I do the following

1) Have a camera input (ie from webcam)
2) Create a simple title (second entry)
3) Create a multiview (blank and add the 2 previous as layers).
4) Add the resulting composite in a multiview (a caller window in a multiview)

Now.

a) Zoom, Pan, Borders work fine.
b) Cropping does not work.

Now why I want to do that.

I want to be able to create a "Caller Output", that I can use in small or larger boxes, and turn on and off the title when I want with shortcuts.

I tried adding the title in the original input (camera) as a layer.
Cropping works, but ONLY to the main input, so still it cannot be used.

Creating a virtual camera from the layer also would not work (both not cropping, and not updating when turning the titles on and off)

Any ideas?

Not cropping looks more like a bug to me.

Thank you. The mix workaround seems a weird solution!!

A similar weird one is NDI out of that mv and back in as new input

If you have moere than 3 cameras, then you may want to look into a virtual set and switching the input dynamically. Are you putting more than 3 cameras on screen at one time?

I do video backgrounds in virtual sets all the time, where did you run into that issue?

That's correct, but that's why I asked if you had more than 3 cameras at once on screen.

You can set a shortcut to change the camera within a Mix and use that same mix with multiple cameras in the same virtual set.

Not to hijack the thread, but we're in the same scenario as the OP and would love to be able to do this or have a viable workaround if one exists. Slightly different application on our end, but same needs. Mixes don't solve it because we run out of

I'm betting that there are a lot of operators and programmers that are suddenly using Vmix now with the current state of things that would really love to be able to do this. We're in the same situation as the original poster and cannot use mix layers for this due to too many inputs that need to be composited upstream of the final "3rd-tier" multilayer input. We need a "2nd-tier" multilayer input where we can turn layers on and off dynamically and then have that input feed several more "3rd-tier" inputs where scaling and cropping of the composited 2nd-tier input are different in each one. We shouldn't do this on the 1st-tier input due to the multilayers not scaling independently from the Main layer; it makes a mess of things. That "1st tier input is where we would correct color, but not sizing, positioning or crops.

Effectively, we're trying to use the "2nd-tier" inputs to get around the limit of the number of mixes available and create a sort of upstream router that can be automated be external control. To your point regarding number of cameras onscreen at once, we have up to 8, and need an easy way to toggle any of those camera feeds to a headshot or similar with a single button punch that then appears across a lot of different inputs simultaneously.



But, you're saying it's always been that way, and we're now on V24....if there are no crops or positioning changes on the sub multi views, it would be great to be able to do it on the final one. I can't imagine that there aren't more people trying this; it seems to be a common question between all of our local operators.