RTCEncodedVideoFrameType in case of encrypted content

[youennf]

On receiver side, when SFrame encrypted content is retrieved, the frame type cannot be computed from the media content itself.

It can be computed from:

• RTP header extensions such as frame marking or AV.1 DD.
• From SFrame packetisation format itself if it provides this ability (design under discussion).

Following on last WebRTC WG meeting, the suggestion is to:

• Use information from RTP level if possible (SFrame packetization if available, RTP header extensions otherwise).
• If no information at RTP level, add a new unknown value to RTCEncodedVideoFrameType.