multi-track audio

[bravedog]

Hi, I'm new to OBS, successfully using scenes, have audio questions

I have not fully explored what already works in audio, except for reasons unclear to me, if I am running a Capture Device with audio input, that is my default streamed and recorded audio; but if I plug in a computer microphone, audio switches to that and silences the Device audio. I'd appreciate clarification of why that is, though guess it's not so important?

More broadly, I'd like to set up possibility of using audio from 3 [maybe more later] immediate sources, for both streaming and recording [must I do things differently for stream vs record?] with the ability to use zero, 1, 2, or all 3 at a time; of the Device Capture audio; the computer mike audio; and my main audio mix bus Focusrite 1 and 2, available in OBS [though why only 1 and 2, as opposed to Master, or choices?]

I am referring to

Advanced Recording Guide With Multi Track Audio

OBS Studio - Free and open source software for live streaming and screen recording - obsproject/obs-studio

obsproject.com

which unlike quickstart and general obs use pages, I find confusing.

1] Elsewhere I am cautioned not to use advanced features unless I am confident, which I am not. Will trying this screw up my audio settings and be hard to at minimum return to my present audio settings? Can't start experimenting until this one's answered!
2] It says "pick any encoder that is not "same as stream"". Really? It makes no difference which of the half dozen options I see in my OBS? That SEEMS illogical.
2] If I "pick any encoder that is not "same as stream"", will this disable or otherwise harm my audio when I do stream? Must I go into Settings and restore/change the earlier settings each time I want to toggle between stream-suitable audio and recording multi-track audio?

Once question 1 is clear, I can experiment and will revisit this thread with any questions which arise. Thanks!

 

[bravedog]

Also, just without going into Advanced, I don't understand obs audio handling, even referring to obs help resources

1]My OBS Audio Output Sources available in my Settings>Audio include 2 different Speakers options, Digital Output Nvidia, and Digital Device SPDIF. I understand these are kind of my desktop audio but don't understand distinction between them or so far see any difference in OBS behavior choosing differeing variations?

2] OBS Resource says can add Audio Output Capture and Audio Input Capture as individual audio sources which can be toggled in a Scene, so all such sources should be turned off in Settings>Global. In Settings> Audio, I have Disabled all audio global settings there, and have created a Test Scene to which I have added an Audio Input Device and an Audio Output Device

I guess I don't understand what OBS means by Global Settings. In my experiments I can Disable all or set various ways and see no different outcome, can someone suggest a specific setup where I can see such effects?

3] When loading my above-described Test Scene in Program, the capture card's audio moves in the Audio Mixer as Audio Input Capture 2 - but cannot be heard any more. What is going on? maybe it is being muted so it doesn't also appear in Audio Input Capture, which both displays in Audio Mixer and can be heard.

 

[bravedog]

All my Post #1 questions remain, but Post #2is are mostly answered, from that only I'm unclear if potential feedbacking Sources are always automatically muted for computer speaker output though not for recording/streaming - could swear not as I remember hearing mild feedback after which my computer audio level had reset to zero. Can't find relevant documentation - I also remain unclear of the differences if any among my OBS Audio Output Sources options referred to Post 2 #1

.I think I have answer for post 2 #3, it appears OBS has simply muted this scene automatically presumably because otherwise it would feedback. I don't find this discussed in documentation or google search yet.

For post 2 #2, Global Settings adds volume meters globally for Scenes and records them, this was unclear to me because they are automatically (?) muted in computer sound if might cause feedback [mike, system sound] so I believed them absent.

 

[bravedog]

OK found the Mute controls, one must put Program on that has the relevant Sources, then tight click the Source's volume meter in Audio Mixer [not in Source itself] click Advanced bottom option, there can set various Audio Sources to mute, output, or output plus monitor. Except for distinction between different Output settings Post 2 is all answered.

For Post 1, I've figured out how to add any combination of my 3 audio sources so that's the main thing. I guess I misunderstood what was meant by Multi Track recording, thought meant multiple simultaneous inputs but no, apparently mkv actaully has multi-track audio stems which can be independently manipulated in post. Thought all video had either mono or stereo audio only!

Spent like 8 hours on OBS today almost all on audio which for me was much harder to grasp than the video sources.

 

[Suslik V]

Any questions left?
There are more guides: https://obsproject.com/forum/resources/categories/guides-studio.8/

 

[bravedog]

Thanks! Looked through a lot of resources but this looks like more!

