Forums

ah ok, look as some parts use UVsets and another use multiple textures. sorry, this workflow is out of my knowledge. - related: https://manual.reallusion.com/Character-Creator-4/Content/ENU/4.0/05_Character/UV-Optimizations.htm?Highlight=uv Head and neck (blue areas in the illustration) are merged in one UV for enhanced scan head and animated blend-map.

There are no plans to reinstate support for Angel Script; it was removed from the project entirely due to incompatibility with new Apple processors. I think 2000 isn't that much, and I can convert it to Python in a few days. Please send me the Angel Script source code and any other files needed for it to work.

using the treat materials as separate textures option during the baking process... i hope i didnt confuse when i said it gives 2 channels. i meant that it gives 2 texture sets, naming them channel 0 and channel 1, but they are just scramblings of the diff uv sets. even worse the file being used has the various texture sets set at varying resolutions.

and the mesh that gets baked out is another hot mess. painting direct on the model is impossible. this is using the import tiles as uv sets option when generating the mesh for the paint room. it gives 2 channels as materials...

so i checked my export options, and i believe i do have the udim option enabled?!? for texture work it didnt seem to matter, but when you try to bake out the mesh from the retopo room it causes problems w/ the 1001 uv set, combining the head, eyes and teeth into 1 scrambled mess of a uv map.

we have 2 options on import (as far as i know of). 1. treat materials as separate textures: this is the better option but yields the results as shown in original post. 2. import tiles as uv sets: this option completely ruins the texture settings and the uv sets, mashing several sets onto 1 and renaming the sets in 1001, 1002, 1003 etc fashion. i will try to see if there is some option on export to help w/ this. as always, many thanks.  

https://pilgway.com/files/3dcoat/PythonAPI/classcoat_1_1_scene.html all of these layer getters create new layers  it seems theres no current way to query the layer state, or count, using either the ui.cmd or python $LayersPanel::AddNewLayer $LayersPanel::AddLFolder $LayersPanel::AddLayerMask $LayersPanel::MergeDown $LayersPanel::DuplicateLayer $LayersPanel::MoveLayerUp $LayersPanel::MoveLayerDown $LayersPanel::DeleteLayer -- delete contents of layer but not the layer itself $LayersPanel::TrashLayer -- deletes the layer $MERGE_VISIBLE $MERGE_DOWN $MERGE_UP $FILLLAYER1 -- fills everything that is currently visible and not ghosted with the brush the above seems to be everything we have access to in ui.cmd, and all except one python method spawns new layers if a query is not found. There is no "navigate up" or "navigate down" to switch which is the active layer to query the current layer. I'm trying to do a smart consolidate that basically ensures there is only ever a layer 0 (sculpt) and a layer 1 (paint) in the layer stack. I would name the bottom two layers "Paint" and "Sculpt" but 3Dcoat implicitly spawns layers named specifically "Layer 0", "Layer1" or "Layer 1" in so many tools that I'm just going to give up on that and try to consolidate to layer 0 / layer 1. But in python there is no way for me to check - how many layers there are, or get an array of layers - navigate up or down layers without spawning new ones - merge specific layers, or perform any operation really unless you know their layer IDs / names ahead of time - You can't actually get layer names by iterating ids without spawning new layers - reorder layers intelligently because you can't get their index in the stack More api request! - For 3DC extensions there is no way to tear down the extension, or unregister it from 3dcoats runtime once it has been started. Would request a hook for being able to unload the cModule without having to restart 3dcoat - edit edit I was able to work around it though. Been banging claude over the head with a wrench all weekend.  Here is a short video of my layer consolidator. It merges all the color data into layer 1, and all the sculpt data into layer 0. Why is this useful? Often you will accidentally paint color or depth on a color you didn't mean to when sculpting, and this ensures that there is one layer with only color, and one with only depth API weirdness: You have to have a paint brush open to use the $MERGE_UP command. Doesn't work otherwise via code.   2026-01-04 22-29-16.mp4 I'm finding, when I use iterate subtree, we are traversing through instance links and iterating the same objects multiple times. This is not desirable: if an object is instanced it doesn't need to be processed more than once - Identified the bug: Collect selected returns multiple copies of instances. If an object is instanced, but only one is selected, collect selected returns the selected sculpt element, and its instance, sometimes more than once. Next, one click autopo to multires. It's a bit finnicky but it works. Keeps you in Sculpt Mode. Sure wish I could suppress that "getting to sculpt" popup that keeps coming up. Tried using the canned confirm ui cmd but it just resamples my mesh instead of closing the dialog.  2026-01-04 23-58-13.mp4   Lastly, an overview of all the other tools in the extension. 3DC's api cmds is a struggle but the more python it gets the better A lot of the issues you are seeing in this video is due to 3dcoat's builtin methods for collecting selected objects returning instances as well even if you only have one side selected. It's why you see the leg getting processed twice, and actions selecting the other side. I've tried to code a workaround for this but its not 100%. Ideally 3dcoat api just gets fixed so selections don't return instances as well unless they are actually selected I made a safe resymm that works in surface mode because the majority of the time, applying symmetry in surface sculpt mode crashes 3dcoat instantly. It converts your object to voxels at 4x, symmetrizes, and then decimates back down the original polycount. The point of the dynamic subdiv is to make the brush setting for dynamic subdiv global across brushes. IMO dynamic subdiv settings don't make sense to be brush bound, so this is a workaround so I can switch brushes without having to setup my subdiv levels again each time.  I've mapped increment / decrement dynamic subdiv level to pg up and pg down. The setting is just stored as json so future invokes of the tool remember the last value.    2026-01-04 23-48-23.mp4   If this seems useful to you, I've  attached a zip of the cmodule. You can extract the LKS folder to  C:\Program Files\3DCoat-2025\UserPrefs\StdScripts\cModules\ Or at least, that worked for me. There are some button in the Extension tab that are for inserting a bunch of action scripts into the scripts menu so they can be bound to shortcuts. I am thinking I might make an embedded shortcut editor for the builtin scripts assuming I have access to shortcut assignment     LKS_3DCoat_cModule_20260105.zip

