Devil’s advocate…cough cough…3ds Max dev team Reform. 
3 Likes
100%. This is what I’ve said a few times. A good clay engine is more important than anything else for sculpting. Sadly, Blender clay just doesn’t compare to ZB, imo. This obviously includes polish/trim brushes, which are part of the clay engine.
13 Likes
Their past accomplishment might be the best of its kind in human history, but that doesn’t mean I can throw them into cyles or the brush engine or manta flow or obj export code and expect them to immediately start producing production ready code. Especially if I assign them with tasks in all of those different areas simultaneously.
1 Like
No, quite the opposite. Blender’s open movies have been generally used to develop specific features during productions. This has been quite hit and miss as to whether the features ever ended up in master or just stayed in a separate branch used for the production.
6 Likes
Yeah, that’s the only (relatively recent) auto-retopo algorithm that has managed to equal ZRemesher / Quad Remesher, or so I’ve been told. 3D-Coat’s Autopo is a very competent runner-up.
2 Likes
Glad to to see that they took my lengthy feedback post seriously way back when they made a task on developer.blender.org some years back. The node system discussed a while back I also approve of if it means that we finally can have brushes as proper assets that can be easily shared and installed for anyone who has the files.
Just hope we get to see some serious development being put into Sculpt Mode in the near future. 
3 Likes
A post was split to a new topic: Image texture is pixelated despite subdivision
I’m also hoping that the Brush Management work considers Grease Pencil too, otherwise the UI will become a giant mess. I for one, would like to know why particular strategic choices were made. To me it feels like Blender development, apart from performance improvements here and there, has slowed down noticibly in the last 6 months or so. At least, I unfortunately don’t see as many exciting things as there used to be in previous years (could be my workflow tho). I understand that devs need to fix things under the hood here and there, but I would also like to see some, albeit small, QoL improvements across the board.
Hopefully brush management will become straightforward and intuitive from an artist perspective rather than conceptually framed from a development POV.
I’m justnot quite seeing how you could easily change alphas if a brush is an asset. And the whole library override system sounds way too confusing terminology wise.
If you have various brushes that only differ in terms of alpha related settings (e.g. spacing), doesn’t the Asset Browser method add lots of overhead? And wouldn’t Alpha, Stencil and VDM settings benefit from being a brush tip module that acts as a sort of submode for a brush?
4 Likes
In fairness to the devs, this is what users actually asked for. Less new features and more performance and polish improvements. Two performance changes that I think affected Blender sculpting very positively this year are .obj-file improvements for importing and exporting and Eevee being fast enough to do live PBR sculpting. The former makes working between multiple programs a lot easier and saves literal hours on a per project basis. The latter changes workflow by allowing multiple lighting setups, realistic materials that match your sculpt object, and makes sculpt painting actually feasible.
However, I agree that things have slowed down pretty substantially for Sculpt Mode. I am still very much looking forward to DynTopo finally getting a proper overhaul in the Master branch. I would really like to know what’s taking so long considering that it was being hyped up to be nearly finished as early as last year.
6 Likes
Definitely, those performance improvements are very exiting indeed. And very much welcome and requested. What I meant to say, though, is that there are many promising and long term project on the horizon, but I’m not seeing any of those being worked on actively. It appears to be that Blender is in a transition phase of finishing previous work that seemed to have taken longer than anticipated (I may be wrong on this) and that work on either short term QoL or long term major features are both not in a state where it is reasonable to expect either to land soon. Hope I don’t sound to confusing lol.
5 Likes
Just for the record, I’m not blaming anyone here. I am, however, questioning some decisions. I would just like to see some promising proposals that were made by the community for the UI or for reducing clicks be adopted by the developers in new proposals. I’m not seeing enough streamlining workflows. I can’t help but think that work on the grooming (hair curves) started too soon and that instead the brush management + UI should have been tackled first, so as to not be constrained by the new features/ tools that will then get in the way of a major UI overhaul. I feel like in this case, as with Sculpt and Grease Pencil, that if the basic tool/ brush management would have been thought out on beforehand, the tools could have been implemented in a more consistent workflow manner, thus requiring less clicks. Now some of it feels very… let’s say… left untouched after delivering a feature (e.g. we have been asking for crease + pinch brushes combined into one, that’s a tangible target QoL improvement).
6 Likes
I dont think any major UI development will happen until they hire another full time UI/UX expert.
1 Like
Speaking of priorities: the position of UI developer does not even appear on the jobs screen immediately, but only if you click on any of the 4 positions they are hiring for and then scroll to the bottom for other positions…
5 Likes
An optimization patch from Hans from GeoNodes team. (Recently he’s been doing a lot of cleanup and generalization regarding attributes throughout many parts of Blender.)
https://developer.blender.org/D15937
Sculpt: Make storing mask and face set layers optional
When these features aren’t used, there is no sense in storing the
corresponding data layers and using their values for computations.
Avoiding that should increase performance in many operations that
would otherwise have to read, write, or propagate these values.
It also means decreased memory usage-- not just for sculpt mode
but for any mesh that was in sculpt mode. Previously the mask, face set,
and hide status layers were always allocated by sculpt mode.Here are a few basic tests when masking and face sets are not used:
Test Before After Subsurf Modifier 148 ms 126 ms Sculpt Overlay Extraction 24 ms every redraw 0 ms Memory usage 252 MB 236 MB I wouldn’t expect any difference when they are used though.