Ended up spending almost 10 hours on OBS audio yesterday and got most of it down. Can't edit the above questions down. My remaining audio questions are:

1] Elsewhere I am cautioned not to use advanced features unless I am confident, which I am not. Will changing Advanced Audio Settings away from Same As Streaming screw up my audio settings or streaming and be hard to at minimum return to my present audio settings?

2] In those Settings for Multi Track Adio Recording

Advanced Recording Guide With Multi Track Audio

OBS Studio - Free and open source software for live streaming and screen recording - obsproject/obs-studio

obsproject.com

It says "pick any encoder that is not "same as stream"". Really? It makes no difference which of the half dozen options I see in my OBS? That SEEMS illogical. Which should I actually pick?

3] If I "pick any encoder that is not "same as stream"" there, will this disable or otherwise harm my audio when I do stream? Must I go into Settings and restore/change the earlier settings each time I want to toggle between stream-suitable audio and recording multi-track audio?

4] My OBS Audio Output Sources available in my Settings>Audio include 2 different Speakers options, Digital Output Nvidia, and Digital Device SPDIF. I understand these are kind of my desktop audio but don't understand distinction between them or so far see any difference in OBS behavior choosing differeing variations? I have SPDIF output only on my Focusrite audio bus for desktop Ableton, which I can't easily route to OBS [I can input only Hardware Inputs 1 and 2, not DAW output], maybe on Focusrite I can connect SPDIP out to SPDIF in and so access Focusrite bus? Have to try today.

5] Interested in other possibilities to route Ableton audio to OBS [mostly for midi instrument availanility for OBS]. I have audio outs from the Focusrite but to get back into computer, would seem to need to route bak into Focusrite 1 and 2 or [maybe] SPDIF without creating feedback, to access in OBS.

 

[Suslik V]

1. Will changing Advanced Audio Settings away from Same As Streaming screw up my audio settings or streaming

I don't see "Advanced Audio Settings" in OBS. Please, clarify your question (maybe attach screenshot of the UI where you see this).​

​

Anyway, in Settings > Output > Output Mode: Advanced each tab (Streaming, Recording, Audio, Replay Buffer) modifies only its own parameters, literally "Recording" tab only modifies available settings for the recording (audio track number, output path, video encoder itself - here it can be named "Encoder" or "Video Encoder", if you see "Audio Encoder" - the encoder for audio can be set). Recording process in OBS can re-use already encoded video data from the streaming, thus without encoding the video twice the streamed data can be copied to the local disk (this is what the "Use stream encoder" setting exactly means). But usually, content creators running live shows (streaming) at lower quality, lower encoding settings, because network bandwidth is limited and in primetime network connection can be "overheated" (thus users will see only stutters, because of the network lag). While running theirs shows at lower quality content creators still can save (locally on theirs PCs) their video/audio tracks at higher quality. So, there is "Recording" tab with its own settings in OBS.​

2. In those Settings for Multi Track Audio Recording... It says "pick any encoder that is not "same as stream""...Which should I actually pick?

The one that has best quality/resource_consumption balance for you. Yes, you should decide it for yourself. Different hardware has different capabilities. Make few test recordings and make your decision.​

3. If I "pick any encoder that is not "same as stream""... will this disable or otherwise harm my audio when I do stream?

No, it is not. OBS sends only one audio track to the stream, its quality is defined by "Audio" tab when you are using Settings > Output > Output Mode: Advanced. All other tracks can be routed to the local recording (via checkboxes at "Recording" tab). All audio for stream is encoded into AAC format - this by standard, local recordings also uses AAC, because it is compatible with .mp4 media container and has nice quality/resource_consumption balance. Thus, at "Audio" tab you only can set bitrate for each track. If you define track#1 for streaming - you can set it at 160kbps, if you define tracks #2, #3, #4 for local recording - you can increase bitrate up to 320kbps for each of them for better quality.​

4. ...Audio Output Sources available in my Settings>Audio include... Digital Output Nvidia, and Digital Device SPDIF...

In general, the Settings > Audio > Global Audio Devices was made for lazy people. This sources of audio (sound from this devices) will appear in each scene you made (wish you this or not). Usually, professional users of OBS set this all to Disabled state and adding each audio source to each scene manually. Thus, "Audio Input Capture" source that represents your mic can be added to scene where you speaking. "Audio Output Capture" to scene where your PC playbacks something, both "Audio Input Capture" and "Audio Output Capture" to scene where you speaking and your PC playbacks something. Finally, you adding no other sources to your "start streaming" scene (where usually only some of your intro media playbacks) etc.​

5. Interested in other possibilities to route Ableton audio to OBS...

I cannot give you exact settings for your hardware and setup. Sorry.​

 

[bravedog]

Thanks for your helpful reply!

