Mixing of HDR and SDR videos in teleconferencing contexts

[tidoust]

Laptops and smartphones now frequently ship with an HDR webcam and display. In a teleconferencing scenario that uses WebRTC, there may be a mix of participants with HDR cameras and displays, and participants with SDR cameras and displays. Unless I missed something, it is not possible today to negotiate the color space when establishing a WebRTC communication (as in not supported by the underlying SDP protocol), so there cannot be any guarantee that a conference room will be HDR only or SDR only.

My understanding is that mixing HDR and SDR content can sometimes be tricky. Is that the case when mixing HDR and SDR videos (be it in a pure peer-to-peer scenario or when the streams go through an intermediary Selective Forwarding Unit (SFU)? Or do video streams include all the necessary information to mix and render HDR videos and SDR videos?

Cc @dontcallmedom.