Title

Dolby Vision Metadata Levels

URL Name

Dolby-Vision-Metadata-Levels

Created By

MichaelH, 4/1/2025 7:02 AM

Last Modified By

MichaelH, 4/1/2025 7:16 AM

Validation Status

Validated

Publication Status

Published

Assigned To

Assigned By

Assignment Due Date

Article Content

Dolby Vision Metadata

The Dolby Vision Mastering Metadata comprises global type, track type, shot type and frame-based structures. Global and track type metadata are static metadata and applied to the entire deliverable while shot-based and frame-based metadata are dynamic and applied on a per-shot and per frame in certain specific cases involving dissolves, transitions or animations. Dolby Vision metadata is created by the colorist during the color correction/grading process.

The metadata that is created can be exported from the color correction system as an XML.

There are two versions of Dolby Vision namely, CMv2.9 and CMv4.0. CMv4.0 uses improved algorithms and an advanced and more complex tone curve that allows for better mapping and more controls during the Dolby Vision trim pass process.
Dolby Vision metadata travels with the associated content and instructs Dolby Vision playback devices how a piece of content must be played back.
As there are two versions of Dolby Vision CM algorithms, there are also two versions of XML to support each CM version.

Note:
While there is built-in backwards compatibility in the metadata, the XMLs themselves are not backwards compatible.

In this article we're covering the following topics:

Levels of Metadata
Metadata Version 2.0.5 XML for CMv2.9
Metadata Version 4.0.2 XML for CMv4
Metadata Version 5.1.0 XML for CMv4

User-added image

Levels of Metadata

The Dolby Vision metadata in the XML is divided into a few levels based on the different aspects of the project and content that have to be recorded.

User-added image

L0 – Level 0 (zero) – Mastering/Target Display Characteristics

The different displays used to create the XML are listed as follows,

Mastering Display

The colorists’ selection (based on the HDR Mastering Display that is used to create/grade the HDR Master) in the Dolby Vision preferences on the color correction system, is recorded in the XML.

Target Displays

The different monitors used for the target trims done by the colorist during the Dolby Vision trim pass are recorded in this section of the XML. The 100nit/Rec709 trim is often mandatory for all Dolby Vision projects but additional trims may be requested by the studio or client.

L1 – Level 1 – (min, avg, max)

The Dolby Vision analysis produces three values per shot (or per frame) to describe the image and its dynamic range.

L2 – Level 2 – Target Trim Metadata

The trims and metadata adjustments done by the colorist during the trim pass in Dolby Vision version 2.9 are recorded as L2 metadata in the XML. L2 metadata is generated automatically by an algorithm and inserted into Dolby Vision v4.0 XMLs (for backward compatibility with legacy v2.9 devices).

L3 – Level 3 – L1 Offset Metadata

L3 metadata in the XML is calculated mathematically using a combination of the data obtained from the Dolby Vision analysis and the adjustments made to the Mid Tone Offset trim control in Dolby Vision v4.0.
The user interface of color grading applications usually displays the value of the manual adjustments made, whereas the value recorded in the XML reflects a combination of both - the mathematical calculation as well as the manual adjustment.

User-added image

L5 – Level 5 – Timeline Aspect Ratio Description

The aspect ratio of the project is described as L5 metadata in the exported XML. L5 is described as two values namely CanvasAspectRatio and ImageAspectRatio.

User-added image

L5 metadata normally applies to the entire project but individual shots can have their own unique aspect ratios. The global L5 values are listed at the beginning of the XML and per shot L5 metadata is included along with all the metadata (L0, L1, L2/L3/L8, etc.) associated with each shot.

L6 – Level 6 – MAXCLL/MAXFALL

L6 metadata is required for HDR10 content and includes two numbers namely, MAXCLL and MAXFALL. These values are not normally calculated as part of the Dolby Vision analysis and will usually have to be calculated separately.
[refer to How-To: Calculation of MaxFALL and MaxCLL metadata]

L8 – Level 8 – Target Trim Metadata

All the trim adjustments performed in Dolby Vision v4.0 are recorded as L8 metadata in the XML. This includes all the primary as well as secondary trims.

L9 – Level 9 – Mastering Display Primaries and Whitepoint

L9 metadata describes the color primaries and white point of the Mastering Monitor/Display used for the project. L9 is calculated automatically from the mastering display selection made by the colorist during the Dolby Vision content creation process.
L9 is listed per shot to improve the alignment of the XML to the bitstream metadata carriage.

