I assume that you are registered to the use the playoutDelayHint on the origin trial experiment, right?
From my tests the neteq jitter buffer does not instantly adds
delay to the jitter buffer to match the hint delay, but lowers the
playout so it starts to ramp up slowly. I am trying to modify that
behavior, because I also think that adding the audio should match
the hint as fast as possible (adding silence at the beginning).
Also take note that the jitterBufferDelay stat is inaccurate, so
you can't rely on it for double checking if it is working or not.
Best regards
Sergio
--
---
You received this message because you are subscribed to the Google Groups "discuss-webrtc" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss-webrt...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/9796e6ab-a261-477b-90e8-022a453e8f94%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/d395cb2a-41c4-d0e7-0d8f-cdaa3634cd9e%40gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/CAJLUFi%3D0U67OwM35DUWoP8S0HHfZN6k0_J_ikxeqYwx0CH%2B0Dg%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/2361d335-7eaa-f9d3-bbd5-e0d52625d43b%40gmail.com.
I assume that you are registered to the use the playoutDelayHint on the origin trial experiment, right?
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/d395cb2a-41c4-d0e7-0d8f-cdaa3634cd9e%40gmail.com.
At least for local development that isn't required when one sets
Yes, that works too, thanks for the tip.
Best regards
Sergio
--
---
You received this message because you are subscribed to the Google Groups "discuss-webrtc" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss-webrt...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/111c32f5-b86d-0381-b73a-131edb63ac14%40gmail.com.
Hi Sergio,
I've noticed that videoReceiver.playoutDelayHint delays both video and audio.while audioReceiver.playoutDelayHint has no effect on audio delay.
const [audioReceiver, videoReceiver] = pc2.getReceivers();// Add additional 2 seconds of buffering.// audioReceiver.playoutDelayHint = 2; // Has no effect on audio delayvideoReceiver.playoutDelayHint = 2; // Delays both audio and video
Is this a design choice?
Probably you are also triggering lip sync which may affect
latency of both flows.
Are you only trying to modify the ramp behavior so video/audio gets in sync quicker?
My use case is audio only.
Is there plans to be able to control video delay and audio delay independent from each other? Ie, giving developers the ability to make asynchronous video/audio
Try disabling lip sync (by sending them in two different streams and RTCP RR).
Best regards
Sergio
--
---
You received this message because you are subscribed to the Google Groups "discuss-webrtc" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss-webrt...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/41734222-b53a-e10a-31ef-7be4c685cf31%40gmail.com.
--
---
You received this message because you are subscribed to the Google Groups "discuss-webrtc" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss-webrt...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/discuss-webrtc/e1dd2216-e35a-0f85-48e6-895cacc6dc77%40gmail.com.