3D Projection Mapping- Is it all about UV Mapping?

[videoboy]

Hello!

 

Trying to get a very broad level of understanding of how 3d Projection mapping works in Watchout. Is all of the mapping basically done in the 3d Model application when setting up the UV map? Is that where I would define regions and scaling and position of media I want to project on to the object?

 

For example, lets say I have a complex object I want to project on to using multiple projectors. Let's say it's the Statue of Liberty. Once I have my 3d model of the statue, do I then need to use UV mapping to define where on the statue I want to project? Let's say I only want to project on the face so it looks like the face is moving. Do I need to define that face region in the UV map. I can't, in Watchout, just put a projector projecting on to that region and restrict it to the face alone?

 

From what I'm gathering from documentation, Watchout assumes a perfect map is already created and just simply applies it on and plays it back and has no ability to manipulate what is being projected from within Watchout?

 

Hope that makes sense, just looking to clarify how it works. Thank you!

[Josef Swanberg]

Yes, if the whole statue model is one single texturable area, you need to know where in the UV map the face is.

 

In the 3d model application it is often quite simple to generate an image of the UV map mesh, such as CarMesh.tif, which is seen in this video:

 

http://www.dataton.com/3d-texturing-using-virtual-displays

 

From that image you should be able to distinguish the different areas of the 3d model, in this case the statue

[Mike Fahl]

Is all of the mapping basically done in the 3d Model application when setting up the UV map?

 

Mapping is a multi-step process. The first step is mapping content to the "virtual 3D model" appearing in WATCHOUT. This mapping is all done through UV mapping in a 3D program. See below for the following steps.

 

Is that where I would define regions and scaling and position of media I want to project on to the object?

 

Yes. You can set up several independently texturable areas (sub-meshes) in the 3D program, as well as assign UV to those there.

 

Let's say I only want to project on the face so it looks like the face is moving. Do I need to define that face region in the UV map.

 

In this case, I'd make the face as a separately texturable area (a sub-mesh) in the 3D program. Apply UV to the face there, and make sure a test still image maps properly in the 3D program. If this looks good, it should work the same once the model and image are brought into WATCHOUT and applied there.

 

Take a look at the mapping related tutorials found here. They should hopefully help clarifying some details:

 

  http://academy.dataton.com/cookbook

 

As mentioned above, mapping onto a physical 3D object is a multi step process:

 

1. Create a matching 3D model in a 3D program, possibly with multiple, separately texturable areas.
2. Apply UV coordinates in the 3D program, and test with some sample texture images.
3. Bring the resulting 3D model file into WATCHOUT.
4. Apply your sample texture to the model in WATCHOUT, to make sure it looks OK in the stage window.
5. Add a "3D Mapping Projector" to the Stage window in WATCHOUT, pointing it roughly at the virtual 3D model in the Stage window.
6. Point your corresponding real-world projector at the physical object so it covers it the way you want.
7. Use the semi-automatic calibration feature of WATCHOUT to position the projector in the stage window soit matches the position of the real-world projector.
8. Add more sophisticated texture content, either by applying a video or by creating a multi-layered texture in WATCHOUT using a Virtual Display.

 

Hope this helps.

 

Mike

[videoboy]

Thanks guys, I get it now. Very helpful!