Thanks for your help AaronD.
I haven’t played around with it a lot yet, but I wanted to ask this ahead of time. We’re a church that uses one computer to run PowerPoint type presentations and then I have another computer solely for streaming. My goal with this plugin was following a video played on the presentation computer, the streaming computer would switch scenes. We’re pulling in the presentation and video via NDI. Is it possible to do this or will this only work through one machine. Thanks in advance!
That rig with NDI and Adv. SS is actually for
a church service!
Faithlife Proclaim sends lower-thirds lyrics and Scripture to OBS via NDI, so that those stay synchronized with the audience displays. The NDI feed is actually full-screen over the cameras, and completely transparent except for those cases.
One of the things to watch out for in Proclaim is that new content, by default, is not transparent to that feed, which makes it cover our cameras, so we have to manually fix it every time we add something new. For the times that we forget to do that, I also have a pair of hotkeys to hide the NDI source and show it again.
For a direct copy of the audience displays, which includes any fullscreen content like pictures, videos, announcement slides, etc., I use a physical signal capture instead. Better control that way, for a dedicated Broadcast Engineer. The only exception to that is the Welcome Video, which is played directly in OBS from a local file, to the stream only.
Sound comes from a post-fade aux send in the
FOH console, that feeds a
USB line-in, as OBS's only audio source except for that video. NDI is there, but always muted.
Actually, it's more complicated than just a post-fade aux: it's actually several post-fade aux mixes that feed several matrices, or not, and one of those matrices goes to the stream. Another matrix goes to the lobby and hearing assistance transmitter. The primary difference there is whether it has the room mics (stream) or not (lobby/assist). Lots of compression involved too, at several different points along the way, to end up with a broadcast mix instead of a live one like what goes to the room.
For automation, I do a lot with Adv. SS, but the part that I think you're describing is and probably always will be manual. I consider it to be part of the art of live media, and "You can't automate art!"
For me, Adv. SS sets up the first scene and starts the stream automatically at 10:00am on Sunday, automates fading the audio in and out from the FOH board at various times (out for the Welcome Video and back in, for one example), turns the Noise Gate off during the service proper and back on again at the end, and other miscellanea that I started to forget out of boredom once I had it more-or-less figured out. And it handles the ending slides as an escapable sequence of scenes once I show the first one, which ultimately has it stopping the stream and closing OBS so that the trainees that I hope to have aren't playing whack-a-mole with multiple windows that don't come back automatically.
But seeing something relevant on a different source/scene and showing it to the audience, is part of the art to me.
You can't automate that, partly because the criteria keeps changing! Automation only works with well-defined triggers, and well-defined actions in response to those triggers. If you have both, then it works well. If you're missing one or the other, then you need someone who appreciates the art to do it manually.
You might try the Video condition, to try and do some sort of machine vision or pattern matching on the picture, or you might generate a control signal of some kind from your presentation software that Adv. SS can respond to (Websockets exist now in the standard release, and have for quite a while, and there's basic MIDI in the latest testing version, which you can get by scrolling up in this thread), but that's probably about it, short of a dedicated operator.
github.com
