tiburbage

Yeah, per Layer 0 being "an auxiliary layer" which should be left alone, that was my take on it, and I'm planning to never touch it -- just build layers on top of it. Carlosan, thanks for the long explanation and suggest you might want to add that to the base documentation if it isn't there already. The thing I was really trying to get clear in my mind was if Layers can store data other than shading data (PBR+Depth+Opacity), what are they? Perhaps pose data, which I think of as Morph (aka Blend Shape) data, is just stored as displacement? Just curious... But since the topic became about Layer 0, is there anything you can ONLY do via Layer 0, or can one simply leave it alone and build what you need in layers on top of it?

In this video, which is about the 2022 Multi-Resolution Sculpt Object support, I think to demonstrate how mesh data changes propagate through a Multi-Res Level stack, the Pose Tool is used to pose the character's hands. But what I wasn't expecting is that the pose is being stored on a Layers panel layer. What is being stored there? Is it vector displacement data? Or can a Layers panel Layer store more than just pixel data? This raised broader questions I've never seen fully explained about Layers panel layers and all of the types of data they do and don't contain, and whether the mysterious Layer 0 is just an ordinary Layer or has some "special properties"...

I'm not sure, really. Probably just by watching a bazillion videos like the more recent videos on the Paint Room, and I dug in a lot on videos relating to Freezing and the Freeze tool, and the Pose Tool in the Sculpt room. Just trying to understand how 3DCoat masking support relates to ZBrush, which has excellent, very fluid masking support. Because 3DCoat doesn't support grayscale layer masks as in Photoshop or Substance Painter, but rather uses Paint Layer Transparency to stand for what black means in regular compositing. I'm still not 100% sure whether the Freeze state of pixels separate data, or is just Opacity/transparency data. But 3DCoat will at least translate between Opacity/Transparency state and Freeze state.

The "masked" areas are in 3DC terms in a "freeze" state. You just need to clear it: Ctrl+D Freeze > Unfreeze All.

Also noticed by accident that when a Smart Material is active, you can preview and transform those mappings in the viewport via Ctrl+RDRAG. I guess it is showing the top-most layer (image) of the material...

This past week, I learned a couple of useful things I hadn't been aware of before: 1) View Masking: If you've created mask layers, and you want to be reminded of what the defined mask is, just Ctrl+CLICK on the layer in the Layers panel. The transparent (masked) regions will display on the model as frozen/masked. In fact, this will work for any layer which has some transparent regions. (see also Layers popup menu: Freeze Transparent Pixels). 2) Preview Projection Mappings (Cube, UV, Cylindrical, Spherical) to get visual feedback while tuning them and without the Smart Material Preview: After selecting a Smart Material or Stencil, in the Material/Stencil Control Panel, select e.g. Cube Mapping (or UV-Mapping), then click Settings to its right. Show Preview ON. Now you can use the transform buttons on the Control Panel or the 2D gizmo to offset, scale, or rotate the mapping and see what you're doing. I like the gizmo: RCLICK on the tiny center/pivot square to open (if you don't see it, click the recenter button on the Control Panel). For Cube Mapping, you can then also see, and so make visual adjustments to how the projections blend using MapSharpness. It works nicely for UV-Map projection as well, but the Settings are different. Not exactly sure what map we're seeing for the Smart Materials, but it is a big help especially in dialing in the scaling.

Wow. Thanks for the thoughtful reply! I just finished a re-watch of the "Quick Start Smart Materials" video series just to firm up my understanding of the design intent and usage of Smart Materials. The point at which I became less enthusiastic was with the Attach To Layer workflow, which is an attempt to introduce a procedural method to Smart Material use. The fact that it isn't currently possible to render Smart Materials directly in viewport (meaning without the Smart Material Preview window) is a significant workflow limitation, and gets worse if you have multiple attached SMs because the Preview will only render one at a time so you can't get a fully composited view of your model. My hope is that Andrew will continue to think about how to allow the texturing process to be more procedural (while still remaining "artist friendly", and not turning into a Tech Artist tool), and to make improvements to viewport rendering which might, ideally, remove the need for the Preview window entirely. To be performant, that would almost surely have to be GPU rather than CPU based. An yes, in terms of the Layers structure, there seems like a lot of room for user friendly improvements there. I could also mention the aging Texture Editor window, but I think I've said enough!

