I work with imported objects quite often. Many of these objects are curved but I need to flatten them out so it’s easy to generate the shape for UV mapping.
Unrolling the sop may be the wrong terminology but think of a flag blowing in the wind. I get the flag with a bunch of curves but I need to flatten it out before I can make the uv map.
Do you know how I can approach this? (the models can’t be exported flat.)
I’ve attached a candy cane as an example but the candy cane model looks much more complex than the objects I’m used to working with.
Hi Matt,
I thought this might be as simple as appending a spring SOP to the imported geometry since the spring SOP creates a cloth-like effect with a grid SOP. Or perhaps using a grid SOP as a template for laying out XYZ coordinates.
I agree that using other 3D environments may be a bit easier (depending on one’s skill level) but if there’s an easy way to tackle this in TD I want to find it.
Hi, if you’re working with lots of imported geometry, then you woukd be very well served to pick up some texture/uv mapping skills. I don’t mean this as an admonition, but just from general experience.
Often, by the time that playback media and model geometry has landed in my lap to get applied to some projection target, it has wound it’s way through a Workflow that is more or less complicated by a whole host of different apps. Raganmd’s suggestion of blender is actually really on point. I get a lot of eyerolls from colleagues who prefer maya or cinema4d, for my love of blender. But even they concede that blender’s UV editing framework is probably the most intuitive and flexible out there.
blender is free and you can quickly google up about a hundred uv mapping tutorials for it. adding some uv-editing and simple modeling skills to your arsenal is a smart investment of time and effort.
you can probably chain together a sop-tree that’ll do what you want, but you’ll wind up burning cpu cycles during playback by using the wrong tool for the job. render unto caesar.
The model also has a very high poly count which is undesirable if you plan on moving it around. I use blender - its great for most things. You might be able to export as quads and use a texture SOP (rows and columns).
Thanks for the replies and insight so far! I’ve attached another file for you to take a look at so you have a clearer understanding of what I’m working with. Crescendo 14x22.zip (92.2 KB)
I have Blender, AutoCAD, Vectorworks, and AfterEffects but that doesn’t mean I know how to really use any of them in depth. Photoshop is 2nd nature me and TouchDesigner, where 95% of my time is spent now, is very close behind Photoshop. I know my 3D skill-set is dismal at best, it’s a work in progress.
@flowb
My workflow in this instance is all about fast function rather than creative design. I get CAD files after an event has been planned and sold and am typically just providing the technical direction to make everything work within these set parameters. I export the staging CAD files and texture the projection screens in the drawing. Many of the set designs have unique shaped screen surfaces that need linear measurements turned into a pixel map for content development. I rarely have time to recreate the screens in other environments but it sounds like this is what I would need to do.
@asterix
Take a look at the attached model. The flatten/unrolled surface would be used to create the template for content development, it would not move or cook once I unrolled it.
I think my biggest shortcoming here is that I don’t know what type of object my screen model is to begin with. I’m sure it’s a mesh but do I need to convert it before attempting to unroll it?
Again, your insight is much appreciated.
Thanks for sharing with the noob!
This would be a great time to wax philosophical about the life of Technical Artists/Show Programmers. The way that one stands at the intersection of Creative and Technical Departments without really being accepted as either… sigh.
I’m looking at your model. Here’s another pseudo-philosophical rant: Scenic Designers don’t know what topology is. Which is to say, they want to show the internal, structural supports for their sets, but don’t care too much about overlaying six different pieces of geometry to do it.
Unlike them, we care, a lot. Because we need to render the thing. The first thing that I started to do when I looked at your model is delete the geometry for the internal supports. They are worthless to us, and our computers waste precious resources by even being aware of their existence. This highlights the fact that there are multiple faces and vertexes overlaid on top of one another. Often this happens because the set designer is automating part of their build out (ie: copy/pasting willy nilly without realizing that they are adding geometry on top of itself). We need to double check to make sure that we don’t have doubled up faces on our projection surface, or our renderer might decide to create ugly z-fighting artifacts there.
Are the edge supports necessary? The ones wrapped around the whole thing, or are you only planning on projecting on the stretched material on the inside?
Also, it might not be a terrible idea to do some kind of subsurf, to smooth out the screen geometry, unless it’s actually going to look like a bunch of triangular facets.
What was the model originally built in and how many steps did you go through to produce the 3ds?
Your original Geometry is on Layer1. I removed the internal support columns from it.
Layers 2-4 are the different parts of it split into separate models. They can be reassembled into a single model if needed but this lets you treat each model with it’s own simple set of UVs.
I did a quick pass to fix the topology on the faces. Blender has some quick utilities for merging duplicate vertexes which made it pretty easy. You will probably want to tweak the UV maps on the faces.
Or, you can just export the UVs as they are and make the content animators deal with it. This isn’t a bad idea honestly. Yes, you can streamline their work by giving them straighter edges to align to, but you can also do what the builder of the original model did and kick that can down the road. creschendo_topofix_v1.blend.zip (467 KB)
Hey flowb,
Your example has cost me about 20 hrs of Blender tutorials and exploring the application a little more in depth. THANK YOU!!
I believe I stated that I have to work with quick turn-around times and don’t have the time to recreate my 3D objects in other applications. Thanks to you I now realize how much quicker things can be by having the right toolset to aid in the workflow. I know I can do this in TD easily but I’ll figure that out in my spare time after I have a solid understanding of 3D editing.
I’m pretty sure that my exported UV layout is being produced flattened-out which is what I was after to begin with.
Question:
Do you know how to export the color grid with the UV layout? My exported image is just a solid green color.
