Question / Help How do you capture sound twice when running two instances of OBS?

So after fiddling around most afternoon I finally have a setup that allows to record to hard drive at 720p60 3000kbps (will try to push that as I figure out more of how this works) and stream at the same time. I do a full screen preview and project the preview out to a virtual 2nd monitor. I can run a second OBS instance, capture from the preview window, and output to twitch at 720p30 1000kbps. I finally figured out settings that let me do that without pushing my CPU usage through the roof.

The only problem is all the sound from my capture device and microphone(s) is stuck in the recording OBS and I'm not sure how I'm supposed to clone it over to the streaming OBS. I use Voice Meeter and VB cable for mics so capturing mic sound from 2 virtual cables shouldn't be a problem, but i don't know how to grab sound from my elgato and desktop for the second instance. Is there something obvious that I might be missing here?

Just to cover all my bases, here are my log files.

Recording OBS
https://gist.github.com/5e19ba911f99de34d463

Streaming OBS
https://gist.github.com/b6525f4be7b0d0e9dc79
 

FerretBomb

Active Member
Just set the second instance of OBS to listen to the same channels as the main OBS.

This also has a LOT of potential to lead to severe desync, as the preview window is not necessarily real-time. Same with artifacting. Better option is to just run OBS with the -multi switch, and use Manycam to replicate any webcam inputs between both instances, controlling both through hotkeys. At that point everything is sync'd on both instances. Can just use two Profiles, one set for streaming to remote, one for the local-recording instance.

Also need to point out that 1000kbps is nowhere near enough for 720p@30 unless you're streaming something ultra-low-motion, like Hearthstone. May want to drop to 480p, or raise your bitrate to at least 1500kbps (preferably 2000).
Similar, 3000kbps isn't really enough for 720@60; want to be in the 4500 range if you're recording locally, or higher if you're using QSV, NVENC, or Ultrafast x264 to minimize artifacting as much as possible.
 
I'm not entirely sure what you mean by listen to the same channels. In the primary OBS it's pulling sound from the elgato capture device in the Scene, the mics are coming through VB cable A, and the systems sounds from default.

I should be able to have to second instance capture mic sound from VB Cable (i might have to set it to cable B) and I'm pretty sure I can capture system sounds (music I play in a media player or something), but how do I get it to listen to just the audio of the elgato?

With regards the bitrates. I'll definitely push the recording bitrate higher. The streaming bitrate was set low to account for my actual internet at home and the fact that the venue were we play games weekly has horrendous internet. Hence why I want to local record simultaneously. Actual live streaming from there may be down right impossible.
 
Just in case anyone stumbles across this thread via Google search the work around I found is to have the elgato output sound to desktop. Then I can capture desktop audio. It's not as elegant a solution as I wanted, but it works.
 
Top