791.pdf
https://www.hdm-stuttgart.de/~froehlichj/diss/Jan_Froehlich_Thesis_2018_02_15_Acrobat_300dpi.pdf

The 12-bit dual stream Dolby Vision is also documented in streaming DASH-IF Interoperability Points.
https://dashif.org/docs/DASH-IF-IOP-v4.2-clean.htm


3. Are 4K UHD Blu-ray discs with DV Profile 7 (dual layers) the only available 12-bit source in consumer market?
Yes, because the industry standard HEVC encoders / decoders are 10-bit for the time being.


4. It's my understanding that in early days of DV, it required a hardware module. However, these days, DV is capable of running on software (provided that silicon has enough additional headroom to run additional post-processing job required by DV).
Is Dolby Vision's secret sauce (12-bit reconstruction technique) in any way affected by the presence (or absence) of dedicated DV hardware?
In other words, any difference in hardware-based vs. software-based implementation on display (TV) and/or source player (AppleTV and/or 4k blu-ray player) in terms of 12-bit reconstruction technique?
"Dolby Vision is a format + processing/mapping software" [FPGA, …] and there is no specific Dolby Vision hardware chip.
https://www.avsforum.com/forum/465-...al-hdr-compliant-displays-4.html#post50302361
Dolby Vision requires a powerful processor (SoC) due to its dynamic metadata / color appearance model display adaptation and its 12-bit dual layer architecture.
Before the introduction of the MEL (Minimal Enhancement Layer), Dolby Vision required a SoC to be capable of simultaneously decoding two video streams (base layer video stream and Full Enhancement Layer video stream).
https://www.avsforum.com/forum/465-...l-hdr-compliant-displays-12.html#post56988004
https://www.avsforum.com/forum/465-...l-hdr-compliant-displays-13.html#post57025528

By the way, Apple TV 4K can only decode a 10-bit single layer Profile 5 bitstream.
https://www.apple.com/apple-tv-4k/specs/
 
  • Like
Reactions: Madmax67

·
Registered
Joined
·
3,274 Posts
Thanks, Danba. Great stuff and reading.
Stacey Spears mentioned that the full enhancement layer contains data from the original HDR master. He also said he wish Sony went with the original DV format over the low latency model.

ICtCp color space is very noticeable, at least for me it is.

Sent from my SM-G550T1 using Tapatalk
 

·
Registered
Joined
·
87 Posts
Stacey Spears mentioned that the full enhancement layer contains data from the original HDR master. He also said he wish Sony went with the original DV format over the low latency model.

ICtCp color space is very noticeable, at least for me it is.

Sent from my SM-G550T1 using Tapatalk
Any way to tell or know which 4k bluray titles have MEL or FEL? Is there a such list, by any chance?
 
1 - 12 of 12 Posts
Top