AbnRanger, do you have any idea if there are any significant improvements to the Paint room on the agenda for 2023? I looked briefly at their Trello Roadmap, but couldn't tell anything really from that. One of the things I've enjoyed most since I adopted Blender as my core hard surface modeling app is how much insight we get regarding what major initiatives are planned, what's actually being worked on, even generally when these things are scheduled to arrive. I'd love to see Andrew open up a little more about where he sees 3DCoat as headed, and what the priorities are. All of their work in the Sculpting area has paid off with some real recognition among Concept Artists, and I think another wave of focused updates in the Paint room could gain 3DC more recognition in the production asset creation area, especially for game asset development.

Sometimes I think it just means they are working on something larger scale, which may take weeks rather than hours or even days to do. It is common for a team's main branch to become "frozen" for a period of time while somebody has a big set of changes in their sandbox they need to get stabilized before they can push that to main and free it up for more general submissions... Just a thought.

I really like the way that Substance Painter synchronizes the 3D and 2D/UV views, and am suggesting some of what they do would be a real enhancement to the Paint room as well. 1) You can cycle among the main texture channels via a single hotkey (C) and both views show that data shading (basecolor, roughness, etc.). You can also choose these from a viewport toolbar dropdown list. 2) You can cycle among the baked mesh map channels via a single hotkey (B) and both views show that data shading (AO, Curvature, Position, etc.). You can also choose these from a viewport toolbar dropdown list. 3) Hotkeys such as focus (i.e. Shift+A, Shift+Z in the 3DC viewport) are also applied to the 2D/UV view, making it easy to re-center the view and show the full extent (in UV terms, frame the main tile).

I find the current draw method for showing selected faces to just be visually distracting, and would much prefer the approach taken by Blender or Maya: just highlight the faces' boundary edges, and then within that use a more subtle, semi-transparent shading over the face surface itself. Simple, clean. Sorry if this is a dup, somebody else has to have complained about this.

+1 to being able to change WIREFRAME COLOR on the fly. There are several feature request posts relating to wireframe color in both a 3D mesh edge context as well as in the UV view/Texture Editor. It's fine to have a global default, especially for mesh wireframe, but in the Texture Editor, a good high contrast color depends on the particular channel being displayed, and its specific shading. For me, I'd often just be setting it to BLACK or WHITE when the current channel is base color, but for grayscale channels, some hue tinting is needed. Perhaps the user could select from a Color Swatch set, or just a dropdown of a small set of pre-defined colors including just black and white... Carlosan, let me know if I should create a more visible separate [FEATURE] request for this. I think it is needed.

Carlosan, I hope the team will give us some insight into where the texturing system is heading, and in a general sense, when we will start to see the implementation of it in the Paint room. I came here to "+1" being able to have Masks on Layers in the Substance Painter sense of that because after just working through a complex vehicle model baking/texturing process using Painter, it is obvious how core it is to getting the work done in a layer-based system. It is more than just being able to use various auto-masking methods, with manual tweaks on top as needed, it is about fast visualization (e.g. Alt+Click on the layer (or folder/group) mask swatch to immediately switch the viewport to Mask mode). However, I've also had the sense that a layer-based system, despite how familiar they are to most artists, can tend to break down in the face of larger scale, more complex model projects, especially when you are having to manage many texture channels at once. If Andrew et al have a "next generation" vision for 3D painting/texturing, I'd love to see that happen. Painter is certainly a powerful 3D texturing solution, and provides many nice UI/UX features, but I think there is room for competition even in the pro/production asset creation space if a solution which is robust, has high ease of use (more artist than tech artist focused), where the workflow is as non-"destructive" as possible. There are certainly many who really want a production texturing alternative to the Adobe/Substance 3D offerings. Based on many Google and YouTube searches, I don't get the impression that Textura or 3DCoat proper have had much of an impact on production 3D texturing (e.g. for game or vfx asset creation). I think the opportunity is there though.

