How does Blender not have sculpting layers?

I’ve finally taken the time to sink into Blender’s sculpting tools, and have to say I’m quite impressed. The amount of precision and control you can have is something I never felt with Zbrush. But one thing I don’t understand is while sculpting, at least with the multi-resolution modifier, the workflow seems very destructive? There are no layers to adjust/turn off, and shape keys don’t seem to work in that manner (at least when I tried).

If any of this is wrong, feel free to correct me. But this seems to be a not so insignificant feature that is missing from an otherwise great tool set.

The work required to code it is probably not so insignificant either :wink: We still have a multires feature that´s pretty much unusable, can´t imagine coding a layer system being any easier to do well.

Answer is probably that we have no dedicated sculpt coder, Cambpell and others have improved sculpt and fix bugs, but they also have other coding duties (and bugs, bugs, bugs)

I see, thanks for the response.

I think it’s a bit extreme to call it unusable beside the very rare random spike bug, I 've n’t had this happen to my in years now(knocks on wood), It works pretty well and will definitely allow you to get higher polycounts then dynamic topology which is vital for detailing.

I have an extremely modest graphics card but I can easily sculpt on meshes with 6 million polys no sweat.

I personally love going dynamic topology to multires

Multires is pretty broken. I’ve never had a single multires sculpting experience in Blender in nearly a decade that didn’t eventually lead to some kind of broken showstopper of a problem. The devs are well aware that the existing system is beyond subpar, but finding the time and a developer to fix things are two high hurdles to jump over.

Well we have masks, and there is a great add on free “sculpt tools”. (which does do some remeshing as well)
Personally if i’m about to do things that might destroy my sculpt or im not sure about it, i simply copy the object first.
From what i heard Blender is richer in tool set, plain mesh editing, and sculpt tools, and can handle more faces.
But i’m no expert to compare, since i only use Blender.
Though i think you better watch some youtubes about what you can o with sculpting in blender, to learn the tricks.
Which you wont get into if your playing and not sure what setting to change to get certain brush effects.

Because no one had coded it yet, I believe if somebody codes it, we will find a way to use.

Oh man, I really don’t like when people make broad statements about some feature without explaining what the real issue is. Can someone be more specific other than saying “It’s broken” or “Unusable”. What exactly is broken with it?

It’s only fine if you’re trying to model a hedgehog

It’s perhaps a bit convoluted and requires a retopo solution (and knowledge that it’s even possible), but dyntopo > retopo > multires, with applied shrinkwrap > final detail pass never seems to break on me. You get all the benefits of dynamic topology and the resultant efficiency and quality of a quadified mesh in multires, for the cost of a few steps in the middle. And no more wrestling with hedgehogs… usually…

Spikes have been an issue since the 2.5 upgrade or so. There was a bad design decision, and now it’s haunting Blender all these years.

Blender needs a new multires system that separates the displacement density and the tesselation by allowing users to “paint” vector displacement textures (UV or Ptex) directly with the sculpting tools and using interactive viewport tesselation to draw them.

Guess you couldn’t call that multires though since there would be no discrete subdivision levels anymore, just the displacement texture resolution - which could be per-patch with Ptex, or dependent on the UV-layout with UVs.

There is no need to throw out the baby with the bath water. The spiking bug is persistent and annoying and hard for the dev to sort out because it pretty difficult to reproduce.

But lets pretend it does n’t exist(the spike bug) then what is there is very usable.

Personally I think a bug fix has a higher chance of happening then a complete rewrite…That is one area of Blender that hasn’t had a regular developer in a really long time.

not creating pefect quads but inside the sculpt tools there is some kind of remesher, realy nice tool since it also allows for bolean add operations.

Ahhhh. Now I see. Thanks for informing us. :slight_smile:

Yes, this is the method I’ve used as well. I’ve only used it a few times though so it’s not like I have a year of no problems.

This sounds very similar to what Modo used to do before they had the multires system. It’s funny because back then people were saying that the old system was “Fake sculpting” because it was only simulating the sculpting by painting instead.

As far as I know, the spikes bug would mainly be an issue if you’re trying to do something extreme (like turning the default cube into a highly detailed face).

The last sculpt I did made use of a base mesh with a fairly decent amount of polygons (and a smaller multires level as a result), the only times I had spikes was because of bad geometry made during the retopo process (after the initial Dyntopo sculpt).

In short, create a preference for a higher polygon base mesh coupled with a lower multires level (with the added bonus that it will be easier to make textures and shading masks for).

To op sorry to go off topic.

To Richard and others i agreed with you guys saying that Multires is broken but being unusable it is a little extreme, i will say people are using it “wrong” (not the Blender way).
Mostly spikes issue more often happen when people sculpt on different subdivision level and go back and forth, high and low level. I have this principal when sculpting when i go high i don’t get low. I have just to be sure that i have enough details on given level, so that when i subdivided i don’t have to back on that level to touch it again. As many has pointed out dyntopo -> retopo-> multires usual solve that problem.
Another problem is the size, you don’t want to sculpt with real life size you will end up with some problems.
Those are workarounds that i found when sculpting with multires. I used multires in all my finished work and i didn’t have spikes or any another problem for long long long time.

interesting thread, I also have the same question

That is exactly sole purpose of having multiresolution sculpting and basic workflow in every software that support it.

People seem to manage just fine with Dyntopo and its complete lack of resolution controls.

Multires still has a use if you leave it for the final detail pass (ie. for scales, folds, and whatnot), not being able to use lower levels is less of an issue now due to 2.77 having better performance compared to years back. I do agree though that it would be nice to just not have to worry about spikes, but it boils down to the usual issue of finding a developer willing to work on it and the BF not having enough funds to cover every area at once (and there’s dozens of areas that could still use improvement).

Who use dyntopo in real world? a tool that don’t allow to make polish surfaces? really?

Dyntopo, Multires and sculpt in real world are unusable. The devs must accept this, you cannot use a tool that don’t work the 90% of situations. I have used sculpt in a few works and always happens the same, it’s not extreme cases. It’s everytime.

Nobody sculpt in dyntopo, dyntopo it’s only a way to make sketchs, concepts,… But when you work you need multiresolution. First because you work in different levels to change the shape without delete the details. It really simple see who everybody works with Zbrush, the vast mayority of people work with a few tools

  • Subdivisions
  • Zremesher
  • Layers
  • less than ten brushes
  • clip brushes

You see this list, what do we have implemented in Blender?

  • Subdivisions (multires)
  • Brushes

And one of this tools don’t work correctly. Ok, we can forgot clip brushes because we can model in realtime with classic tools, and well, we can do the retopo in a traditional way. But subdivisions are mandatory, and layers will help a lot to make complex models.

It’s sad that the tool is really good, and have a lot of features and we can use with the modeling tools, deleting polygons, change base shape,… but you cannot work at 100% with this, at the end it’s only a secondary feature that you can use sometimes in a few cases. And in the real world when you don’t trust in a tool… you use othe tool.