Proper Implementation of Lee Griggs' Volume Displacement. Need help!

[futuremotionrecovery]

I've been somewhat obsessed lately over Lee's hyper-real volume displacement images and artwork. If you don't know what I'm talking about, take a look through his gallery here:

https://leegriggs.com/volume-mesh

He was also kind enough to author a small glimpse into his process with this tutorial, and this video.

The very general idea of the effect is simply converting a polymesh to a volume, cranking the density, and then displacing it with a texture. I'm sure he's also further manipulating the volume in other creative ways.

Despite a decent amount of information on the subject, in practice, it's very hard to execute at the level and detail that Lee does. But I am very determined and tenacious. 🙂

I emailed @Lee Griggs yesterday and he suggested I make a post here about what problems I was having with the technique. So here I am!

Today I went through the tutorial found here, and documented every step along the way. I did this for reasons twofold: 1. So other artists can follow along, and 2. So @Lee Griggs can tell me what I'm doing wrong and how to improve on my results.

Edit: I've created a fully commented Imgur gallery since I hit the image/attachment limit. Find it here:

https://imgur.com/a/ZW9hKsW

This is where I stopped. I just couldn't figure out how to improve the look any further. The end result was kind of cool I guess, but nowhere near Lee's artworks.

Reflecting on the process:

1. I need more surface detail. The surface should look more like geometry than plumes of smoke. Lee goes into that here. But I have no idea how to get there.

2. The smoke is too even, and lacks depth. I think it's a combination of my lighting, volume density, or some other magical techniques I don't quite understand yet.

3. Look at the extreme detail in this image:

I am at a complete loss as to how that kind of micro detail and contrast was infused into the image. While I feel like I learned a bit about volumes and displacement, I still remain extremely confused. There has to be more to it, or I'm doing something wrong.

Either way, here's my final graph:

One thing to note is that there is no camera projection here in the graph, because I didn't know how to implement it. I'll implement it in my next attempt at this technique.

And many apologies for the length of this post, but I felt it was necessary to break things down step by step for clarity.

Thanks!

Jay

[lee_griggs]

Hi Jay, did you try the camera_projection? You could also use the Maya projection (set Projection Type to Perspective) as used at the bottom of this tutorial. You could also insert a ramp before the range to control where the effect appears. Other than that you seem to be doing everything right. It really depends on the quality of the texture map that you are using. Trust me, I have tried many textures that look bad before finding one that works well in a given situation. Please share your results when you have something that you are pleased with.

[futuremotionrecovery]

Thanks! I am going to try all your suggestions soon. You're right about the source image. I actually came across a combination that looks much better (although I'm not completely happy with it):

I don't like how the face itself is all displaced. I need to try the ramp I guess.

[lee_griggs]

Yes, the ramp will help. It takes a lot of experimentation to get something that works. Keep at it!

[futuremotionrecovery · Accepted solution]

Some more progress. I think I'm getting the hang of things.

Nodetree:


I think there is still a lot to learn about this technique. Admittedly I don't even know exactly what I'm doing in some cases. Just a lot of tinkering to get good results.

[lee_griggs]

Looking good Jay.

[futuremotionrecovery]

Thanks Lee. Your work is an inspiration.