Sculpting layers
I used the sculpting features in Blender quite extensively lately and they are great! The brushes and options all work really well. The only thing that’s missing in my opinion is layers. Basically you could be able to add layers and sculpt “on” them. At any point, you can switch a layer on or off without ruining the sculpt details on other layers. Also, you can dial the detail on certain layers down or even increase them (layer “weight” could go above 100%). Where it’s very useful is in adding fine details on a mesh while making sure the larger detail you sculpted remains intact. Another great thing is that you can have an erase brush that selectively erases detail sculpted only on the selected layer, leaving the rest of the sculpt untouched. This system is really well developed in Mudbox but also present in most other sculpting applications. I was checking around and other 3d content suites have the feature even though they are not dedicated sculpting tools. It seems to be fairly popular.
Thumbnail/Proposal image by @KickboxingBanana .
Duplicates
- Sculpt layers
- Sculpt layers based of Photoshop/Painter
- Sculpting: Working on Layers
- Sculpt layers with opacity per layer
- Shape keys for multi-resolution modifier
- Shapekeys on multiRes Layers for sculpting and HD-Morphs
- sculpting and texturing layers
- LAYERS⚡in sculpt paint mode
- Shape keys for multiresolution?
- Sculpting layer system
- Layers in sculpting
- sculpting layers :>
27 Comments
People always make reference to zbrush but actually in mudbox layer system implemented much better. In real production with deadlines and feedbacks using blender as sculpting tool almost impossible, layers should exist in blender.
it could work something like this
@1loginlogin321: Nice mockup. Making it part of the Multires might work indeed.
Agreed that layers in zbrush are in zbrush are surprisingly underdeveloped. Also agree that the implementation in Mudbox is the best there is.
The idea sounds cool, but I think a lot more thought needs to be put into how exactly it would work. Would it be as simple as adding a second Multi-resolution modifier to the mesh and that modifier would serve as the layer? Or should layers be an additional feature built-in to the existing Multires modifier? Or should sculpt layers be a modifier of their own?
Of course, you could try to think of ways this might work outside of the modifier system, but we have it and (for the time being), it's stack-based. Perhaps that can be taken advantage of.
P.S. mock-up images are a big help!
A suggestion by itself has a little bit of use, but if you want the feature to actually be implemented, it's really best to put as much thought into the concept as possible. Consider how you would design the feature to work and how you would like to use it in Blender. That gives both users and developers the clearest understanding of your idea. Then we can work together to refine it... and perhaps even convince a developer to work on implementing it.
That's the purpose of this site. It's really not meant to be just a dumping ground of ideas. In the ideal case, we're tying to put together a useful resource.
I actually did a little mock-up but I cannot figure out how to post the image. :) Should I upload it somewhere and add it with a html tag?
There was an update to the site software last night. Editing is now done using Markdown. When you edit or create posts, there's a handy link on the side to a Markdown cheatsheet if you're not familiar with the syntax.
Sculpt layers are easily one of my top desires for Blender to implement (after fixing multires).
Off the top of my head I think this would be something that would go inside of the multires modifier instead of being its own modifier because in either scenario (whether it's part of multires or its own separate modifier) it would have to synchronize with multires because you might have sculpt layers where you were sculpting on subdivision level 1, and others on level 2, or 3, 4, etc. and it seems like it would be harder to do that if layers were external to multires and had to synchronize with it. Simpler to just build it into multires itself. But I'm no programmer so I don't know which of those is actually easier to do.
With that being said, the features I would expect are thus:
The beautiful thing about being able to work with sculpt layers is that it is a semi non-destructive workflow. You can turn layers on and off at will and experiment with different strokes, textures, weights, and so on, or you can selectively modify or erase details while preserving others, all without having to make destructive changes to your sculpt to try something different. This gives HUGE flexibility. It is impossible to overstate just how flexible this is. Especially if you can change the layer intensity; it lets you retroactively change the strength of your sculpting to make it stronger, or weaker. All of this is great for variations, iteration, and change feedback from supervisors.
Open-ended ideas/questions:
* @Chris Kohl * Very nice post. Yes, I also think it should be an improvement to the existing multires. I am not absolutely sure but I think something like this is part of the roadmap of improving sculpt (the stuff Pablo Dobarro has been doing). From this doc: docs.google.com/presentation/d/1fmjdtajPzeD3ixpGIvseKMRsAzhmKC4etIE8YqacwLk/edit
"Should support layer operations Provide multiple structures to store the deformation when it is copied back from the proxy. Then that data can be weighted and added to create the final result. "
If Pablo is involved I believe it's going to get done eventually! :)
Hey guys this would turn blender into god mode With the recent improvements to Sculpt, Blender has turned into a real alternative for sculpting, the one thing missing is layers for your sculpt I have attached an image to give some idea of how it could potentially work? For example i can use an alpha on one layer then use the slider to tone down the strength, while it not affecting the other layers alpha
![Image][1] ⋮ [1]: ibb.co/PZqxcXH
* @jaylukeherbert * Yeah, that looks good. Placing the layers in the Multires is a good idea! Honestly, I don't think that sculpt layers are such a mystery in terms of how they work from a user perspective. Implementing them seems to not be so straightforward ... or there is no developer who is interested in that. There seems to be enough interest from users...
