Sculptris - Mudbox for Blender?


#1

Hello guys,

I haven’t been here in a while. Been working on something and it’s been pulling a lot of my time from the boards.

I recently heard of a free open Sculpting Software… like Mudbox… called Sculptris (http://www.sculptris.com/)

Has anyone used this software with Blender? I know some colleagues who are telling me this can be used to produce some really nice normal maps… But has anyone done a proven workflow with Sculptris?

How does it compare with Blender’s own Sculpt Mode? And what about UV information? Does the UV information survive a “.obj” import/export movement between Sculptris and Blender? Or can “Bake Selected to Active” work even if say the Sculptris imported version has no UV information?

I think this can be an important development for those of us producing material with Open-Source Software. What do you think?


#2

Pretty much everyone is aware of sculptris,

Has anyone used this software with Blender? I know some colleagues who are telling me this can be used to produce some really nice normal maps… But has anyone done a proven workflow with Sculptris?

How does it compare with Blender’s own Sculpt Mode? And what about UV information? Does the UV information survive a “.obj” import/export movement between Sculptris and Blender? Or can “Bake Selected to Active” work even if say the Sculptris imported version has no UV information?

Sculptris works by locally tesselating wherever the user is sculpting based on a tesselation factor, so detail can be added where needed, and the user doesn’t have to worry about not being enough poly density to get good detail.

The brushes in Sculptris are implemented quite well, and are currently better behaved than those in Blender. However I suspect that with the GSoC project for sculpt we will pull ahead of it and others regarding Brushes.

We have 4 GSoC projects that relate to functionality around sculpting and painting

  1. Multires improvements - including masking support
  2. Sculpt tool improvements - huge number of brushes and interface improvements
  3. Paint tool improvements - will improve stroke, generalize the sculpt related brushes to be able to work with painting (ie rake), add multichannel support and layers support
  4. Pen tablet and space ball input improvements - vastly improved interaction when using these tools

Also Farsthary has done a dynamic tesselation implementation for blender sculpting.

Also another of the GSoC projects is a quad dominant remeshing algorithm.

In short I would be surprised if Blender hasn’t surpassed what is available in sculptris by the end of summer.


#3

You might be interested to read this about the another development being done in Blender which does the same thing as Sculptris.

Blender Nation Link


#4

I for one am ecstatic that this even exists. Free software of this level that one developer put together as a personal project?! this rocks!!!

Add to this that Blender gets more and more pumped everyday! Why choose? this is all good. I can’t wait to see what Blender has in store.


#5

Sculptris may be good for someone who uses Lightwave for example, but with Blender it has to compete with integrated sculpt. The idea of dynamically tessalating the geometry is not as great as many may think. It causes problems in animation and doesn’t as of now allow to differentiate resolution in different parts like nose and cheek which need different level of detail.


#6

You would use a re-topo’d mesh for animation and from what I’ve seen you can have varying levels of tessellation on the mesh, as it dynamicly divides the geometry based upon the HQ setting (Blender doesn’t have this by the looks of it, but Sculptris does).


#7


#8


#9

My testing I couldn’t export textures with sculptris. So presumably that part of the obj export for sculptris is broken. Presumably he will fix it when back from vacation.


#10

Guys, what we have found is that Sculptris (or even Blender’s Sculpt Mode) is great at producing Hi-Res “Baking Sources”. As long as the Lower-Res Baking Target’s topology is pretty strong, you can have very clean projection and zero increase in the number of vertices and not worry about rigging issues. As a “Baking Source” the Sculptris/Blender Sculpt result can produce any maps and you don’t have to think about .OBJ anymore. All you have to do is import the Sculptris mesh.

You must also be very very meticulous about what details are Actual Topology in your original model (and thus won’t be sculpted) and what does become sculpted. Good design is key. Random and ill-planned abuse of Sculpting (either in Sculptris or Blender) will result in bloated sculpted meshes that will cause slow-down and hang the system.

The main difference, though, right now between Sculptris and Blender is the speed and convenience at which you can produce really fine “chips in armor/engraved script lettering” style sculpts to make a good “Baking Source”.

In Blender 2.49b and the current 2.50 versions available it takes a lot of workarounds using applied black-and-white Displacement Maps with the sculpting brushes and GIMP-ing over maps to get a very nice, economical result.

In Sculptris (and similarly Mudbox), it’s just a very organic process and you’re finished quite quickly.

The next part of the current workflow for us is to use the heavy sculpted model as a Baking Source like I mentioned. And using Bake “Selected To Active” you can then project that high-detail on the original model with much lower vert count.

My main question about this work by Farsthary and the GSoC is “What is the Time-to-Promise?” As a small fyi, some of us are in Production right now and we need our own “current solutions” while these features have not yet been implemented in a current Blender release. Like someone asked above: “Is this for 2.60?”


#11

I’m mentoring the sculpt SoC right now.

Farsthary has mentioned he plans to commit soon, we shall have a look at integrating it with our work once he does a commit. As to whether it will be release with this next release - I find the odds to be medium to low of me being able to convince the rest of the team but we shall see :slight_smile:

Right now we are fixing the brush implementation in Blender giving them more natural and predictable results.


#12

Your efforts are greatly appreciated! :slight_smile:

I actually like it that Blender (unlike other 3D Apps) actually has its own sculpt mode. My main concern only was that the “user experience” compared to Sculptris was quite different. At the end of the day, though, the workflow was kind of the same, since our goal was to project the detail onto a model for animation purposes.

In an ideal world… you’d hit “Sculpt Mode” and Blender would become Mudbox… but that obviously cannot happen overnight.

Again, your efforts are highly lauded… And my question was really more focused on “current workflows” for projects already in the pipeline.


#13

well it is free, but is it actually open? I didn’t see anything about it being open-source.

yeah, the attraction of Sculptris isn’t so much what it does or its capability, but how it does it, and just how intuitive, well designed and well implemented it is. It’s great that Blender will have similar features, but sometimes merely adding features to the list misses the point. Which is not intended to detract from the efforts of the Blender team, and I look forward to seeing what they come up with.


#14

Fastharys patch is integrated into our branch, but the general consensus is that it needs a lot of work, and probably needs to be reimplemented on top of bmesh instead of mesh. Right now it basically does face selection while you stroke, then it enters editmode, tesselates the selected faces, then exits editmode.

It has a long way to go before it can be considered a competitor to sculptris.

On the other hand, the core sculpting tools have improved dramatically in just the week+ time that SoC has been going on - one artist felt it was so much better that the student could stop now and he would be happy with the progresss for the summer.


#15

Well I remember seeing somewhere that Sculptris was supposed to be released under an Open Licence… but… it seems it probably was not.

My mistake.

And really, although Sculptris looks exciting… I must say I’d prefer Blender’s sculpting simply evolved into its own Perfected Version.

We basically have this workflow already:

  1. Modeling - Blender
  2. Sculpting - Sculptris
  3. Baking from Sculpt- Blender
  4. UV - Blender
  5. Rigging and Animation - Blender
  6. Visual Effects - Blender
  7. Post-Production and Editing - Blender

It’s practically all Blender… And I’d really prefer that Item 2 was also Blender. Sculptris only won out (for now) because of speed and brushes and tesselation logic.

But everything changes. And once Blender’s Sculpt Mode is up to speed we will shift Workflow study to that.

To LetterRip: Carry on. :slight_smile:


#16

You can give feedback on the sculpt SoC project here, users are regularly posting test builds.

http://blenderartists.org/forum/showthread.php?t=186339&page=14

LetterRip


#17

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.