Resolume Forum

Hey everyone,

**Long post disclaimer**

So, this project probably isn't super interesting for you veterans, but this is the most complex mapping thing I've ever done. Only been doing mapping with Arena for about 2 years maybe, so I'm still decently new. Anyways, onto the project. I'd love you're guys' input on what I can do to make my composition/output better. So, here's the design:
This is being projected on with a single 16K projector. In addition to this, we also have two 8K projectors as side screens. The center screen design is a two layer thing. So the 5 panels in the back are sitting roughly 6 inches behind the 3 horizontal bands. The way I've mapped it so far works fairly well. I'm limited a little bit with what I can do, but not a ton. Only issue is, it uses a ton of layers and layer routers. I've got 20 layers going so far and to add more functionality and options, I could easily double that if I keep going the way I am now. Also, right now I don't have control over the side screens. I'm just doing the center screen.

So here is what I would like to be able to do with the whole setup.

-I need 2 outputs. One for the side screens and one for the center mapping setup.

-I want the ability to have a different image/video on every single surface, which would be a total of 9 surfaces (5 panels, 3 bands, 1 screen for sides). Ideally, each of these would be able to have a few layers of blended video. Layering is a big thing for me. Usually I will never layer more than 3 video files to make a comp with the majority of the time being 2 layers. I can see times with this setup where I'd have the need for more though.

-I also want a place to be able to have all the panels be their own blended surface, the bands be a blended surface, and then the ability to have blended video across the whole screen.

Now, here is how my composition is currently setup. Again, right now just the one output for the center screen. It's 1920x1080. It's setup with 20 layers. The bottom 8 layers are the "outputs" if you will. I've got 8 slices that take these 8 layers directly as their source. So I have a layer for every panel and band. That takes care of the need to have something different on every panel. It does not, however, fulfill the need to have blended video on every surface. It's just a single layer for each surface.

Next in the chain I have 3 blocks of 4 layers each. In each of these blocks the bottom is a layer for a black solid. Then 2 layers for blending video. Then a 4th layer that houses a layer router using the "layers below" setting. The 3 blocks are reserved for a full screen composite, a panel only composite, and a band only composite. That takes care of the ability to have blended video on the panels together, the bands together, and the full thing together. In the 8 layers at the bottom, I just put a layer router that routes to the top layer of those blocks where I need them.

Here's a screenshot of the advanced output screen and the actual output:
I hope that all makes sense. I'm thinking there's a significantly better way to do this. Maybe using more slices in the output settings. The trickiest thing about all this is retaining the ability to use blend modes and composite video layers together. If I didn't need that, 9 layers would cover everything.

I'm wondering if it's better that I've kept the composition 1920x1080 and used the layer to slice method, instead of making a larger composition and spreading things around while having more slices that are set to take the composition. Basically just creating spaces on the "artboard" for each section I need.

Who knows. I've got a decent system right now, but I'd love to hear how you guys would map it. I've also attached my composition should you like to view the madness. You can download the screensetup xml file here: bit(dot)ly(slash)ScreenSetupFile

-Tyler

To be honest, I read your post real quick. Just a pass of an eye ; )
I am used to work with such setups. There´s my 2 cents on it:

Yes, I would consider your side panels as a different input. A different system.
Now, on the center...
What I usually do is that I develop/produce custom visuals for each setup very often n my life. Meaning I keep my full 1920x1080 visuals untouched and I simply map, center or distort them to fit the whole big picture, no matter what.
Then, besides that, I calculate areas, pixels, and create a base template in photoshop, dividing in layers seprate areas/parts/modules/screens whatsoever.
Having that graphic information, I will produce one of two dozen visuals specifically to match those areas, all separate. The rest of the canvas for each area/video is empty - an alpha information.
Because of that, I will need just a couple of slices, sometimes just one. And it will provide me enough content and variations to make a very dynamic and impressive presentation.

Paulo

Here is how I would layout my composition (of course sizes are arbitrary):
This is based on a 2x2 1920x1080 comp which is 3840x2160 (4K).

I would create both Photoshop and After Effects documents with masked layers if you want different media for each panel/band. That way you can quickly render media (stills and video) to match this artboard.

This should let you do everything you're describing with 3-6 layers and let you manage the negative space (red in the png file) in your composition.