Content Mapping Version - Level254 & Algorithm Version

The content mapping version is composed of two values. The main CM algorithm and the trim algorithm version.
In XML version 2.0.5 these values are listed under the <AlgorithmVersion> tag as e.g.:

<AlgorithmVersion>2,0</AlgorithmVersion>

In XML version 4.0.2 and above the CM version is listed under the <Level254> tag as e.g.:

<Level254 level="254">

</Level254>

CM Version	'Algorithm Version' (XML version 2.0.5)	'Level 254' (XML version 4.0.2 and above)	Notes
2.9	2,0 \| 2,1	n/a	current CMv2.9
2.9	2,1	n/a	CMv4.0 downgraded to CMv2.9 (using old backward compatible algorithm) - deprecated
2.9	2,2	n/a	CMv4.0 downgraded to CMv2.9 (using new backward compatible algorithm)
2.9	2,4	n/a	analyzed with CM algorithm version 4 (using cm_analyze)
2.9	2,5	n/a	CMv4.0 downgraded to CMv2.9 with artistically edited v2.9 metadata (L2)
4.0	n/a	4 0	Indicates old backwards backward compatible algorithm - deprecated
4.0	n/a	4 1	Indicates latest backwards backward compatible algorithm
4.0	n/a	4 2	Indicates artistically edited v2.9 metadata - no update required

When creating v4.0 content, v2.9 metadata (L2) will be generated automatically and inserted in the XML along with the 4.0 metadata (when using the latest backwards L2FromL8 algorithm).
This is done to support backward compatibility with legacy devices in the market that only support Dolby Vision v2.9. The metadata thus generated will include L2 metadata for three targets namely, 100nits, 600nits and 1000nits.

User-added image

L11 – Level 11 – Automatic Picture Optimization (APO)

L11 metadata stores the Content Type and intended White Point settings

Metadata Version 2.0.5 XML for CMv2.9

Metadata Level/Field	Description
LEVEL 0	GLOBAL METADATA (STATIC)
Mastering Display	Describes the characteristics of the mastering display used for the project
Global Aspect Ratio	Ratio of canvas and image (active area)
Frame Rate	Frame Rate
Target Display	Describes the characteristics of each target display used for L2 trim metadata
Color Encoding	Describes the image container deliverable
Algorithm/Trim Version	CM algorithm version and Trim version
LEVEL 1	ANALYSIS METADATA (DYNAMIC)
L1 Min, Mid, Max	Three floating point values that characterizes the dynamic range of the shot or frame Shot-based L1 metadata is created by analyzing each frame contained in a shot in LMS color space and combined to describe the entire shot as L1Min, L1Mid, L1Max
LEVEL 2	PER-TARGET TRIM METADATA (DYNAMIC)
Reserved1, Reserved2, Reserved3, Lift, Gain, Gamma, Chroma Weight, Saturation, and Tone Detail	User defined image controls to adjust the content-mapped images on a per- target basis.
LEVEL 5	PER-SHOT ASPECT RATIO (DYNAMIC)
Canvas, Image	Used for defining shots that have different aspect ratios than the global L0 aspect ratio
LEVEL 6	OPTIONAL HDR10 METADATA (STATIC)
MaxFALL, MaxCLL	Metadata for HDR10 MaxCLL - Maximum Content Light Level MaxFALL - Maximum Frame Average Light Level

