Eugene Voityuk

Hello @ggarber! The goal of back-porting was not just to simply upgrade libwebrtc version, but use Loss V2 estimator, multiple bugfixes was a bonus. When loss v2 estimator is enabled...

@ggarber You are more than welcome to checkout this branch and do testing and playing with params, if you want, you can contact me on telegram, viber, whatsapp: +380630593550. I...

> @sarumjanuch `libwebrtcFieldTrials` PR is merged in v3. Please, when appropriate follow steps at [#960 (comment)](https://github.com/versatica/mediasoup/pull/960#issuecomment-1332517890) @ibc Done.

Hello @vpalmisano! So initially we started testing with artificial congestion created by traffic shapers, but they are mostly useless, because they do not reflect how the real network queues behaves....

Now few words about importance of Loss v2 estimator. Loss v2 is a giant leap forward in comparison what GCC was doing before. Prior it loss adjustments to BW estimate...

> Thanks for the explanation. It I've correctly understood, you are running: > > * 1 single producer at HD resolution on machine A; > * 19 consumers running on...

You can see loss percent (blue line) as a second column on y axis on a first chart, at every screenshot. Avarage loss rates might be higher, in case of...

> Can you give an example of how to manipulate the `fieldTrial` string so we can have multiple scenarios? createWorker has a new param now, which allow you to override:...

Hello @vsviatetskyi, I believe i have seen your questions on google webrtc groups. New loss estimator can be adjusted to tolerate this by changing this flags over filed trial setting...

@ggarber After additional testing and this branch includes fix from #962.