1 and 3] By 'Advanced Audio Settings' I meant to refer to those in the link in question #2, you're right these are within Output Settings, thought link would make this clear, sorry for lack of clarity then, Anyway I think you answered my worry - these Settings can easily be changed/rolled back and won't affect my Stream settings.

2] Hm ok varies from rig to rig, must test if want to distinguish. Probably will stick w default then.

4] I too have set all Global Settings to Disabled. Was unclear at 1st how they worked, now find them unhelpful. - The Digital Output Settings options is a different question, but I've tentatively concluded that for me all these outputs [SPDIF, speakers, etc] are the same, in theory [don't know or have strong interest in how] these could be set to route different output buses within my PC.

5] I've succeeded in routing Ableton output to OBS by running a cable from Focusrite Control hadphone #2 I don't use, to an input on my desktop, which as marked 'Microphone' I epected to have wrong impedance and be mono, but which seems to use my stereo signal fine.

So all these audio questions seem answered!
------------------------
I continue to spend a lot of time on OBS, but down to 6 hours yesterday. New minor questions which have arisen include, why does Mask filter seem to apply differently to a video source, from one applied to a Group containing only that video source? I do understand creating and using Groups to move things together but apparently don't understand Groups completely. Won't post separate question for this, will experiment more.

I'm also unsure whether to record at 30fps [now] or 60 fps, have read strong opinions both sides, thinking to stay at 30 as my video has major changes between frames. Can test but only down the road when I have better more challenging samples to test with readied.

Lastly I'd like a way to move different layers of my running stream or recording in real time, but haven't googled this yet, seems maybe impossible.

 

[koala]

@bravedog The best thing to increase your knowledge is to experiment with OBS and its settings and how changes actually work out if you try some recording. It seems you're already doing this.
I'd like to give some generic info you might have missed so far. Now you investigated all the settings thoroughly, look back to the big picture.

OBS can be configured to have a number of audio sources. These sources can be a global audio source, or a scene specific audio source. A global audio source is available (is playing) in every scene, a scene specific audio source is specific to the scene only where it is configured, so it's only active if this scene is the active scene. Global audio sources are configured in Settings->Audio->Global Audio Devices. Here, you can select Windows audio devices. Anything that is recognized by Windows as audio device is available here. Whatever you make Windows play or record on its audio devices, you can capture it here.
If you have an external audio mixer, it will show up as Windows audio device and can be configured here among all the other Windows audio devices.

Everything else is a scene specific audio source. Capture devices usually come with their own audio source. You can also add a Windows audio device as scene specific audio device instead of a global audio device: add a "audio input capture" or "audio output capture" source. The browser source includes audio. And so on.

Every audio source that is made known to OBS can be seen in Edit->Advanced Audio Properties. Make sure to switch to the correct scene if you want to see a scene specific audio source.
If it comes to where these audio sources go in your recording and your streaming, look at the Tracks matrix. Here you can define which audio source is going to which audio track. If you check multiple audio sources for the same track, they are mixed and output together to that track.

Finally, you need to define where the tracks should go or should be ignored. If you're using the recommended simple output mode, simply go to Settings->Output. For recording, there is a "Audio Track" setting where you can define which tracks should actually be written to the video file. Use multiple tracks only, if you intend to postprocess your recorded audio and intend to mix them later. For streaming, Track 1 is hardcoded to go out to the stream. It's only one track that can be used for streaming.

In advanced output mode, you can configure which audio track instead of track 1 is going out for streaming. For recording, it's the same as in simple output mode. If you stream to Twitch, there is a special feature. You can select a second Twitch audio track. This second track is the track that will appear as audio track in your VODs. The primary audio track is what appears live. So you can have different audio in your VODs than in your lives.

There is another thing with audio: monitoring. If you configure a source as "monitor off", it goes out to the activated tracks. If configured as "monitor only", it doesn't go out to the activated tracks. Instead, it's output to the Windows audio device you configure in settings->Audio->Advanced->Monitoring device. You use headphones here, usually. Don't also record these headphones to avoid feedback loops.
The setting "Monitor and output" does both: it sends the audio of that source to the tracks as well as to the monitoring device.

 

[bravedog]

Thanks! Yes I have been experimenting a lot. My experience has been that I found OBS use of layers and visual sources pretty clear and understandable, but its treatment of audio I found confusing and even quirky with audio controls spread about and even hard to find - just my experience, and I've got a grasp now I think. Personally I don't like Global. I am just adding all audio sources I ever want, clearly labeled, within each scene and have them turned off by default till wanted.

