You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am currently testing ABR WebRTC (and LL-HLS), and I realized that the data from availableIncomingBitrate becomes completely inaccurate when I stream still images (such as a PowerPoint presentation, for example). I believe the issue has already been identified here: #1066 (reply in thread)
With a still image, even with CBR, the stream is much lower than the configured bitrate, and the availableIncomingBitrate assumes that the utilized bandwidth is very low, resulting in an estimate of available bandwidth that is far below what is actually possible.
If I recall correctly, I had the same issue with WebRTC and mediasoup when I enabled simulcast with still images, back when only goog-remb was available. Since switching to transport-cc, I haven't encountered this problem anymore.
On OME, transport-cc is not yet available:
I wanted to know if there are plans to support it in future releases?
Additionally, and this isn't an issue with OME, I have the same problem with LLHLS. When I send a still image, the packets are very small, and the DNS resolution times and TCP request initialization have too much impact on the actual packet loading time. This also results in a low bandwidth detection with the hls.js library. Based on my tests, once a packet exceeds 30kB, the data starts to become relevant. Has anyone in the community encountered this issue before?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I am currently testing ABR WebRTC (and LL-HLS), and I realized that the data from availableIncomingBitrate becomes completely inaccurate when I stream still images (such as a PowerPoint presentation, for example). I believe the issue has already been identified here: #1066 (reply in thread)
With a still image, even with CBR, the stream is much lower than the configured bitrate, and the availableIncomingBitrate assumes that the utilized bandwidth is very low, resulting in an estimate of available bandwidth that is far below what is actually possible.
If I recall correctly, I had the same issue with WebRTC and mediasoup when I enabled simulcast with still images, back when only goog-remb was available. Since switching to transport-cc, I haven't encountered this problem anymore.
On OME, transport-cc is not yet available:
OvenMediaEngine/src/projects/config/items/virtual_hosts/applications/publishers/webrtc_publisher.h
Line 72 in 9d011aa
I wanted to know if there are plans to support it in future releases?
Additionally, and this isn't an issue with OME, I have the same problem with LLHLS. When I send a still image, the packets are very small, and the DNS resolution times and TCP request initialization have too much impact on the actual packet loading time. This also results in a low bandwidth detection with the hls.js library. Based on my tests, once a packet exceeds 30kB, the data starts to become relevant. Has anyone in the community encountered this issue before?
Beta Was this translation helpful? Give feedback.
All reactions