Hey guys I need your help. When I try to put AO and the model is on the flat view (or num 2) it doesn't show. Please can you help me with this

As I remember, you need to import the model UVs as uDims  

greetings all. admittedly ive been away from 3d for a while, but i dont recall having this issue w/ character creator figures in the past. at first i didnt even know what was happening, it was causing all kinds of issues moving between the paint>retopo>sculpt rooms and while i dont have the technical know how to explain properly, seems like uv maps are getting combined or moved to other positions in the uv grid?!? can someone please help explain whats going on and how i can work around it? on import i am using the "treat materials as separate textures" feature which provides all the diff texture maps. if not,  when in the retopo room when i bake out the model i get two channels, and i dont get the 001-006 uv sets which correspond to "some" of the texture maps.

that would save so much time! 

I use wine to run 3D Coat, on my linux.  The Blender applink wont work since, Winne + coat will output files to some predefined  (hardcoded / not-editable)  folder, and blender will write to its own folder (than god folder it is editable). And on top of it  when exporting, some temp export.txt file is created where  another output path is created. I'm tired to  trying to follow what is exported  where. I was debugging these  output paths 2-3 weeks ago, so I may get the details wrong.  But the gist is, its more complicated than it needs to be. Make it dumb:  - in coat:  set user  customizable  exchange folder  - blender: set it to same  exchange folder.  That is it. No magic  exchange.txt files, or three places where things need to be set. 

They dev team want to transform the whole gui to qt.

minor python api bug. coat.pyi says cextension exists in coat (coat.cExtension) but it doesn't seem to. should be cCore.cExtension Though, maybe its different depending on whether you are using a cmodule or an addon script? Not sure just throwing it up

Just a thought, I was reading the latest here and looking at the add on MJonathan was working on   Makes me think a penny of time spend extending / making complete the python API is worth a pound of core 3dcoat dev time, because it unlocks the ability for the community to "Grow their own tools".  For example, there was API for lower level access to mesh and volume data, so we could use and interpret primitives (even if its slow) a bit like blender bsurfaces / their api for lower level mesh data, we could do things like implement our own uv seams algorithms or face modifiers ourselves. One tool I know is possible but haven't been able to do because I don't have access to low level mesh data with python is a decimation tool that uses an shape error tolerance instead of an absolute polycount for the output mesh. You could decimate your meshes until they reach some level of deviation from the original shape and then stop. But am unable to access those sorts of datas I think? Just speculating. And of course, if we can extend the GUI api to just allow spawning of python windows, then we could use external libraries for the GUI instead of requiring 3dcoat to be extended to do everything we need. For example, if I want to plot some mesh statistics on a graph, python has a ton of libs for that already, and 3DCoat would basically just be able to send off a data blob to python for interpretation. Same with custom editors and stuff. It would take a very long time for the devs to implement all the nice features for UI building that come stock in pure python libs, so it would be neat to just extend the amount of hooks and data access to allow for building out our own tooling etc   --edit: as for the second part it looks like cModule does exactly that, it comes with qt so you can build your own interfaces! super cool 

Same. Come on please fix this in 3dc2025.17 on apple silicon. I am really starting to tire of the mac bugs! Almost every time I sit down with coat, something comes up! I am here to make 3D, not report bugs. I dont have Windows and it is frustrating to use coat on mac, because bugs show up so often. Coat on mac feels second class.