The big Blender Sculpt Mode thread (Part 1)

Guys, any news about face sets and sculpt colors support for Multires?

2 Likes

I guess not: https://developer.blender.org/D8104 but are face sets not already supported ( https://developer.blender.org/D7168 )? I could have mixed it up.

2 Likes

Breaking news from the Blender developers front! :+1:

39 Likes

That’s great… but man, the sculpting module needs more devs, that’s a lot of work for just one dev…

20 Likes

Agreed, but at least Joseph can work full-time on Sculpt Mode starting in January. :+1:

18 Likes

finally damn it x"D…

8 Likes

Now that’s some good news. And yes, it would be optimal to have more people in the sculpt area, but let’s rejoice we at have Joe back on board :slight_smile:

9 Likes

Sounds like a Christmas presents :wink: fuck yea )) I love this news:)
image

11 Likes

Great news!

edit: Also strongly agree that this is an opportunity for Blender sculpting should the powers that be decide to give it more resources. Joe Eagar full time is a big step in the right direction, but like TheRedWaxPolice said, this area needs more than just one full time dev.

9 Likes

Thanks for the links.

Nah, just tested the latest sculptdev branch. No face set support for multires.

I am curious why these haven’t been merged to master. Is there a reason why these codes are just left this way? Its been over a year.

Dyntopo seems to be getting a lot of focus from my tinkering with the new dev build.

2 Likes

Face Set support for multires is the same as in master in sculpt-dev branch.
That means that face set creation and storage is limited to base mesh level.
But Face Set auto-masking is working at any level of multires modifier.

The big patch about SVC was containing work about undo, propagation, etc… Sergey asked for a split for an easier review and Pablo explained that will be complicated.
Since then, Joe proposed to solved problems encountered by multires modifier and Dyntopo with bmesh code.
This summer, he started to improve multires modifier. Since autumn, he is working on dyntopo.
And Pablo let Joe take the lead.

Recently, Joe published a roadmap for sculpt mode.
https://developer.blender.org/T93226

Priority will be SVC. Then, merge of sculpt-dev brushes. Then, Vector Displacement maps and Sculpt Layers. Then, Multires improvements.

13 Likes

Are there plans by the devs to work towards removing this limitation because I think the purpose of using multires is to have as low a basemesh level you can for optimization and good viewport performance? Maybe have the face set created at a particular multires level stored at that level? I think Zbrush doesn’t have this limitation if I am correct.

Good to know we will be getting sculpt vertex colors soon but with the very near term being:
‘‘Move sculpt colors out of experimental in master.’’
Does this mean sculpt colors will now work with multires or nah because the latest sculpt dev build still does not support painting with multires.

I appreciate the work Joe is doing and very happy to see the new roadmap for sculpt mode. This should have been the route a year ago. Multires has been in limbo for far too long and it is heavily used for very highend work.

2 Likes

Just found out Dyntopo now supports uvs in the latest sculpt dev build. Thats really awesome. Was this implemented by Joe?
Also noticed that if I change the sculpt brush stroke to anchored. It doesn’t tessellate the mesh as I drag out with dyntopo and crashes most of the time

Since dyntopo now supports uvs, would there be a bake displacement using dyntopo?

2 Likes

This is great!! This means that we will have two developers when Pablo cames back!! And it will be awesome!!! I have my perfect wish for this holidays :slight_smile:

5 Likes

No.I think issue with SVC is the same as for Face Sets.
I think that when Joe refers to Custom Attributes, PBVH Grids to PBVH Faces : it means changing storage at basemesh level to storage at faces of higher levels.
I think that is the purpose of all the mentioned Multires Improvements.

They want to introduce a new system that works with Vector Displacement maps.
Use of VDM should help to improve viewport performance and reach higher levels of details.
But this system has to be in place before modifying Multires to support it.
That is why Multires modifications are scheduled after VDMs.

Yes. Joe started with an UVsmooth brush for Dyntopo as an experimental feature.
But eventually, he improved a lot support of UVs for all brushes, one month ago.

Because there are such things too difficult to ask to your machine : Dyntopo is automatically disabled for them in master.
And it is because Sculpt-Dev branch is allowing user to do not judicious choice : that is an experimental branch.

Baking of stuff requires render engine developers to be involved.
But initial Pablo’s proposal was to put in place a way to bake SVC and custom attributes as material textures.
The VDM support implies that highest displacement will be a texture.
But it is not really clear, yet. I don’t know if we will use same brushes for sculpting and to create those VDMs.
Things are not really clear about that.
Maybe, we will have VD baking as a regular baking system option for Cycles, before sculpt VDM system will be ready.

7 Likes

You can check out a number of Joe’s additions in his YouTube channel:

https://www.youtube.com/user/joeedh

Sculpt-Dev issues can be reported (and are very useful and much appreciated) in the Sculpt-paint-texture-module at Blender.chat.

9 Likes

Thanks for the explanations. Understand more clearly now.
Joe is doing excellent work.

Thats great news. Finally, we will be able to use face sets in higher multires levels. This has been a long awaited feature :+1:
VDMs will definitely be a huge plus for Blender sculpting.

That is great. Thanks to Joe for this feature. I think Blender sculpt is going to be very phenomenal next year. Also congrats to him getting a fulltime job with Blender. He is definitely on the right track with what users want with Blender sculpt. Really impressed with all the info I am seeing so far. The Blender foundation are in for a lot of good and robust features that will solidify Blender sculpts and he has added so much within such a short time.

The new roadmap is excellent.

I am aware its an experimental branch :slightly_smiling_face:. Thats why I posted it so hopefully he would see it and it can be be looked into.

Noted. Hopefully we will hear more info about how this can be achieved. Based on all the info you have given, I already have a lot of trust and confidence in Joe’s ability to push Blender sculpt’s features in the right direction. Never would I have expected dyntopo to work with uvs and he added it within a short period. Next year is looking good.

I am really happy to know he is now full time working for Blender and thanks to him for the awesome things we are going to get for Blender sculpt. He is doing an excellent job!!

3 Likes

Yeah, checked it out before from the amplify video posted earlier on this thread. Excellent work!

Thanks for the link. Will look into joining. :slightly_smiling_face: :+1:

3 Likes

@zeauro Do you think when Multires is improved, the addon would work with it. From the dev, who is currently updating it, he mentioned unless ‘‘multi-reso mask data storage is officially supported’’

I am wondering if we can get vertex group support with multires.

1 Like

Currently, mask data storage is supporting multires levels.
The issue is that the mask is unique. That is why face sets are needed.

The roadmap talks about custom attribute support.
The purpose of a custom attribute is to use it to manage something not supported by regular data or to manage several kinds of data as once by conversion of custom attribute into a regular data channel.

Maybe that will have no interest for the addon.
But if others improvements are meaning that attributes and face sets will be supported by multires levels. The addon would probably become obsolete.

And maybe sculpt layers, that are planed before multires improvements will be sufficient to modify workflow and reduce a lot the need to store several masks.

1 Like