Large resoltuon notch block multiple UV cameras

Hi, how would one best approach an interactive notch block that spans a whole room of LED tiles and virtual pixel gaps. upstage screen, ceiling screens, sides etc.
It’s on 2 pixels maps, of around 4k each.
Played back on a D3 server.

UV camera? Is it possible to output 2 UV cameras from a notch block / D3? For each pixel map?

How well will it run at those resolutions on D3?

Haven’t tried this before. I have my usual go-to methods i’d use to make this as rendered content, but as a notch block it’s something thats brought me here.


The UV camera can capture multiple objects, you just need to make sure the uvmaps don’t overlap (e.g, split them side by side). Then just render out the image at 4kx2k and you should be fine.

as for how well it runs, depends on your content :slight_smile: the overhead of the UV cam itself is minimal.

– Ryan


Would you suggest this being the best solution, using a single UV camera that has the UV’s for each screen section laid out in a tile space?

My scenes are general fly through scenes that I want to be immersed across the LED, flying through tunnels along splines for example.

The problem is the ceiling alone is made up of 20 or so pods with a total pixel map resolution of 5k x 2.5k and with virtual gaps to make a circle a circle, would be more like 7k x 7k.

I’m assuming that I can add in 3D model of the screen layout thats UV mapped to match the pixel maps, and then do a render to texture to feed my typical cameras view (front, pointing up, sides etc) in the scene over to be a material on the 3D screen layout scene.

Would an alternative option be adding in a camera for each LED pod / Tile? One for upstage, one for pod one in the roof etc. It would be a lot of cameras…
Is it possible to have lots of different cameras in the scene and have each camera show up as an output in D3 that can be moved and arranged server end to match the pixel map?
I’d have to somehow set the resolution of each camera…

It’s complicated. apologies if none of this makes any sense.

I’m not entirely sure if I follow, but I would think the UV cameras are the way to go as opposed to loads of cameras. If you need content rendered from a particular perspective, you can always re-project that content onto the surface before it gets onto the UV camera - example below.
ReProjectionExample.dfx (21.0 KB)

– Ryan