Pablo Dobarro's master plan for sculpting and painting, development news

There’s already a patch made to use OIDN in the viewport, you can try it out in here

2 Likes

Dyntopo Roadmap

https://developer.blender.org/T73934

PBVH_TEXTURE Design

https://developer.blender.org/T73935

12 Likes

I’ve feared Pablo would break my heart like this for a long time. :cry:

But seriously, Voxel Remesh is just not on par with ZBrush Dynamesh. It still changes a sculpt too much each time you remesh your model. Some parts lose details, other areas show a crudely angled polygonal structure, and parts that are close to each other such as fingers risk melting together. I also dislike having to keep track of polygon stretching and manually remesh frequently. It feels like taking a step back when compared to dynamic topology sculpting.

Cranking up the voxel detail level is undesirable because it makes the mesh less suitable for manipulation (smoothing gets harder, etc.).

I guess an important cause of the voxel remesh issues is the lack of a really efficient smoothing / averaging / rounding algorithm to translate the OpenVDB voxel cage to a rounded mesh.

The best OpenVDB voxel remesher I know for Blender is still OpenVDB Remesh by @ambi. It features really good smoothing, which brings it almost on par with Zbrush Dynamesh.

Concluding: ZBrush Dynamesh is better than Voxel Remesh, and ZBrush Sculptris Pro offers dynamic topology sculpting with no noticeable speed difference from the Dynamesh workflow, while Blender’s Dyntopo will probably be set aside to perish. I think that will make Blender lose a powerful, dynamic creation method, and I will personally seriously consider a return to ZBrush for sculpting again.

19 Likes

On the one hand, I agree with your statement, but on the other hand, I don’t. Because if there’s an alternative that can be improved than maintaining something old.Why would you panic?Pablo never offered anything without an alternative. So I’ll wait for his decision.

4 Likes

I expected this. He added many useless features for dyntopo users and now wants to remove it.

1 Like

Do not post that news in this thread:
https://blenderartists.org/t/dyntopo-tests/550302

We could get a river with crying users.

5 Likes

Agreed. I’ve been lurking these threads for a while and this Roadmap post pushed me to make an account to post. I really can’t see using Voxel Remesh through the whole workflow myself. It’s great for setting up meshes, but a real pain for me otherwise. Ive given it tries since the Sculpt Mode Features branch and through 2.8x releases and I just don’t find it very usable. Trying to find a good Voxel size for a sculpt (down to 4-5 decimal places at scale), then it smoothing things, losing details that you had already worked on, etc. and at the end you end up with many more vertices than I would have needed if I had just done spot detailing (eye lids, mouth area, nose, ear, etc. as an example) with Dyntopo.

With Dyntopo I can go constant detail, pick a resolution and work on areas, spot fill other areas with a different resolution. I can then turn off Dyntopo when I’m happy with the resolution in that area, sculpt it some more without any performance issues if I start hitting higher numbers of verts. No progress lost, no undoing just because the 3rd decimal place in the Voxel menu was off so I lost all of my edges and creases.

I don’t agree with the claims that it Dyntopo is “unusable” for anything beyond simple sculpts. Sure, if you’re trying to create and push hundreds of thousands of verts at the same time it will lag. I’ve never felt the need to do that while Blending other than just messing around.

We felt this was coming but it’s still a punch in the gut.

7 Likes

To be honest, I understand the Pablo. He doesn’t want to do double work instead of adding new features. He himself says he’s looking for ways to fix bugs in dyntopo in new functions rather than doing anything else.

5 Likes

Reading through the whole thing, it appears trashing dyntopo for something else is just one potential solution he made, not the solution. Users appear to be responding already on removal not being a good idea.

One of the other points is optimizing and improving the whole dyntopo system itself, which would be the one I would endorse. It would be great if dyntopo also started preserving existing data like vertex colors and vertex weights (even if they still have to be destroyed if a spot is sculpted on).

6 Likes

I don’t know what you’re all attached to removing dyntopo. If there’s an alternative and an alternative that’s better than dyntopo, I’d be happy to. This is better than to develop two paths (I mean voxel remesh and dyntopo) and do twice as much work, it is better to have one path.

Yes, if I understand you correctly this is very much the way I use DynTopo as well, like a local multi-resolution where I enter the Dyntopo mode mainly to add/subtract polygon density in areas, then leave Dyntopo mode and continue sculpting.

I VERY much would hate to see Dyntopo go, and I agree with Metin_Seven, voxel remesh is not a substitute for Dyntopo, the resolution needed to keep details when remeshing quickly makes the mesh unsuitable for anything other than further detail work in my opinion.

3 Likes

Pablo saying:

  • When developing new features I’m always testing them with the Voxel Remesher workflow, but some people keep using Dyntopo as their primary dynamic topology system. This makes that almost all bugs I fix during the development of new features are related to Dyntopo compatibility and not to the feature itself, which does not make much sense as Dyntopo is something that is not active being developed.
1 Like

Why can’t anyone hear that replacing it with another type that’s can be better than dyntopo is not bad.

1 Like

While I usually don’t use the raw Dyntopo mesh as the final one for renders, it is an incredibly useful way to quickly bang out a complex model which you then retopo with the shrinkwrap modifier. However, even the raw mesh is good for renders if you don’t plan to use texture painting or armatures.

The only thing that could be seen as a superior replacement is actual volume sculpting, because you can then make holes and easily make details such as membranes (for wings and such). If anything, Dyntopo shouldn’t be removed until we have volume sculpting and a fully working multires system. Voxel remeshing wouldn’t quite cut it because it’s all grid based with no directional fields.

3 Likes

Its not better. Dyntopo>>Dynamesh.

2 Likes

He offers an alternative in the form of volume sculpting. I’d like to take a look at this alternative before screaming about how bad the removal is.
Pablo offers three solutions. I agree with the second solution as dyntopo with volume sculpting.

1 Like

Faster Sculptris style dyntopo with a specialized triangle mesh acceleration structure would probably be the optimal solution.

8 Likes

And the second solution is dyntopo with volume sculpting, what do you think?

…is not a replacement for a mesh based sculpting solution

All right. We’ll see what the developers decide. But I generally understand why Pablo offered such solutions. It’s twice as much work instead of making new features.