I wanted to ask you how do you tackle RTP-forwarded streams that have dynamically changing resolutions due to poor bandwidth conditions.
I am using with the VideoRoom plugin and I RTP-forward the participants so I can process them and compose their contributions into an enriched video scene (with graphics, animations, ...).
Strong visual artifacts ocur when a publisher has a low bandwidth because that triggers resolution changes on the connection and introduces artifacts that are visible if you consume (with VLC or fflpay) that RTP-forwarded stream or any manipulation done upon it (gstreamer, or others...).
I was investigating how to mitigate those issues, and the approaches I thought about were:
I guess it’s pretty common to RTP-forward a participant to post-process that video further down the pipeline... What would be the best way to protect yourself from those cases? Do you do any post-processing to “sanitize” those artifacts (e.g. freezing the image until the bitrate is good again)?
I am sorry if you think this question is out of Janus’ scope, but I believe it’s very related since it’s about getting out its media streams from it.
Thank you!