The code changes are mostly just making sculpt features safe for when
the layers aren’t stored, and some changes to the conversion to and
from the hide layers. Use of the “.hide_poly” attribute replaces testing
whether face sets are negative in many places.
You can help him by downloading the test build and see if there are any regressions:
I personally don’t see any.
9 Likes
EEVEE next and Viewport Compositor are long term projects that have been worked actively, during this past year.
That is obvious if you test master build.
IMO, with simulation nodes, they are last missing bricks of 2.8 design.
All releases of the 3.x series will be full of QoL improvements.
If you take a look to 3.3 release notes, except new Hair object, at each section, there are few new features giving new abilities and most of what have been done are QoL improvements.
For geometry nodes, nodes to UV Unwrap and Edge Path nodes are releasing new abilities. But what will be done with Volume Cube Primitive, Points primitive, Mesh to Volume nodes ans Intersecting Edges socket could have been done in 3.2, in a more fastidious way.
For Grease Pencil, except Light & Shadow contours of Line Art modifier, the rest of what have been done are only QoL improvements for its animation.
The whole section about Library Overrides is about improvements of UI for a better UX.
Image Plane Marker is a simplification of workflow.
And the last section with a detailing per module is mainly about QoL improvements.
There are QoL improvements everywhere in Blender, at each release. But when your workflow is focusing on modes to paint and sculpt ; that will be difficult to feel any improvement, until brush management + UI will not have been touched.
Sculpt and Texture module is behind. Because it was behind in 2.8 design.
William choices for its UI were not the good ones.
Pablo had to deal with them. He expended mode behind what was anticipated. But he did not have a clear idea of what would be an appropriate UI for his work.
Now, Julien has a new solid idea for a new paint mode that could share a lot of its features, with sculpt mode. And the work for a design of a correct brush management has just started.
Marking a brush as an asset just means, that a thumbnail of it can be displayed in a catalog of assets, in Asset Browser or a minimized asset browser in a toolbar. That is just an internal way to simplify brushes libraries management. That has no effect on how a brush works, when it is selected.
You can still modify its settings. Asset browser allows you to append datablocks. It does not force you to link them and create library overrides. The role of overrides is to let to user the ability to change locally settings of a linked datablock.
So, whatever way you activate your asset, it is EASY to modify its settings.
There is no reason to fear lots of overheard because the word “asset” is used.
Here, it is used to simplify brushes libraries management, access to brushes (not brush behaviour).
Anything can be an asset. Brushes but alphas , too.
The same way as there is a thumbnail selector for brushes, there is one for textures/alphas.
If an asset browser widget is done for brushes ; there is a strong probability, that one for alphas will follow.
And, if you don’t want to create custom brushes libraries, you will still be able to ignore that option.
9 Likes
Oh yeah indeed lots of QoL, just checked a 3.3 vid and hadn’t kept up all the way. Still there is plenty of QoL that has yet to be initiated.
As for the Asset Browser + brush management, if entire brushes incl. tip settings are supposed to be assets (meaning the entire brush and all of its settings) there will be lots if duplicate data. So if instead, it were to be more granularly modular, then it would be easier to swap brush tips. I’m hoping it will ne more of a bottom-up than top-down design so you can mix and match more easily without this manual workflow of: duplicate brush, edit settings save brush. Instead I hope it will be like: pick brush, pick tip (if not the default) and go. You can always save brush + tip together as an asset anyways.
OK. I get your point.
Brush Settings are a massive monolithic datablock.
Textures and color palettes are datablocks that can be used by it.
But we don’t have same ease to reproduce a texture brush mapping, a stroke set-up, a sculpt plane set-up or an auto-masking, through several brushes.
Now, I understand more why you can be afraid, if we pass from 2,3 asset browser widgets to 6.
There is almost zero chance that set-ups I mentioned would be represented by a thumbnail.
But it could be possible to create sub-presets to handle that.
There are lots of brushes modes in Blender. A unique Paint mode for meshes could become reality.
It took months to make a paint brush able to share a lot of settings with a sculpt brush.
But there is almost nothing in common between a Grease Pencil Brush and a Mesh brush.
One is creating dots or strokes floating in 3D space, most of time, projected on a plane.
The other one is displacing or painting vertices of a surface, most of time, curved in the 3 dimensions.
The way to map textures is different.
The way to define stroke variation is different.
The way to mask is different.
I doubt that will be possible to share, for example, a tip shape, between a GP brush and a Paint brush.
Even though, GP brush were refactored to mainly draw strokes on 3D surfaces.
The nature of a GP stroke as an envelope around a succession of GP points would still result, in a different mapping of textures.
To have a possible fusion between Mesh and GP brushes natures, GP points should be simple voxels.
That would imply that a 3D scene, showing meshes textured by millions of pixels, should be able to handle billions of voxels.
The whole point of keeping spacing between GP points, is to avoid a too heavy 3D scene.
I just chip in here in this thread, which is a conversation between five people.
Last time i sculpted with 25 mio it lagged a bit with a texture on the brush but with 3.4 its super smooth. Changing modes is not that fast but otherwise smooth…
Thats nice, exited for other improvements.
1 Like
at 2.5 million verts on a sphere the clay brush is super laggy while the clay strips brush is super smooth
Turn off Auto-Smooth…
3 Likes