https://media.discordapp.net/attachments/587761956526751784/944726091430498324/hmmmm.mp4
Feels quite short… 
8 Likes
It would probably help the developers if people did not start rumors of [insert project here] being abandoned whenever the last update becomes more than a week old, or even more than a few days old in some cases.
Unless one of the long-term effects of a Covid infection is a sudden onset of ADHD, then this type of posting demonstrates a lack of understanding of just what all goes on terms of patch review and user testing (since for starters, a studio artist gave Sculpt Vertex Colors a test run and found bugs and other usability issues that needed to be looked at).
That is fine if you have a deadline to meet, but I do not believe the BF should allow Hollywood to impose the way it makes content on all users (as not everyone here works in a studio or is in training to work on future blockbusters). Just because the likes of ILM and Weta has a specific workflow does not mean the Blender devs. should enforce the idea of it being the only valid one on the code level.
4 Likes
What? No…I guess you didn’t watch the video…
Watch it, then you’ll understand what I mean…
1 Like
I see now, but it does not necessarily mean the contract can’t be renewed once the three months are up (which is possible with the resources the BF has now). It also does not mean that the BF plans to sweep sculpt development under the rug and forget about it in three months either, especially when noting that they have been hitting a lot of high notes recently (such as getting editmode and subsurf to a performance level above 2.79).
1 Like
Yeah yeah sure… but c’mon, I mean, c’mon…
8 Likes
We are talking about an approach that works and helps achieve good results 
. From the vid I posted above, it demonstrates why Multires/subdivds work better because they have some sort of viewport optimization when working with very high poly mesh which is advantageous compared to dynamesh/dynotopo.
Thats why I keep mentioning why Blender devs should give Multires some more attention. Dynotopo seems to get more focus which is fine but Multires is very important especially for very high end work like the project they are currently working on which is ‘‘Hollywood’’ sorta related in workflow.
Noone said its the only valid one. Multires is taking a backseat to dynotopo and if this is the specific workflow by big companies like ILM and WETA, why not? More dev grants for Blender potential, if you ask me tbh.
Joe is doing very excellent phenomenal work. Hoping it gets extended. 
4 Likes
Multires requires the implication of an experienced developer.
Often, Pablo’s addition had to be reviewed and completed by Campbell and Sergey.
And those developers are responsible of other modules.
Pablo also introduced a workflow based on a OpenVDB voxel remesher addition for intuitive creation of basemesh.
At that moment, developers asked themselves if it was necessary to continue Dyntopo support or to abandon it.
To that question, Dyntopo users replied : NNnnOOOOoooo !
Contrary to OpenVDB voxel remesher, you don’t have to frequently remesh to progress with dyntopo.
With a detail size corresponding to brush purpose, you can experiment a pleasant feeling of direct satisfaction about reshaping intuitively basemesh with dyntopo without being interrupted.
What stops that feeling is the performance decreasing with polycount increasing when you start to add details to shape. So, we asked for an attempt to improve Dyntopo performances.
At that moment, Joe replied that Dyntopo developers used Bmesh, but it was not a good choice because Bmesh was not created for that purpose.
As Bmesh creator, he proposed to develop a better Bmesh for Dyntopo.
But he started to improve Bmesh for Multires modifier into temp_bmesh_multires branch until September. In parallel, he improved Dyntopo until start of February.
He also had to modify Sculpt Vertex Colors to make them compatible with new attributes that would be useable by Geometry Nodes.
To prepare merge into master of Sculpt Vertex Colors, he created a temp_sculpt_colors branch and focused on that branch, during whole February.
It is not necessary to ask developers to work on multires modifier. They planned to do it.
But until then, Joe is following the roadmap he settled, which consist on assuring working SVC and dyntopo improvements, first.
Then, merging Pablo’s brushes and working on brushes management.
And eventually working on Multires, sculpt layers and VDMs.
Dyntopo improvements showed into the article will be soon merged into master.
At the same time, they will probably be the last improvements made to Dyntopo, before a long period of development that will focus on other stuff.
11 Likes
New auto-retopo algorithm. A possible new candidate for an updated auto-remesher, @joeedh ?
18 Likes
looks dope, will be perfect if it can be adjusted to take into account face groups .
7 Likes
I can enjoy looking at such neat auto-retopo results like viewing a Leonardo da Vinci painting. 
5 Likes
Any hope we’ll someday get an algorithm that can cleanly put a sharp bevel on something shaped like this?
2 Likes
THISS. Great point, @CDMJ, The brush tool bar is full of great brushes but they’ve piled up and now it’s unmanageable - so I just use my most used brushes from a custom pie menu.
I would rather have an abbreviated, custom set of brushes in the toolbar. Ideally, we would be able to filter which brushes appear on the left of the screen instead of that long, long list.
Also…“meat space” LOL
3 Likes
some texturing news Texturing meeting
5 Likes
That all sounds great! One small comment - I don’t see that many 16k textures used in film. I do see UDIMs everywhere so hopefully that side (baking to udim, painting across tiles) is also under consideration.
Impressive topology even able to generate correct low poly and able to keep edges correct.
This could become the new big retopo tool after InstantMesh.
Will have to find some builds and try it 
2 Likes
Hmmmm… 
8 Likes
That is like using Quadriflow 
4 Likes
This looks like we’ll need to keep using Quad Remesher once again. But the examples in the paper look very promising.
4 Likes
Hello all,
Is Sculpt Vertex Colors actually working in 3.1.0 ?
I am using the unofficial win7 build, and the related UI elements do show up as expected after running
C.preferences.experimental.use_sculpt_vertex_colors = True
However the Value filter seems to be the only thing that actually affects sculpt colors - I can’t seem to get anything else to anything. Painting a color doesn’t do anything, and the Store/Load buttons don’t seem to quite do what they should. I can make existing Vcols show up in sculpt mode after some random clicking of the buttons, but that’s about it.
Note that I am familiar with the (temporary) workflow as I have been using the special 2.80 Sculpt/Pablo build for a while now and everything behaves as expected there (again besides the very odd labeling of the store and load buttons which makes no sense whatsoever, as they doen’t say what they are loading/storing from and to But that’s probably hardly relevant since these are supposed to disappear soon anyways).
Now of course this could very well just be a case of the win7 build not supporting it to begin with. Would love to see someone confirm that it is working for them.
