OBS doesen't RECORD the screen

[EvoL000LovE]

Hi all,

I had installed OBS for streaming, so I installed it with the streaming configuration, now I'm looking to record a video with audio, but the video is not recorded, it appears black.

I tried with the "screen capture" and then with the "window capture" fonts, but with both is the same result: dark video. Instead of the preview where is visible the screen and all what I do. (I have only one monitor, Windows 10 updated, last version of OBS installed)

Here is the log, what can I do?

https://obsproject.com/logs/Dy6OcFdllFxNJBau

 

[koala]

You have 3 different display capture sources. They stomp on each other. Remove all of them, then add a single new display capture. This one should work. If you need another display capture in any other of your scenes, add the additional capture sources with the "add existing" option.

 

[EvoL000LovE]

You have 3 different display capture sources. They stomp on each other. Remove all of them, then add a single new display capture. This one should work. If you need another display capture in any other of your scenes, add the additional capture sources with the "add existing" option.

Hello Koala and thanks for the answer,

as you can see in the screenshots, there is only one display capture source in the fonts tab, the other is the audio source, where should I look for the 3 sources that you are indicating?

Thanks.

 

[EvoL000LovE]

Oh, probably you are talking about the tab in this screenshot, there are only 2, I just tried both of them (only 1 for every try), but nothing changed. How can I delete those 2 to have another try?

Thanks!

 

[koala]

Scenes and sources are global. All sources in all scenes are always active. Check all your different scenes for display capture sources.

 

[EvoL000LovE]

Scenes and sources are global. All sources in all scenes are always active. Check all your different scenes for display capture sources.

Well thanks for the answer, nothing had changed erasing scenes, it was for the transactions, deleting the transactions it went fine!

Now there is another issue: the recorded video is like in the clouds (see the attached images) and the audio is with echo. But the set up for the audio is at best values: 48Khz Stereo and Metrics I tried with all the values.

 

[AaronD]

Using other apps (like Zoom) with OBS complicates the audio a lot. It's certainly manageable, but YOU have to know and manage it. It's not plug and play. The two common problems are not getting a connection at all, or creating a feedback loop from too many connections. That feedback is where the echo comes from.

I have a rig that uses two simultaneous instances of OBS, an external audio processor (Ardour), and an online meeting. (Jitsi, but Zoom or anything else works the same way) It works perfectly - been using it twice a month for about a year now with no changes at all: just run my custom setup script and there it is - and I've attached the diagram that I have for it.

It'd be good for you to keep a similar diagram for yours. Do the engineering based on the diagrams, and then make the settings match that. Don't design it directly on the settings.

I've also attached the flowchart for that setup script. Same idea there: do the engineering on the documentation, make that work in your head, both as a user and as a technical system engineer, and then make the product match the documents.

Some of the specifics of my flowchart are because it runs on Linux, which itself is in the process of making its audio a lot better. In the meantime though, I'm using the older, reliable-though-wonky method, which the script needs to manage. So that's what Jack and PulseAudio are about, and the bridges between them.

 

[AaronD]

One of the problems that often comes up with a meeting like that, is that the mic picks up the speaker through the air. Often better than it picks up the person that it's supposed to pick up. Normally, that's a feedback loop, except for a bit of processing called an Echo Canceller. That takes two inputs - the final speaker and the raw mic - and gives one output - the mic minus the speaker - so that you can use the system somewhat sensibly as a non-technical user thinks of it. Most meeting platforms have an echo canceller built into them, so you don't have to think about it.

The problem is, as soon as you do anything beyond the arrangement that the echo canceller is meant for, it doesn't work anymore. It can only adapt so far, in terms of matching its copy of the speaker with what the mic picks up. So if you use your own mic and/or speaker, or use your own processing so that the echo canceller gets something other than the final speaker and the raw mic, then you're back to effectively not having it at all, same as a live concert or similar event.

I solved that on my rig - the one I posted the diagram for - by ducking the processed mics under the meeting return. That's the red arrow on the top right of each page, that takes the meeting return as a control signal back to the ducker that the mics run through. That ducker is actually a compressor on the mic mix, that is set to use an external side-chain (the control signal), and absolutely squash the mics with very aggressive settings.

The upshot of that, is that the remote people (or their noise) can override the local mics, but the local mics can't override the remote people. Or in other words, the remote people have priority over local. That's fine for what I do with mine, but think it through for yours and see if you need to do something different. Be creative, know the gear, and think that through too.

 

[MattEatsMochi]

I'm having this issues as well. Blank scene, one display capture, the capture is visible when you look at preview but not on the canvas. I did find that it works when I play a Youtube video so this might be related to GPU (chrome is using GPU acceleration)? Commenting here so I know when we figure out a solution.

 

[demy1942]

I have the same problem, has anyone tried another streaming service to test?