For me Ableton [music DAW] is not configured to play through my desktop audio and has a dedicated audio driver so its audio was not available to OBS without my now-successful special cabling.Ableton in/outs through a Focusrite Scarlett 18i8; only app Focusrite Control's s hardware inouts 1 and 2 are available to OBS for technical reasons unclear to me. Since midi instruments are DAW outputs not reachable through hardware inputs 1 and 2, I had to and did find a separate routing.

I'd like to be able to move my OBS layers [visual sources within a scene] freely manually in real time like in Preview, during realtime stream/record, but am thinking this is impossible. except partially through mt today's OBS focus, plugins, which appear to give some ability but only in pre-programmed ways, with use of hotkeys. There are half-dozen plugins I'll try today, not just Motion.

IOS plugin looked exciting though its OBS page says $16 now $70 or $11/mo so probably won't use.

 

[Suslik V]

...why does Mask filter seem to apply differently to a video source, from one applied to a Group containing only that video source?

Maybe you discovered bug in OBS. Make new thread about this (post your examples, screenshots, write step by step how to reproduce the issue etc etc).

...I'd like a way to move different layers of my running stream or recording in real time

Just use mouse. OBS don't care if you streaming or not. Scenes can be modified in any time.
Later, when you learn OBS more, you'll discover Studio Mode of OBS (which is next step in modifying scenes).

 

[bravedog]

ha ha I have been constantly using Studio Mode and it didn't even occur to me I could simply manipulate placement in real time when not! This is a perfect example of how my wrong assumptions trip me up. Thanks for helping clear this instance!

FWIW for ios addition to OBS, I tried out half dozen apps from app store and ended up buying IVCam for $25. It seems to work well and frees up desktop capture as a separate Source, earlier I accomplished this using screen mirroring. Droidcam also worked about as well but to remove watermark etc is subscription - I won't buy apps on subscription model...

I've downloaded and installed half dozen plugins ok, guess I have to watch some videos to understand configuration...

 

[bravedog]

...why does Mask filter seem to apply differently to a video source, from one applied to a Group containing only that video source?

"Maybe you discovered bug in OBS. Make new thread about this (post your examples, screenshots, write step by step how to reproduce the issue etc etc)."

can't rule it out but maybe not? For instance this tutorial video seems to show the same difference achieving a helpful effect

- YouTube

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

www.youtube.com

Maybe it's connected to the fact that a filter applied to a video source, produces a different effect [at least sometimes] than the same filter applied to a Scene containing only that Source; for example today I discovered the Recursion plugin filter works

Recursion Effect

Plugin for add recursion effect to a source using a filter. examples of usage: Donations You can donate to me via PayPal, GitHub or Patreon.

obsproject.com

applied to a Scene containing only one [video] Source, but doesn't work when applied instead to the Source itself within that Scene.

Somewhat confused.

 

[Suslik V]

The video you posted shows two sources: cropped_source (above) and full_source but "green-screened" (below).
It has nothing to do with the grouping - just simple trick above/under. In text form similar effect was described in: https://obsproject.com/forum/threads/choosing-an-area-on-screen-for-chromakey.134874/

There, the groups were used because "Video Capture Device" source in OBS was reused (reference copy) and reference copy of the source cannot has different sets of filters - they always the same, only Transform (size) can be different. So, to apply different sets of filters to the same source (literally same source, source that has one feed, same input signal, same properties) you encapsulating that source into a Group. This trick allows to you to apply different filters to different groups (and thus to all theirs contents). If you don't do this (encapsulating into groups) then when you adding any filter to any referenced copy of the source then same filter will be "added" to all copies of this source (actually, nothing is added - it is still the same source, it's a programming stuff...).

About the plugin - better ask in the plugin's thread (author of the plugin usually knows more).

 

[bravedog]

The video you posted shows two sources: cropped_source (above) and full_source but "green-screened" (below).
It has nothing to do with the grouping - just simple trick above/under. In text form similar effect was described in: https://obsproject.com/forum/threads/choosing-an-area-on-screen-for-chromakey.134874/

There, the groups were used because "Video Capture Device" source in OBS was reused (reference copy) and reference copy of the source cannot has different sets of filters - they always the same, only Transform (size) can be different. So, to apply different sets of filters to the same source (literally same source, source that has one feed, same input signal, same properties) you encapsulating that source into a Group. This trick allows to you to apply different filters to different groups (and thus to all theirs contents). If you don't do this (encapsulating into groups) then when you adding any filter to any referenced copy of the source then same filter will be "added" to all copies of this source (actually, nothing is added - it is still the same source, it's a programming stuff...).

