Originally Posted by: DWAM I don't get it Richard!
Well, let's see if I can explain it differently/elaborated.
First of all, our ambition is to provide cost effective productions for our customers. Of course some clients have large budgets that could swallow extra work hours required if one builds up a parallell production using triggers, Multiview tricks, and u name it. But why should they pay extra if the result could be obtained more effectively? And thus give us time to do more produtions for other clients.
The point is that we do one main production (going to Output) using Cameras, Graphics, Transitions, Multiviews, Overlays, Triggers, Audio, and soforth. Quite frequently we add stuff to the setup when live. One typical case is when we realize we need an extra Multiview combining a Mentimeter result with shots of the audience. Another case is when we add a Virtual Input providing a zoomed-in shot of a Camera. In both these cases a "dedicated multiview", as you suggest, would complicate things - especially when live (extra mental burden for the operator).
Now, consider one of our cases (
documented on our web page), and the snapshot below. The upper right part (original production) was produced on one PC using all the stuff mentioned in the previous paragraph. That part shown in the snapshot is actually a combination of one cropped camera Input (the presenter) and another Input with a copy of the video signal going the the room's projector (her slides). As a matter of fact, we also used Merge to go from/to that particular shot to/from just the camera Input. Try doing that in a "dedicated multiview".
As examplified in my previous post the original production was then picked up in another PC. As this particular production was made in 2015 we did it via Fullscreen fed to a Magewell USB dongle on the second PC, where we added the sign interpreter and the subtitles. Nowadays we instead use NDI for PC one to PC two. But, imagine if we could do everything in one PC. The only missing piece is having access to Output. If we had that we could add a Multiview Input and with it do whet we did on the second PC. Then Fullscreen that Multiview to an external streamer, and record it in vMix's second Recorder.
Then to the risky part. What happens if a Multiview Input is added as a layer on itself? No one knows, because you can't - vMix simply does not allow it. I am pretty sure vMix could stop the operator from transitioning an Input that contains the Output to the Output. What happens if an NDI Input with Output as a source is transitioned to Output? It just looks ugly, but vMix does not crash and one can easily swith to another source.
I am pretty sure that if this Feature was implemented people would find all kinds of mindblowing uses.
Side note, I to my side don´t understand what OUT2,3, or 4 has anything to do with my request. Neither of those are directly available to Inputs.