Skip to content

w3c/media-capabilities

main
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Media Capabilities API

Problem Statement

To deliver media across an enormous range of devices and networks, modern services stream their content in a wide range of formats - varying by codec, resolution, container format, encryption scheme, and other dimensions. Unlike many devices, browsers run across a wide range of platforms and hardware, so identifying which formats are optimal - or even supported - is challenging. Existing Web APIs, like isTypeSupported and canPlayType are vague and insufficiently expressive; thus sites often pick an arbitrary target independent of actual capabilities.

Main Requirements

The Media Capabilities API seeks to provide definitive answers on several playback dimensions:

  • Given its properties, can a specified piece of media content be played at all?
  • Will playback have high quality (smoothness)? Will it be power efficient?
  • Do output capabilities (e.g. color gamut, dynamic range, audio channels) match content?
  • Are security requirements (e.g. encryption, HDCP, etc) supported or playback-impacting?
  • Given multiple possible media formats, which is preferable?

The sum of these requirements is simple; given the reality that content items are stored in multiple formats and levels, which of these should be selected for delivery to a particular UA?

Solution Outline

The proposed solution is to introduce a new API and corresponding set of properties that meet the requirements above:

  • A new query-based capability API allows pages to identify whether a given combination of properties - codec and profile, resolution, frame/sample rate, bit rate, bit depth, pixel format, color space, EOTF, content protection scheme, and others - is supported. If so, estimates of playback quality and power efficiency will also be provided. This mostly corresponds to the decoding capability of the UA.

  • Additional output properties (e.g. on window.screen), such as color gamut and dynamic range, will identify what can meaningfully be rendered to the user. This helps sites avoid resource-intensive formats that can be decoded but not properly displayed (e.g. HDR content on a SDR display, or multichannel audio with stereo speakers).

  • Even with the above, existing APIs and heuristics remain important. Network and CDN throughput aren’t addressed by the above but often constrain viable bitrates. Real-time contention for CPU/GPU, detected using dropped frame count or perhaps the new Media Playback Quality API, may also dynamically limit quality (e.g. forcing lower resolution).

The above allows sites to make optimal decisions, most of the time, before streaming begins.