Possible Fix for Max Audio Buffer issues

[RavenStreamer]

Thought I'd post this for anyone troubleshooting a similar issue. My setup uses 3 cameras - one camcorder fed via HDMI to the HD60S into the laptop via USB, and two smartphone cameras using DroidCam for OBS. The camcorder feed is video and audio; the Droidcam feeds are video only so I needed the mic from the camcorder to capture the crowd sounds at live sporting events, regardless of which camera was active.

I HAD set this up as an Audio Input Capture source in OBS, selecting Elgato HD60S Game Capture from the audio source dropdown box. This led to Max Audio Buffer errors, which glitched ALL the audio, including my Voice-over audio (analog mixer to a Presonus Audiobox interface using the ASIO-OBS plugin as ASIO Source) and VLC Source music playlists. Huge Headache!

What I HAVE NOW done is scrap the Audio Input Capture source and set the Elgato Game Capture audio as a global audio device (in my case, Mic/Aux 1) in Settings. The Use Device Time stamp tick box is UN-checked, fwiw. I added the new audio source to the applicable scene and voila! A few test logs show no audio buffering issues! Yay!

I will continue to test and report back if anything significant changes, but this may help a few people solve their Max Audio Buffer problems.

 

[konsolenritter]

What I HAVE NOW done is scrap the Audio Input Capture source [...]

That would have been the most important hint to you. ;)

As a global audio device the benefit is the already-held-open-connection thru the usb. Otherwise each scene switch would enforce a close and re-open of that device (each time with a reasonable one-time-delay added, hence the safety margin added to the audio buffers).

The other possibility would have been if the hd60s was added to each scene (either in fore- or background) without closing it between scene changes, hence no dropouts in the audio stream.

 

[RavenStreamer]

That would have been the most important hint to you. ;)

As a global audio device the benefit is the already-held-open-connection thru the usb. Otherwise each scene switch would enforce a close and re-open of that device (each time with a reasonable one-time-delay added, hence the safety margin added to the audio buffers).

The other possibility would have been if the hd60s was added to each scene (either in fore- or background) without closing it between scene changes, hence no dropouts in the audio stream.

The downside of using a global audio device is having to mute it in the scenes I don't need it, but as most of the stream is actual game play where I need the crowd audio, it's not a huge deal. One more hotkey! :-)