The workflows for creating Dolby Vision content for distribution on OTT services are widely deployed and are integrated into many of the common tools that are used for OTT-focused post-production. However there are two high-level workflows to consider as a good starting point when looking to create Dolby Vision. Which of these two workflows to start with will depend on who is commissioning the content you are creating and which platform it will be distributed on. 

The information on this support site does cover both applications so it’s important to understand the difference in the types of files that different distribution platforms require.   

 

Distribution on a social media platform 

If you’re creating for distribution on a social media platform such as Vimeo or Bilibili then it’s likely you’ll be delivering an HEVC-encoded Dolby Vision file, such as a Dolby Vision profile 8.4 file. These files are encoded with substantially more compression than mezzanine formats such as IMF, and they are designed to be played back directly on consumer equipment although they can be further processed if required. If you are looking for information about these workflows see Dolby Vision for Social Media. 

 

Distribution on a major OTT service 

However, if you’re creating for distribution on a major OTT service like Netflix, Apple TV+ or Disney+ then this article can help get you started. It’s likely that you will be required to deliver a mezzanine file in a format such as IMF. These mezzanine files are generally high-data rate files with little video compression. They are designed to be edited, transcoded and/or processed as part of the delivery process and will not playback directly on consumer devices.   

A high-level Dolby Vision workflow





The main steps for creating Dolby Vision for an OTT service 

However, you are working in Dolby Vision there are two key parts you need to create 

• An HDR master – this is your hero video that can include a very wide dynamic range and expansive range of colors  

• Dolby Vision metadata – this controls how your HDR master image is adapted to other display devices that potentially have lower peak brightness or different color gamut to that of your HDR master image 

 

When creating Dolby Vision for an OTT service then it is likely that you will look to use shot-by-shot metadata, potentially with optional manual trim metadata. In this case your color grading system will need to support Dolby Vision metadata creation and content mapping, so you can audition the effects of this metadata on a range of targets, e.g. SDR. For more information about the different methods of creating Dolby Vision metadata see Dolby Vision metadata - shot analysis mode vs. long-play analysis mode.

 

The basic workflow steps for creating Dolby Vision for OTT distribution  

In order to create Dolby Vision content there are four key steps that need to be performed 

• Capture in HDR – capture or generate the source content with wide dynamic range and in a wide color gamut (WCG) 

• Edit/Grade in HDR - edit the content and color grade to create an HDR master 

• Generate Dolby Vision dynamic metadata - with the option to tailor (trim) this metadata if desired 

• Package/export a Dolby Vision mezzanine – package the content with the metadata to create the mezzanine asset as required by the studio/OTT service provider specifications 

 

While it is possible to create an HDR master from SDR content, whether completely or just using certain clips from SDR sources, for simplicity the process of expanding SDR sources is not covered here.  While it is often necessary to use at least some SDR content, starting with as much high dynamic and WCG content as possible will enable more spectacular content to be created.  For more information see Inverse tone mapping - upscaling SDR content to HDR 

 

Capture in HDR 

In order to create Dolby Vision content, it is important to start with content that has a wider dynamic range and uses a wider color gamut compared to SDR content as this will give more to work with in post-production.  For examples and further information about source formats see Dolby Vision Recommended Source Formats.  
These choices include a range of options including manufacturer specific choices for both the transfer function (opto-electrical transfer function or OETF) and the color space.  Other key factors to consider when capturing include the video codec, the data rate, and ensuring that a minimum of 10-bit capture should be used. 

 

Edit/Grade in HDR 

You will then need to use at least a color grading workflow that supports Dolby Vision metadata creation and content mapping. You can find a list of compatible systems at Dolby Vision compatible Color Grading & Packaging Systems. 

When creating Dolby Vision content for an OTT operator you will usually need to create a PQ-based master with up to 1000 nit peak brightness. This doesn't mean that the content has to peak at 1000 nit, the actual peak is a creative choice, just that this is the range available within which the master should be contained.  When creating for Dolby Vision delivery, it is also necessary to select the Mastering Display characteristics which are the closest match to your main reference display.  While the system supports mastering display choices that are >1000 nit, the two primary options are for a display with 1000 nit peak brightness and either P3 or Rec.2020 support. For more information on selecting a mastering display see our article on HDR monitor considerations.

 

Generate Dolby Vision dynamic metadata 

Once you have completed you HDR master grade, it is then time to create your Dolby Vision metadata. For most use cases this will be created by performing shot based analysis on your grade from within your color grading system. Optionally you may then create trim metadata that gives you a much greater level of control as to how tone mapping will be applied downstream when mapping to different target displays or when creating SDR versions from a Dolby Vision master. 

For examples of how to perform shot based analysis see Create Dolby Vision metadata and for more information on trim metadata see Dolby Vision Trim Controls. 

 

Package/export a Dolby Vision mezzanine 

You will then need to master and package your finished project into the format or formats required by the studio. We have published an example Dolby Vision 4.0 Generic Studio Spec and you can find more information as well as links to delivery specifications from major OTT operators and studios at Deliver in Dolby Vision to a studio or OTT service provider . 

The most basic way of carrying Dolby Vision metadata is as a separate sidecar XML file, but this can also be combined with the video within certain file formats such as IMF.  For examples of exporting each of these options see Export a Dolby Vision Master and Export a Dolby Vision IMF. For a list of packaging systems that support Dolby Vision export see Dolby Vision compatible Color Grading & Packaging Systems 

The final Dolby Vision-enabled video encoder at the OTT operator will take this metadata, translate it to the format required by consumer devices, and include it with the encoded video file, such as alongside HEVC video within an MP4 file. 

 

Further Information 

• There’s also lots more information on this site in the Dolby Vision Post-Production section or you can visit the Forums  

• You can see an overview of various training options at Learning options or you can sign-up for instructor-led training and certification on the Dolby Professional Store or see the Dolby Vision Training and Certification FAQs. 

• For more information on why you should create in Dolby Vision see Dolby Vision for Content Creators on professional.dolby.com 

• For more information on why consumers are choosing Dolby Vision and on other services and devices that support Dolby Vision see Dolby Vision on dolby.com