Open
Description
Currently the spec doesn't specify any way for VideoEncoder to signal that a frame has been dropped due to realtime
mode constraints.
I think it would be useful to have such a signal.
Possible options:
- the output callback should be called with an empty encoded chunk
- the output callback should be called with a new flag set in EncodedVideoChunkMetadata
This came up after WPT test change that accounts for frame drops.