Metadata Version 4.0.2 XML for CMv4

Metadata Level/Field	Description
LEVEL 0	GLOBAL METADATA (STATIC)
Mastering Display	Describes the characteristics of the mastering display used for the project
Global Aspect Ratio	Ratio of canvas and image (active area)
Frame Rate	Frame Rate
Target Display	Describes the characteristics of each target display used for L2/8 trim metadata
Color Encoding	Describes the image container deliverable
Algorithm/Trim Version	CM algorithm version and Trim version
LEVEL 1	ANALYSIS METADATA (DYNAMIC)
L1 Min, Mid, Max	Three floating point values that characterizes the dynamic range of the shot or frame Shot-based L1 metadata is created by analyzing each frame contained in a shot in LMS color space and combined to describe the entire shot as L1Min, L1Mid, L1Max Stored as LMS(CMv2.9) and L3 Offsets
LEVEL 3	OFFSETS TO L1 (DYNAMIC)
L1 Min, Mid, Max	Three floating point values that are offsets to L1 Analysis metadata as L3Min, L3Mid, L3Max L3Mid is a global user defined trim control L1 is stored as CMv2.9 computed values, CMv4 reconstructs RGB values with L1 + L3
LEVEL 2	PER-TARGET TRIM METADATA (DYNAMIC)
Reserved1, Reserved2, Reserved3, Lift, Gain, Gamma, Chroma Weight, Saturation, and Tone Detail	Automatically computed from L1, L3 and L8 (lift, gain, gamma, chroma weight, saturation, tone detail) metadata for backwards compatibility with CMv2.9
LEVEL 5	PER-SHOT ASPECT RATIO (DYNAMIC)
Canvas, Image	Used for defining shots that have different aspect ratios than the global L0 aspect ratio
LEVEL 6	OPTIONAL HDR10 METADATA (STATIC)
MaxFALL, MaxCLL	Metadata for HDR10 MaxCLL - Maximum Content Light Level MaxFALL - Maximum Frame Average Light Level
LEVEL 8	PER-TARGET TRIM METADATA (DYNAMIC
Lift, Gain, Gamma, Chroma Weight, Saturation, Tone Detail, Mid Contrast Bias, Highlight Clipping 6-vector (R,Y,G,C,B,M) saturation and 6-vector (R,Y,G,C,B,M) hue trims	User defined image controls to adjust the CMv4 algorithms per target with secondary color controls
LEVEL 9	PER-SHOT SOURCE CONTENT PRIMARIES (DYNAMIC)
Rxy, Gxy, Bxy, WPxy	Stores the mastering display color primaries and white point as per-shot metadata

Metadata Version 5.1.0 XML for CMv4

Metadata Level/Field	Description
LEVEL 0	GLOBAL METADATA (STATIC)
Mastering Display	Describes the characteristics of the mastering display used for the project
Global Aspect Ratio	Ratio of canvas and image (active area)
Frame Rate	Frame Rate
Target Display	Describes the characteristics of each target display used for L2/8 trim metadata
Color Encoding	Describes the image container deliverable
Algorithm/Trim Version	CM algorithm version and Trim version
LEVEL 1	ANALYSIS METADATA (DYNAMIC)
L1 Min, Mid, Max	Three floating point values that characterizes the dynamic range of the shot or frame Shot-based L1 metadata is created by analyzing each frame contained in a shot in LMS color space and combined to describe the entire shot as L1Min, L1Mid, L1Max Stored as LMS(CMv2.9) and L3 Offsets
LEVEL 3	OFFSETS TO L1 (DYNAMIC)
L1 Min, Mid, Max	Three floating point values that are offsets to L1 Analysis metadata as L3Min, L3Mid, L3Max L3Mid is a global user defined trim control L1 is stored as CMv2.9 computed values, CMv4 reconstructs RGB values with L1 + L3
LEVEL 2	PER-TARGET TRIM METADATA (DYNAMIC)
Reserved1, Reserved2, Reserved3, Lift, Gain, Chroma Weight, Gamma, Saturation, and Tone Detail	Automatically computed from L1, L3 and L8 (lift, gain, gamma, chroma weight, saturation, tone detail) metadata for backwards compatibility with CMv2.9
LEVEL 5	PER-SHOT ASPECT RATIO (DYNAMIC)
Canvas, Image	Used for defining shots that have different aspect ratios than the global L0 aspect ratio
LEVEL 6	OPTIONAL HDR10 METADATA (STATIC)
MaxFALL, MaxCLL	Metadata for HDR10 MaxCLL - Maximum Content Light Level MaxFALL - Maximum Frame Average Light Level
LEVEL 8	PER-TARGET TRIM METADATA (DYNAMIC
Lift, Gain, Gamma, Chroma Weight, Saturation, Tone Detail, Mid Contrast Bias, Highlight Clipping 6-vector (R,Y,G,C,B,M) saturation and 6-vector (R,Y,G,C,B,M) hue trims	User defined image controls to adjust the CMv4 algorithms per target with secondary color controls
LEVEL 9	PER-SHOT SOURCE CONTENT PRIMARIES (DYNAMIC)
Rxy, Gxy, Bxy, WPxy	Stores the mastering display color primaries and white point as per-shot metadata
LEVEL 11	Automatic Picture Optimization (APO)
arg1 = Content Type [0-15] arg2 = Intended White Point [0-15]	Stores the Content Type and intended White Point settings