Projecting onto a statue

Beginners
1

Hello,

I am undertaking an ambitions project, and I am new to TD. I have used madmapper and other software in the past, but I don’t think they will cut it for what I am tasked with doing next.

Here is the project:

1 statue, 50 feet tall. It is a person standing in tadasana (straight up with arms at the sides). We will be using 4 projectors (recommendations welcome!) that will likely be mounted on 20 foot poles at an X formation with the statue at the center of the X facing outward. I am also thinking that a + formation would be good, where there is a projector in front, behind and on each side, but i think that there will be shadows on the inside of the arms if we do it this way.

Anywho, the projector location is the least of my worries. What I am most curious about is mapping this sucker. I have two types of media that I would like to project on to the sculpture. The first is general textural media, where the entire body will be covered in a single video texture that has no real visual spacial anchors. That is no problem.

Here is where it gets sticky: I want to take video of people in different costumes standing in this position, and project THAT on to the statue, and to mix between those images and the textures.

Advice on how to capture the video, post process it, and mapping techniques would be very, VERY much appreciated. :smiley:

EDIT: I also have a geometrically accurate 3d parasolid file with the sculpture’s geometry. Could I use that in any way with Kantan mapper?

-t

2

Hi mate,
I would start with Camschnapper,make simpler case and continue to the more complex.
viewtopic.php?f=22&t=4225

3

Ya, using CamSchnapper might be the way to go. Kantan is more of a 2D masking tool. What you can do with CamSchnapper is render projector views of a textured 3D model.

So you’ll probably want to actually texture and figure out UV map of the statue model in a 3D modelling software like Cinema4D or Maya or similar, once that’s all setup, export the model with new UV maps you made, import both into TouchDesigner, and then you’d just texture the model as you would texture any other 3D object.

The other goal might be more complex, but could be do-able if you basically make a component in TouchDesigner that takes your filmed person, and basically crops and transforms all the different parts of that image and lines them up with the UV map you made while working in the 3D modelling software. Once it conforms to your UV map, you just switch between the textures using operators like Switch TOP and Cross TOP.

That make sense?