Resources

This page has useful resources for those interested in or developing with CMAF.

CTA WAVE Specifications

The following specifications developed by the CTA WAVE project are useful resources to understanding CMAF and DASH-HLS Interoperability:

For more detail please see CTA WAVE Project specifications page – and test suites page – and further reading page


Other Specifications

The specifications in this chart either reference or are referenced by the CMAF specification.


Click to view the chart


User Stories

Here are the examples of how CMAF is benefiting the video streaming industry.

Content Provider

As a Content Provider, we would like…

  • to be able to produce linear feeds with variable elements such as audio, video, or captions that define a dynamic experience such as format transitions with low latency and on every streaming device
    • HD or UHD video with SDR and/or HDR picture
    • Single or multi-language tracks with channel- and/or object-based audio mixes
    • Descriptive Video Service or Audio Description for selected programs
    • Closed captions and/or forced narratives
  • to be able to produce file-based assets with predictable Quality of Experience on every streaming device given the source file format and the selected media profiles defined by the CMAF industry standard
  • to be able to produce content that will be securely distributed using encryption or watermarking to enable audience measurement on any streaming device using the best industry standard techniques
  • to be able to produce content that may be associated with alternate advertising content that is inserted at predictable splicing points without degrading the Quality of Experience on the target device

— Mickael Raulet, Chief Technology Officer, Ateme



Commercial (Ad) Content Provider

As a Commercial (Ad) Content Provider, we would like…

  • to be able to enable Digital Ads Insertion (DAI) into linear feeds with variable elements such as audio, video, or captions that define a dynamic experience such as format transitions on every streaming device
    • HD or UHD video with SDR and/or HDR picture
    • Single or multi-language tracks with channel- and/or object-based audio mixes
    • Descriptive Video Service or Audio Description for selected programs
    • Closed captions and/or forced narratives
  • to be able to produce file-based assets with predictable Quality of Experience on every streaming device given the source file format and the selected media profiles defined by CMAF industry standard
  • to be able to produce content that will be safely distributed e.g. protected and marked to enable audience measurement on any streaming device using best industry standard techniques
  • to be able to produce content that may be associated with alternate advertising content that is inserted at predictable splicing points without degrading the Quality of Experience on target device

— Manuel Briand, Lead Video Quality Engineer, Media Delivery, Disney Streaming


Encoder Vendor

As an Encoding vendor, we would like …

  • to be able to produce a common media container at various audio / video bitrates that can be played back on various type of devices.
  • to be able to produce a common media container enabling low latency delivery across various devices at the quality of broadcast, while being backward compatible with standard playback.
  • to be able to produce a common media container enabling fast channel start and fast channel change.
  • to be able to produce common media containers with well-defined structures that can be directly addressed with multiple manifest formats to enable optimization of delivery infrastructure.
  • to be able to produce common media containers with well-defined structures that enable splicing, ad insertion and alternate content replacement without needing to reprocess content.
  • to be able to produce common media containers enabling content protection.

— Mickael Raulet, Chief Technology Officer, Ateme


Content Distribution

As a Content Distributor, we would like …

  • to be able to produce common media containers that fully contain and temporally align the video, audio, text, and images associated with a content presentation to simplify and consolidate the implementation of a rich presentation experience on device.
  • to be able to produce common media containers that enable seamless adaptive switching of video and audio streams on a device in response to changes in the device runtime environment.
  • to be able to produce common media containers that enable seamless adaptive switching of video and audio streams on a device in response to changes in the linear source content format e.g. programming transitions.
  • to be able to produce common media containers with well-defined structures that can be directly and completely addressed with at least one manifest format to enable optimization of delivery infrastructure.
  • to be able to produce common media containers with well-defined structures that enable splicing and combination of presentations via a manifest to enable ad insertion and alternate content replacement without needing to reprocess content.
  • to be able to encrypt common media containers with a single industry-standard secure encryption method in a way that is interoperable with multiple DRM solutions to reduce the processing load, storage volume, and CDN cache size of encrypted content.
  • for these common media containers to define constraint profiles that enable simple detection of player compatibility across a wide ecosystem of players and devices.

— Zachary Cava, Sr. Principal Engineer, Architect, Disney Streaming


Content Delivery Network

As a Content Delivery Network, we would like …

  • to be able to receive common media containers for all LIVE and VOD content being submitted to our network for distribution. Having a single media container for a given asset
    • increases the cache efficiency (since only one copy must compete for space at the edge and mid-tiers) which in turn increases delivery quality-of-service.
    • simplifies workflow actions such as file inspection, concatenation, encryption and modification.
    • minimizes origin storage requirements
    • minimizes midgress traffic
    • removes the requirement for dynamic packaging services, thereby simplifying delivery workflow and raising network capacity.
  • for these containers to have consistent file extensions which predict their internal content - for example, audio, video, captions, timed text data., init segments. This allows delivery behaviors and caching rules to be assigned to media objects without having to parse their internal content.
  • for these containers to have consistent mime-types which map unambiguously to their file extensions. This allows an edge server to set the appropriate mime-type in the response purely by inspecting the file extension of the object being delivered.
  • for these common containers to carry easily parseable identifiers which describe and constrain their internal content, in the event that we need to manipulate or modify that content.
  • for these common media containers to allow multiple DRM solutions to be used concurrently to protect the encapsulated media content, reducing the storage and caching footprint for encrypted content.

— Will Law, Chief Architect, Edge Technology Group, Akamai


Digital Rights Management Provider

As a DRM Provider, we would like to …

  • to be able to manage a single version of an encrypted asset which is accessible using any DRM or device
  • to minimize the head-end key management
  • to be able to consistently manage all DRMs and their associated signaling
  • to support various business models in a consistent way between devices and DRMs, for example, to have the possibility to grant a device access to precise pieces of content across a precise time period or subset of tracks.

For those, media containers should

  • Allow encrypted data to be independent from the DRM used on the device
  • Allow sharing content keys between DRMs
  • Provide standard locations for the carriage of DRM information
  • Support standard mechanisms for supporting business models such as VOD subscription, long-term or short-term Live subscription and Pay-Per-View

— Laurent Piron, Principal Solution Architect, Nagra