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
on PC#2, Configure OBS (30.2.3) to publish to MediaMTX (for example using RTMP or WHIP, it doesn't seem to matter)
on PC#1, Configure OBS (30.2.3) to read from MediaMTX (localhost) as a Media Source (for example using RTMP or SRT)
INCONSISTENTLY Observe audio glitchy, choppy and Video is fine.
This issue occurs intermittently without any change to OBS/PC settings. It happens the moment a publishing PC starts sending data. Sometimes it can be resolved by restarting MediaMTX itself, and sometimes (like we experienced today) we can fight with this problem for an hour restarting and rebooting everything and never get it to work correctly (at which point we fall back on using xObsBeam, which wreaks havoc on our LAN capacity but it works flawlessly.)
We've experienced this using RTSP, RTMP, and WHIP writing/publishing.
We've experienced this using RTMP and SRT reading/streaming.
The issue seems to be within the MediaMTX service, since it can be observed from VLC and OBS as a "reader."
We stream from one, two, or three source PCs (streaming team) to a dedicated "stream pc" responsible for running overlays, switching, and simulcasting to multiple platforms. This happens with any number of PCs, and when it happens the audio is affected over all streams passing through MediaMTX.
We started trying to use MediaMTX for our streaming approximately two weeks ago. Out of 11 streaming sessions this has happened to us three times.
Again, no change to our hardware or software setups. We leave everything untouched between streaming sessions except for stopping/starting streams in OBS (it keeps our temps/bills down to not have encoders running 24/7.)
The PCs are all 13th gen Intel 13700Ks with 4070Ti cards and 64GiB of ram, plenty of headroom end to end. CPU and GPU usage looks great. We've tried source streams both CBR 4Mbps and VBR 6-12Mbps, and depending on the publish protocol have experienced this both with AAC and Opus audio codecs (does not seem to be source-codec specific.)
As this point we've rolled back all of our setups to to use xObsBeam because of this problem. We've had two late streaming sessions and (today) one missed streaming session because of the unreliable audio and the need to reconfigure everything back to xObsBeam. Hopefully this issue can be identified and resolved because when MediaMTX works it works great.
All that aside, as a 30+ year veteran software engineer I know this isn't a lot of help (we have no logs nor pcap to share.) You probably aren't going to be able to repro otherwise you would have experienced this issue yourself by now, and if you can't repro you can't debug, and if you can't debug you probably aren't going to solve this problem.
Hopefully you find a way to repro. <3
Server logs
No response
Network dump
No response
The text was updated successfully, but these errors were encountered:
Which version are you using?
1.9.3
Which operating system are you using?
Windows amd64 standard
Describe how to replicate the issue
This issue occurs intermittently without any change to OBS/PC settings. It happens the moment a publishing PC starts sending data. Sometimes it can be resolved by restarting MediaMTX itself, and sometimes (like we experienced today) we can fight with this problem for an hour restarting and rebooting everything and never get it to work correctly (at which point we fall back on using xObsBeam, which wreaks havoc on our LAN capacity but it works flawlessly.)
We've experienced this using RTSP, RTMP, and WHIP writing/publishing.
We've experienced this using RTMP and SRT reading/streaming.
The issue seems to be within the MediaMTX service, since it can be observed from VLC and OBS as a "reader."
We stream from one, two, or three source PCs (streaming team) to a dedicated "stream pc" responsible for running overlays, switching, and simulcasting to multiple platforms. This happens with any number of PCs, and when it happens the audio is affected over all streams passing through MediaMTX.
We started trying to use MediaMTX for our streaming approximately two weeks ago. Out of 11 streaming sessions this has happened to us three times.
Again, no change to our hardware or software setups. We leave everything untouched between streaming sessions except for stopping/starting streams in OBS (it keeps our temps/bills down to not have encoders running 24/7.)
The PCs are all 13th gen Intel 13700Ks with 4070Ti cards and 64GiB of ram, plenty of headroom end to end. CPU and GPU usage looks great. We've tried source streams both CBR 4Mbps and VBR 6-12Mbps, and depending on the publish protocol have experienced this both with AAC and Opus audio codecs (does not seem to be source-codec specific.)
As this point we've rolled back all of our setups to to use xObsBeam because of this problem. We've had two late streaming sessions and (today) one missed streaming session because of the unreliable audio and the need to reconfigure everything back to xObsBeam. Hopefully this issue can be identified and resolved because when MediaMTX works it works great.
All that aside, as a 30+ year veteran software engineer I know this isn't a lot of help (we have no logs nor pcap to share.) You probably aren't going to be able to repro otherwise you would have experienced this issue yourself by now, and if you can't repro you can't debug, and if you can't debug you probably aren't going to solve this problem.
Hopefully you find a way to repro. <3
Server logs
No response
Network dump
No response
The text was updated successfully, but these errors were encountered: