rockbottom
Active Member
From my perspective OBS handles audio pretty much flawlessly. If there's an issue, it's hardware, driver, settings or something else in the audio chain that is the cause.
If OBS handles everything perfectly then why is causing issue, but another freeware software such as Audacity works with the audio card, not to mention that the Ableton is pretty much solid piece of DAW, which costs around $750 also works fine?
PS: regarding the newer audio driver it's not working with my setup, I've already tested that version 10 days ago, and performed tests - with the same outcome, so I followed the advice of Behringer team member and used the previous driver.
PS: regarding the newer audio driver it's not working with my setup, I've already tested that version 10 days ago, and performed tests - with the same outcome, so I followed the advice of Behringer team member and used the previous driver.
Last edited:
rockbottom
Active Member
Sorry it bust your bubble but it's not OBS. If you don't remove the Behringer & test using the on-board audio, I'm removing myself from this thread.
Suslik V
Active Member
@rockbottom meanwhile, can you perform some tests (because you have nice rig).
Here is the link to the "test files" and steps: https://github.com/obsproject/obs-studio/issues/8064#issuecomment-1597505629
If you'll succeed in reproducing the issue (sometimes it takes 120 minutes to get data flow anomalies) - you can write in related thread on the forum: https://obsproject.com/forum/thread...terval-with-application-audio-capture.166472/
and then you can complete the steps from https://github.com/obsproject/obs-studio/issues/8064#issuecomment-1519111627 on github itself, if this wasn't done yet.
Yes, its not easy task and it is not related to issue mentioned in this thread but at least you may try (if your system is stable). Maybe my tests are too easy task for the powerful PC and thus they are useless, and you'll get perfect results (no "crackling"), here I don't know, really.
Here is the link to the "test files" and steps: https://github.com/obsproject/obs-studio/issues/8064#issuecomment-1597505629
If you'll succeed in reproducing the issue (sometimes it takes 120 minutes to get data flow anomalies) - you can write in related thread on the forum: https://obsproject.com/forum/thread...terval-with-application-audio-capture.166472/
and then you can complete the steps from https://github.com/obsproject/obs-studio/issues/8064#issuecomment-1519111627 on github itself, if this wasn't done yet.
Yes, its not easy task and it is not related to issue mentioned in this thread but at least you may try (if your system is stable). Maybe my tests are too easy task for the powerful PC and thus they are useless, and you'll get perfect results (no "crackling"), here I don't know, really.
rockbottom
Active Member
Alex Presso, the guy that started that thread (8064) has sample-rate mismatches in his logs. Nobody told him to fix it.
Last edited:
rockbottom
Active Member
His issues aren't being caused by OBS either. Further testing would be needed, but I'm fairly certain if NDI was removed from the audio chain, the crackling would have stopped.
Bad plugin: 21:56:29.263: [obs-ndi] hello ! (version 4.10.1)
Bad plugin: 21:56:29.263: [obs-ndi] hello ! (version 4.10.1)
Last edited:
I can confirm that the the audio stuttering has been neutralized. I've recorded successfully 20 minutes with OBS without any drops.
Current OBS settings:
/Output/
fragmented mp4
video encoder - nvidia h.264
audio encoder - ffmpeg aac
audio track 1 (only)
automatic file split - manually
Rate control: CBR
Bitrate 16000 kbps
keyframe int = 0
preset P5 - slow (good q)
tunning - high quality
multipass mode - two passes (quarter reso)
profile - High
Look-ahead - NO
Psycho visual - Yes
GPU - 0
Max b-frames 2
Audio track 1 - 256 bitrate
-
/Audio - 48hz sample rate / stereo
global audio devices: Disabled (all)
-
/Video - Full HD - 60 fps
/Advanced settings
General - Above normal
Video -
renderer - Direct3D11
Color format nv12
color space rec.709
color range - limited
sdr w. level - 300nits
hdr n.p. level - 1000nits
Audio sources:
Input - default (doesn't matter actually)
-------------------------
And the big question: WHAT WAS CAUSING THE audio stuttering?
OBS. Audio sources are asynchronous.
The number of samples provided from the audio sources is sometimes insufficient or too many,
which causes audio glitch as #4600 and audio/video desync.
Who discovered it: https://github.com/floele from Germany & Norihiro Kamae from Japan https://github.com/norihiro
How Has This Been Tested?
The code is tested on Fedora 34 with MOTU M2 and recording a tone signal.
Also analyzed step response and frequency response of the feedback system with my SPICE model.
How the issue has been handled:
Frequency response of the open loop transfer function. Gain curve crosses 0 dB at 5.55 mHz (inverse of 3 minutes) and shows suppression for higher frequency components to eliminate fluctuation such as USB and load of the host. At 5.55 mHz, the phase curve shows 60 degree phase margin so that the feedback system is stable.
Simulated and measured step response of the system. This figure shows step response how the compensation is applied (vc1) and another internal node in the lag-lead filter (vc2), measured with MOTU M2 with Fedora 34. It estimated roughly 4.5 samples are are missing for every second. The implementation gradually started to add samples and reached maximum compensation -5.3 samples per second after 80 seconds from the start, then goes to the steady state compensating 4.5 samples per second. The other node vc2 slowly track vc1 so that it eliminates offset at the steady state. The curves of measured and simulated model are well matched except between 100 seconds to 300 seconds. I think we can rely on the model so that the gain and phase curves are well simulated and conclude the feedback system has enough stability.
More information about the issue here:
github.com
github.com
HOW TO FIX THE ISSUE:
Simply download the PR of norihiro on GIthub, and replace the OBS files (you can backup the entire folder if u like first) and test.
github.com
Congrats to floele and norihiro for discovering and handling this thing!
Current OBS settings:
/Output/
fragmented mp4
video encoder - nvidia h.264
audio encoder - ffmpeg aac
audio track 1 (only)
automatic file split - manually
Rate control: CBR
Bitrate 16000 kbps
keyframe int = 0
preset P5 - slow (good q)
tunning - high quality
multipass mode - two passes (quarter reso)
profile - High
Look-ahead - NO
Psycho visual - Yes
GPU - 0
Max b-frames 2
Audio track 1 - 256 bitrate
-
/Audio - 48hz sample rate / stereo
global audio devices: Disabled (all)
-
/Video - Full HD - 60 fps
/Advanced settings
General - Above normal
Video -
renderer - Direct3D11
Color format nv12
color space rec.709
color range - limited
sdr w. level - 300nits
hdr n.p. level - 1000nits
Audio sources:
Input - default (doesn't matter actually)
-------------------------
And the big question: WHAT WAS CAUSING THE audio stuttering?
OBS. Audio sources are asynchronous.
The number of samples provided from the audio sources is sometimes insufficient or too many,
which causes audio glitch as #4600 and audio/video desync.
Who discovered it: https://github.com/floele from Germany & Norihiro Kamae from Japan https://github.com/norihiro
How Has This Been Tested?
The code is tested on Fedora 34 with MOTU M2 and recording a tone signal.
Also analyzed step response and frequency response of the feedback system with my SPICE model.
How the issue has been handled:
Frequency response of the open loop transfer function. Gain curve crosses 0 dB at 5.55 mHz (inverse of 3 minutes) and shows suppression for higher frequency components to eliminate fluctuation such as USB and load of the host. At 5.55 mHz, the phase curve shows 60 degree phase margin so that the feedback system is stable.
Simulated and measured step response of the system. This figure shows step response how the compensation is applied (vc1) and another internal node in the lag-lead filter (vc2), measured with MOTU M2 with Fedora 34. It estimated roughly 4.5 samples are are missing for every second. The implementation gradually started to add samples and reached maximum compensation -5.3 samples per second after 80 seconds from the start, then goes to the steady state compensating 4.5 samples per second. The other node vc2 slowly track vc1 so that it eliminates offset at the steady state. The curves of measured and simulated model are well matched except between 100 seconds to 300 seconds. I think we can rely on the model so that the gain and phase curves are well simulated and conclude the feedback system has enough stability.
More information about the issue here:
Random short audio dropouts when recording or streaming · Issue #4600 · obsproject/obs-studio
Operating System Info Windows 10 Other OS No response OBS Studio Version 27.0.0-rc2 OBS Studio Version (Other) No response OBS Studio Log URL https://pastebin.com/WfdNU2QN OBS Studio Crash Log URL ...
github.com
libobs: Implement asynchronous sample rate conversion for audio by norihiro · Pull Request #6351 · obsproject/obs-studio
Description This PR implements asynchronous sample-rate conversion by utilizing a soft compensation feature in libswresample. A digital approach of asynchronous sample-rate conversion was propose...
github.com
HOW TO FIX THE ISSUE:
Simply download the PR of norihiro on GIthub, and replace the OBS files (you can backup the entire folder if u like first) and test.
libobs: Implement asynchronous sample rate conversion for audio by norihiro · Pull Request #6351 · obsproject/obs-studio
Description This PR implements asynchronous sample-rate conversion by utilizing a soft compensation feature in libswresample. A digital approach of asynchronous sample-rate conversion was propose...
github.com
Congrats to floele and norihiro for discovering and handling this thing!
rockbottom
Active Member
Behringer, plain & simple.
rockbottom
Active Member
All that filter is doing is changing the sample rates on the fly so the drops are eliminated. A workaround for your flawed rig, nothing more.
rockbottom
Active Member
So what's your fix for Audacity, just record @ 16-bit? As long as the Behringer is restricted to 16-bit, 24 & 32-bit recordings with Audacity will have gaps...