About the plugin - better ask in the plugin's thread (author of the plugin usually knows more).

Thanks for your reply. It may point to a whole new important misyunderstanding of mine where you say or seem to say that any filters applied to a video Source will automatically apply to all instances of that video Source anywahere in OBS [except size transform]. This surprises me and I'll have to experiment with it today. [I already wanted to test to what extent transforms of audio or video Sources added to a Scene as a copy from another Scene also affect that Source globally - maybe the same way you describe? - but I've been thinking all instances have independent life after creation]. Anyway that certainly would explain why Groups were used in the above example and why I hadn't understood that.

PS Yesterday rather than on OBS I focussed on the Panasonic MX-70 mixer. I have the manual which explains use but not meaning of functions and I have basic grasp but am confused by eg aspects of its Chroma Key and of even its implementation of A/B buses [when Internal [color or pattern] is chosen in A or B the previously chosen video source button for that bus stays lit though now seemingly inactive] etc. Anyway I don't expect OBS forum answers to that! but hope for suggestions of forums to ask those kinds of questions? Analog equipment like this seems little used; I am posing such questions on reddit/videosynthesis and videohelp.com but seem to get no responses, looking for more active analog video forums? thx

 

[Suslik V]

...any filters applied to a video Source will automatically apply to all instances of that video Source anywahere in OBS [except size transform].

This true for Paste (Reference) and Add Existing.
For example, if you'll try to rename one copy of the source - every reference copy to this source will got same (your new) name.

Let's say, OBS can make clone of the source and independent copy of the source. Clones - require less resources, works instantly.

 

[bravedog]

OK verified all pasted Sources seem to behave as you say EXCEPT for copies of Display Capture which seem to hold independent filters!. Though actually desktop cannot be copy/pasted but are of identical Source and appear as Display Capture X.

For audio Sources even if added separately and named differently, applied filters seem to apply to all instances. Video Sources added separately by + rather than copy/paste, do not display video from that Source which is only visible where 1st created or copy/paste instances - why not? Anyway this makes filter behavior for such cases irrelevant...

So, to apply filters and presumably other effects to single instances of Sources, they must be applied to the Scene or group containing that instance of the source and NOT to the Source directly, necessitating creating such Groups or scenes as needed; otherwise application will be global. Except for Display capture!

This is not at all the behavior I was assuming, so thanks for the comments.

 

[toplachi]

I have not fully explored what already works in audio, except for reasons unclear to me, if I am running a Capture Device with audio input, that is my default streamed and recorded audio; but if I plug in a computer microphone, audio switches to that and silences the Device audio. I'd appreciate clarification of why that is, though guess it's not so important?

I am sorry I haven't read the whole thread, I am not sure if this is already answered. Just a little guess that may also cause this issue:

Could your settings for MIC/Auxillary Audio on OBS Audio Settings (or maybe from the Capture device audio properties?) be set to "Default"? This might possibly mean that the Windows operating system's selected audio input source at first is coming from your capture device and then when you plug the computer microphone, Windows switches to the microphone port? Not OBS fault.

I wonder if choosing the capture device for "MIC/Auxillary Audio" under OBS audio settings would solve the problem

or

setting it to DISABLE and manually adding an AUDIO INPUT Source that is from your capture device to every scene who needs it

Screenshots of your audio settings could help us better understand your case.

 

[bravedog]

Thanks for your reply! Yes this must have had something to do with setting at Default - I have abandoned use of Default and of Global settings. I have largely solved my audio questions including all those above but as I experiment, continue to generate new questions. Currently trying to unserstand some of the plugins but I haven't eplored tutorials enough to ask questions here.

Small side note, I have a 2nd successful method of adding Ableton output to OBS - when I updated my Focusrite drivers and set to visible, now its SPDIF In [set to loopback] appears in my list of audio sources.in OBS.

 

[Suslik V]

...Video Sources added separately ... do not display video from that Source which is only visible where 1st created - why not?

"Video Capture Device" in OBS has one feed. And this feed is unique - in most cases it is web camera device. When device is busy - you can't get second copy of the digital signal from it. Thus, OBS show to you nothing - device is busy. Most devices unable to produce two (or more) digital signals simultaneously. For example, it either 1920x1080@30fps or 1280x720@60fps but only one at a time. Disable all sources that uses same device and OBS will show to you what this device is transmitting right now. Or, you already know it, use clone (reference copy) of the device.