Cineversity

Hello! Working on a project currently that has me using the Voronoi Fracture object, but have hit a slight snag in executing what I’m aiming to do.

I’ve been able to assign different colors to the pieces of the fractured object with a Mograph Selection tag when used in conjunction with a Plain effector (Position, Scale, Rotation boxes have been unchecked) whose Color Mode has been set to Custom Color.

The snag I’ve hit is that, ideally, I’d like each of the fractured pieces who’ve had a color assigned to them to remain that color regardless of their movement as the sources for the fractures have been keyframed to move. So, while each piece has a color assigned to it, the colors shift wildly as the animation plays back in the timeline.

I’ve tried phrasing my query a handful of different ways but have yet to find the solution that I’m seeking. If anyone knows of a solution, I’d love to hear it!

Additionally, I’m happy to provide my project file if it would assist in troubleshooting.

Thanks for your time!

Hi, helloimneil,

Would you mind to share a scene file?
Just a cube and any relevant part of your set up. Please no 3rd party items at all, thanks.

I’m happy to look into it.

For now, please select the Voronoi Fracture> Transform> Display>  Index. The index needs to be the same to work with the Selection, the Display option allows for the inspection of such.
Here is a scene file that shows the problem. Is that the same that you encounter?
https://www.amazon.com/clouddrive/share/VgfzXIi9RKT0tZEhUYnJAXjZXun10JBqmYngA8wSQ2r

One of the most stable results is available based on manual “tracking”.
https://www.amazon.com/clouddrive/share/3VXFkbVqZOcZeeOqrl3goS1mxJPNPQf6miJogy0tsrZ
Which indicates as well an option to “track” the id, and set via Xpresso a MoGraph Selection.

Perhaps a much simpler option is given by moving Fields with the source, which seems the simplest, but there might be some rare limitations at times.
https://www.amazon.com/clouddrive/share/rPYhd7pWY1AWH5mdjXPCELJRLVVow593Mdf5ZymeAM

All the best

Edited Sunday

P.S.:

If manual adjustments of the ID are the target, this set-up might streamline the work. The example is prepared for ID 0-10.

Scene files and screen capture
https://www.amazon.com/clouddrive/share/Duop91REX6MIiV0rcs2IkXhRTWj0TnjpCV17vg0yZT2

Hey Dr. Sassi,

Intended to respond sooner, but the weekly D&D session went longer than anticipated.

Here’s a link to my project file: https://drive.google.com/file/d/1hZk5MUW58myAtE6wZX6oaZGkK_pZhfjn/view?usp=sharing
I’d like the assigned color values on frame 0 to remain constant throughout playback.

Anyway, in response to your question, the issue I’m encountering is the same one as was present in the first project file link.

The second link you sent is a little confusing to me as I’ve never done manual tracking and am not familiar with the workflow. Whatever technique was used there seems like the same one I should employ on this project file.

Thanks for taking the time to link these reference projects!

Hi, helloimneil,

Thanks for the file, as usual, this is a much clearer representation of the challenge. Yes, I always like to have a scene file, so, thanks!

The first file is pretty much what you have encountered as the problem. Since the amount and position fo the clone fragments is continuously changing, there is no quick remedy.

The tracking is, in a nutshell, to change the ID number in the formula field when the scene changes it. Since there is no way (AFAIK) to have any DIY code producing it, clones fragment are created, or they stay or become deleted, while the overall amount changes as well. Even my little radio-button quick tracking options would need a lot of more set up time, to work for all the parts, even then it is quite some work.

Another approach would be, to place a camera on top of it, and get an image sequence that has an image for each frame. Then you load this sequence in Ae and pain fill the needed color in each frame. This Roto-Work is usable in a Shader Effector.

Both methods produce no future proof results, as with each change, one has to start over.

Besides that, and having mentioned Roto as an option already, I also tried to draw a Spline around the frame zero clones, but also here, to change the spline for 240 frames is not something one would do just for fun.