A bit of an old conversation, but I don't see how the suggestions already given (with examples from other programs) wouldn't be good enough for devs to understand the concept.
Nevertheless, sculpt layers in Blender would probably work like shapekeys that not only work on the base mesh, and their effect would be on a single multires modifier.
I imagine the interface being really simple (no need to complicate it, the code by itself will be complicated enough, I wonder if this will require an entire rewrite. I hope not.), the order on the stack wouldn't affect anything and you'd be able to change the intensity that each layer affects the base layer).
The functionality is pretty similar, not to say identical to shapekeys, the problem with shapekeys is that they're slow as hell and don't support multiresolution.
I think this youtube.com/watch video well explain how sculpt layers must work...this feature in zbrush since 3.0 (i still have video with feature Preview of sculpt layers in It alleged zbrush 2.5 that never came i guess - which (video) dated from 8 aug 2006 cloud.mail.ru/public/NGBQ/aExGiqM2Z - i mean sculpted layers in zbrush almost 10 years - it would be nice to have it in blender too)
Videos of other applications aren't particularly useful... especially since (in this case) Zbrush and Blender have very different workflows from one another. It would be best to describe and show how the feature would work within Blender and how it would fit within Blender's existing interface.
What's the difference here - the current interface? Can not we change it - to improve? And here is a different workflow if a new function for the packet - so it is in any case, change the workflow.
You have strange ideas - if a new feature is meant to be inserted into an existing workflow? And if you need to change processes for new feature - something that now - not to add a feature - to not change workflow?
How then - to add new features?
p.s. those who put the minus - write - why. If you do not write - I think you kind of fanboys.
Because "make it work like X other program" is not an actionable choice. Blender has to maintain its own internal consistency. Sculpting layers, in the abstract sense, are a great feature. The question is what's the best design for implementing them in Blender's environment.
If you're going to add a tree to a landscape, you don't modify the landscape to match the other places you've seen that tree growing. You find the places in that landscape that mix the best with that tree. Then you plant it so it looks like it actually belongs there.
to change landscape - need apply huge amount of resources and in some cases impossible at all. To change the way the software work - need aspiration, programmer and idea of functionality - it's simple if not clinging to old concepts. It's a thousand pities that in blender community many users want to blender work as 10 years ago - supposing developers always listen these users - then we still would sit on somthing like 2.48 and listen buzz of those peoples. If changes need to happen - then they must happen quickly. if you can not do it this way as it was done before - then it should be done other way but - done.
>New features need to be able to mesh well with existing ones. really? - that existing in the blender functions flow smoothly one into another? No! If the sculpt layers not be connected with other functions but will be as beautiful as in the zbrush - I think 95% of those who make sculptures in blender will agree with me - they must exist independently of everything.
Any updates on this?
Would be a very welcome feature!
Hi! Other things i have to say, what about Polygroups like in Zbrush? can we expect something like this in Blender?
Hey pix! If you're interested in that feature try submitting a new post to see what others think. It's not that I mind you posting it here but I guess it's nice to keep separate ideas apart. As far as I'm concerned I never really felt I'm missing Polygroups. As for sculpting layers I keep thinking "I want me some layers!!!" :) That's why I posted this here.
How about something like this: rightclickselect.com/p/sculpting/Hdbbbc/shape-keys-for-multi-resolution-modifier It would (partially) solve the problem of layers.
I think that would achieve a similar result and might have some advantages. I really don't mind any implementation, I just care about the functionality.
cool idea but what's about Dynatopo? Do you think, the layers should get subdivided individually, or would the layers, which aren't activated, just be subdivided smooth.
I am not sure the idea is technically achievable with dyntopo. I guess the way it works is that each layer is represented by a displacement map which are then layered on top of each other (I could be completely wrong here, by the way). Either way I'm not qualified to answer technical questions; I'm just an artist who wants a feature :)
@Niklas Reiprich: That's probably not doable in any way that would make it have a good performance. Dyntopo is mostly for rough sculpts (to then retopo and do multires_ and I think by having layers on multiresolution alone we'd already have a pretty good deal.