Some more details, the folder is set to passthough already and still same result, also I This is Textura 2026. I also tested on trail version of full 3d Coat 2026 and the same issue. I have had someone test on the project file on 3D coat 2024 and the issue didnt happen. Is  this a 2026 issue?

Hi The group layer is set to Pass-through ? Pass-through: Select the method of color blending between the previous layer(s) color and the current color

Hi, im struggling with an issue with grouping in Textura 2026. When I group layers the colours seem to shift to blue. Anyone have any idea why this is happening. Gold is normal, blue is when I add layers to a group

Got it! Will be sending soon.

i am experiencing the same issue

Please send me `Documents\3DCoat2026\Log.txt`. Also, check the box next to "Python Terminal" and restart 3DCoat; this should reveal the errors explaining why AVIF isn't working. .

Yes.

Are you having the same problem using the Legacy version ?

Congrats peeps, this update is making the rounds

Each time the context menu is opened from the layers panel, most of the layers are appended and duplicated to the 'Take another layer as a mask' sub-menu. Version: Textura-2026-12 layers_panel_menu.mp4      

Hi Carlosan, I hope you are well. Sorry, I should have mentioned....I can confirm that I did test the 'Legacy' version as well and that was marginally better but it also froze. Even painting on the terrain there was a small lag and jittering artefacts on the mesh in the view port. 2024 just seems to handle my workflow much better....a little heavy but good. That's why I'm super excited about 2026...to be able to 8K without that heavy feel. Cheers Ady  

So, it's a no with symmetry.

I need advice on how to maintain good workflow between quads and voxel/surface objects.  Sometimes I need to work in other applications, and in that case, voxels are out of the question. My standard workflow for creating armor plates and similar elements is quick shapes in Sculpt Room, then a "clean" retopo mesh(mostly manual) + shell and marked sharp edges, then import to Sculpt Room. I often need to make edits in Sculpt Room, so I thought that Conform Retopo mesh option might be useful, but even with very small adjustments with Move tool (doesn't matter if it's standard Move or Topology move), retopo mesh will be messed up pretty quickly – collapsed edges, vertices, etc. So, is there any other way to maintain adjusted quad mesh without constant manual fixing in Retopo room?

Could you test 2026 legacy for texturing 8K terrains ?

Oh, one question about new Splitter tool - will this eventually work with symmetry? That would be cool.

Hello, I am testing the Linux Flatpak release of 3DCoat 2026.11 on PikaOS GNOME 49.4 Wayland with an AMD Radeon RX 9070 XT. I found two separate issues. Space Navigator / 3Dconnexion 3DCoat initially reported: Space Navigator daemon for 3dconnexion devices: NOT found even though: spacenavd was installed spacenavd.service was running spnavcfg detected the device correctly 3DCoat 2025 native Linux version worked correctly The problem was Flatpak sandbox access. Adding: flatpak override --user \ --filesystem=/run/spnav.sock \ com.pilgway.coat fixed it immediately. After this change: Space Navigator daemon for 3dconnexion devices: FOUND The 3Dconnexion Space Navigator works normally. AMD RX 9070 XT performance problem Unfortunately, the GPU performance of 3DCoat 2026 Flatpak is unusable. The same scene: 3DCoat 2025 native Linux: ~145 FPS 3DCoat 2026 Flatpak: ~3 FPS The host system correctly uses the GPU: AMD Radeon RX 9070 XT radeonsi gfx1201 However, inside the Flatpak: Runtime: org.freedesktop.Platform/x86_64/22.08 The AMD Mesa driver is missing: /usr/lib/x86_64-linux-gnu/dri/radeonsi_dri.so Only: intel-vaapi-driver is present. The application logs: MESA-LOADER: failed to open radeonsi failed to load driver: radeonsi Could somebody with an RDNA4 AMD GPU confirm this? Possible questions for developers: Will the Flatpak runtime be updated? Is 3DCoat 2026 expected to support newer Mesa versions? Is there a recommended workaround for RX 9000 series GPUs? System: PikaOS GNOME 49.4 Wayland AMD Radeon RX 9070 XT Mesa 26.1.4 host Flatpak runtime 22.08 I am currently running the native Linux version of 3DCoat 2025 on the same system without any problems. The only workaround I needed was to provide Python 3.10, since PikaOS ships with a newer Python version. Thanks!

Maybe that is not the right place, but I need to figure out something. Right now, I'm mostly using version 2025.17 and sometimes I need quick UV maps, so here is a scenario - I have multiple objects in Retopo Room, all but one are hidden. I will work(mark seams) on selected//visible object, but when I hit Unwrap, 3DCoat will unwrap all objects, absolutely ignoring the fact that all the other objects are hidden/not selected. That's a bug, right?

Hi, Just an FYI I have been testing workflows in 3DCoat for texturing 8K terrains....regarding the performance of 3DCoat2026-12 on my older system with a 2080ti GPU....I've uninstalled it and reverted back to 3DCoat2024-13. It seems to handle things much better. I'll be testing 2025-17 tomorrow.... I could not see or feel any performance improvements with 3DC2026

api bug with scaling objects: """ On a mirrored instance (negative-scale / flipped transform): Object has a negative determinant / reflected axis in the matrix. GetScaling() returns positive components only. SetScaling(GetScaling() * k) (or any SetScaling with those values) drops the mirror and/or corrupts the intended transform, while editing the 3×3 directly by k preserves the flip. """ basically, if you try to use the api builtins to set scale half and set scale double, it will corrupt the matrices, ignore the scale sign, and return mangled transforms. Expected behavior is that a round trip scale 0.5 and scale 2x restores original object scale / position including instance flips