Of course, there is a method given to go from the source to the clone fragments, but I do not have any idea to set this up and describe the color of the resulting fragment, based on point pairing.

I need a little bit more time, to think about it, but honestly, this time I fear there is no quick workflow for this.
For now, Roto is my only answer to this triple source set up. (With a single source I would have used perhaps a Sorting/object, but with three, there is no simple way). Sorry.

My best wishes

Dr. Sassi,

Thank you for taking the time to examine my project file, it really does mean a great deal to me.

“...to change the spline for 240 frames is not something one would do just for fun.” Ha! I couldn’t agree with you more!

For now, I may dog-ear this element of the project and come back to it later. Perhaps the client will be willing to reconsider their desire for fixed colors!

Again, thank you for your time and wisdom!

Thank you for the feedback, helloimneil, very kind of you.

Yes, the scene you have shared is quite a challenge for anything manual.

I have gotten similar questions in the past, and as usual, I know, the problem comes back until solved. Sadly, I can’t even think of a coded idea, as in, give any clone a unique and very long ID, or other more complex setups.

I have set up a little screen capture, to showcase the parts of the problem. The source points create pairwise the edges or borders. These points sources can change (swapped) and move, while the result seems to stay or not. Well, that is again a cryptic nutshell definition. The screen capture might help here to illustrate the complexity:
https://www.amazon.com/clouddrive/share/WWiIKJ9sGJlzLX5kB67ebKfDGMbP0FVd9j1VsnU2Bel

Yes, I will keep this in mind, and I hope some ideas will come up at one point.

Enjoy your exploration

P.S.: In case you like to “track” the clones, here is a version that allows for ID-numbers up to 110.

Screen capture and scene file
https://www.amazon.com/clouddrive/share/MzgUpiC8lM4ahsReOh5FPkVq6oxCQBOtVrFA7zZZHaQ

Awesome! Thanks Dr. Sassi, I’ll take a look at that scene file!

I found a… workaround… of sorts, to this particular problem. It is by *no* means a solution, but, I found that by adding a material to the voronoi fracture object with a gradient in the color texture channel and adding more knots to the color gradation allowed for the colors to remain static as the animation runs through the timeline.

Thanks for all of your help!

Nice that you found something that comes closer to what you want, helloimneil.

However, I can’t reproduce this here. A color channel with a color gradient (plus lots of knots) keeps the clone’s color selection as you need it? I must miss something.

Care to share a file, or if not for the public, can I send you an upload link?

Cheers

Hey Dr. Sassi, I think my description of the end result may have been phrased incorrectly.

By creating a gradient and applying that to the voronoi fracture, as the pieces move throughout the timeline, the gradient’s colors remain static. They do not adhere to each individual piece, nor are the individual colors of the gradient being assigned and remaining on the fracture pieces, but rather, the gradient as a whole remains constant across the entire fracture object.

Here’s a link to my project file, no doubt this will demonstrate the result I’m talking about better than I’m explaining it:
https://drive.google.com/file/d/1I7OUOAF8nfWMqXLpI4oalmNOZBXWku1G/view?usp=sharing

Thank you very much, helloimneil, for the extra effort.

Yes, a standard projection of a single texture (i.e., a gradient) will be not affected by the changes of the fracturing process.
I understand your “… by *no* means…” remark now more clearly.

You’re right, this demonstrates is with no doubt. I was pretty focused on the initial request, keeping the colors from frame zero on the fragments/clones.

I will continue to explore it, and at the moment I get a little idea of how this could be solved, I will file a suggestion. But I guess more people have to request this option, to get it higher up on the wish list.

So, thanks again, and if there is anything else, I’m happy to look into it.

Have a nice evening

P.S.: Here is another example demonstrating the problems of Source Points vs Clone Axis.
https://www.amazon.com/clouddrive/share/CTk4UAZPqrwbWQmtPkSVBdEqGtqBaNuyLeyu6PNZWe0