What is the best way to create a set of objects that are basically the same, but differ in their (animated) location and their image texture?

Approaches considered

It is possible to create a different material for each object, and a different texture with a different image for each material. However, for many objects I would have many very similar materials, which seems inefficient.

Another possibility might be to use one material and one huge texture atlas. However, each single image texture is already quite big in my case, so merging images into an even bigger file might make things hard to deal with.

The specific problem

I am creating a dashboard that will be placed over a movie, similar to augmented reality or a full-screen HUD. To have depth and animations, I use a couple of objects (currently eight) consisting of a one-face mesh each, holding one "layer" of the dashboard. The layers consist of images I prepared in Inkscape/Illustrator.

At the moment all layers have the same size, basically filling the whole screen. So yes there are a lot of empty spaces right now, which I thought helps to have simpler animations.

I am considering trimming those images and creating a texture atlas, if nothing else works better. However, I would like to keep the full-sized, prepared layers, as they might change in the future. And if they change, I don't have to change a UV-map, but only reload source files.

Update: Visual aids

The images contained in those layers about like the following. At the moment always two layers belong together, one holding a foreground (mostly text) and the other a background.

The layers side by side

The idea is, that the two layers in the leftmost column are always shown and one pair from another column at a time. Here animation comes in, but that's no problem. For presentation, the layers get stacked like the following.

How a viewer sees the layers

The effect I ultimately want to achieve is a three-dimensional interface like the following (exaggerated) image on the right side. This compares to a standard orthographic projection on the left.

Orthographic vs. perspective view

The issue now is that I expect more layers to come and I want to keep things scalable. How to keep required effort minimal when new layers are added?

Possibly related issues

  • $\begingroup$ Can you provide a image or even a sketch of what you are trying to achieve? I get mixed messages, from; overlays and watermarks to Iron man's glowing and highly active HUD. $\endgroup$
    – rob
    Commented Nov 11, 2019 at 14:03
  • $\begingroup$ @rob Thanks for commenting! I added some sketches to clarify my situation. $\endgroup$ Commented Nov 14, 2019 at 10:43
  • $\begingroup$ Looks to me that you should keep everything separate and the use the Compositor or VSE to combine everything at the end. This gives you the most flexibility on the workflow. You can use Linked objects to keep individual Blend files small while still being able to manage timings. $\endgroup$
    – rob
    Commented Nov 14, 2019 at 10:56
  • $\begingroup$ @rob With "keeping everything separate", do you mean having a separate material and texture per image? What do you mean with "combine everything"? $\endgroup$ Commented Nov 14, 2019 at 12:05
  • 1
    $\begingroup$ @rob Thank you for the clarification! The layers itself basically consist of a mesh with one rectangular face, onto which one pre-made image is mapped. So the layers itself are not complicated and don't require any special rendering techniques. Splitting them into multiple files sounds like more work at the moment, but I'll have a look into linking files and data in general! $\endgroup$ Commented Nov 16, 2019 at 10:32


You must log in to answer this question.

Browse other questions tagged .