Don, where you say "I have asked Andrew to change the name..." do you mean changing the use of "Object" for the Pose action choice? If so, really think that should be "Part". +1.

The "standard practice" I've adopted when working on the final res model version is to just go ahead and do Edit > Calculate Occlusion and Edit > Calculate Curvature once the model is more or less finalized, but before kicking off the Paint process. Then you have full control. It should also be mentioned that mesh map baking will take your tangent space normal map into account if you have baked one, so be sure to include that in your project if so before hand. When assigning Shaders in the Sculpt workspace for shapes-in-progress, some shaders will bake automatically, but in that WIP state, I *think* you just let it do what it needs to do...

I thought the topic of fast access UI/UX, and user-extensible UI might be worth discussion. 3DCoat is really powerful, but also quite complex. Entry points for functions are spread across: Tools toolbar, top toolbar, main menus, panel toobars, panel inputs, RCLICK popup menus, viewport overlay toolbar incl Camera icon with it's own dropdown menu. We tend to map the most high frequency usage functions to keyboard shortcuts, but there are practical limits to the number of those most people can absorb. One of the features I've most liked in my other apps experience are faster access, focused UI elements like pie menus, popup panels (which themselves can be hotkey-mapped). Pie menus I don't know offhand whether you can define your own in Blender, but I really like the ones they define, and those implemented by Addon dev MACHIN3. One hotkey assigned and you then have fast mouse/pen access right at the cursor (where you are working) to a set of related functions. For example in a 3DCoat context. Custom menus Back when I was using LightWave Modeler, it let you define 3 custom popup menus. I refined those to the point that I did almost everything either via hotkey or from my own focused popup menus. No more poking around the UI trying to find a function entry point. Custom panels ZBrush calls them "menus" but visually they are effectively popup/dockable panels. They can be built from any UI element defined in ZBrush. They provide a great user-extensible way to create your own fast access to related functions. I use them to define my own little "mini workspaces" to group functions most used for particular phases of asset creation, and now rarely poke around in the main panels for entry points. I don't know if 3DCoat is structured internally to make one or more of these either possible or practical. But thought it worth discussing.. PS: Some devices provide some form of extensibility (like pie menus), I use a Wacom Intuos Pro, and Don Nash has demonstrated in more recent videos that the 3DConnexion SpaceMouse Pro has similar capability. So lacking such facilities in the software, these provide an external solution. But that doesn't help the user just using keyboard and mouse.

I just want to say +1 to your thoughts. I think focus in UI/UX is so important, especially in such a complex environment as 3D. I really prefer that new popup/RCLICK menu in the Modeling workspace to the giant "wall of icons" SPACEBAR tool panel. Of course, the SPACEBAR panel as an "extensibility" function in it i.e. the 1-0 slots at the top, but that's another subject. I also liked the Maya Modeling Toolkit and UV Toolkit in Maya when I was using it.

Carlosan, I think he's referring to the automatic mesh map generation that occurs when you select a Smart Material which requires them but they haven't been baked yet. If I'm reading that right, the question would be what settings does 3DC default to in that case? I think for AO it should be Separate Paint Objects=ON, Lighting Render Target=Into New Layer as you show above.

Well, whether bug or not, you can perform the operation with Symmetry OFF, then turn it back on, then mirror across the symmetry plane...

Thanks for the comment Don. I wish at least that the tool had the option to leave its selections as Frozen. It does respect a predefined Freeze region on entry... Even better would be to have a better more general Parts selection method: perhaps something analogous to the Magic Wand in the Paint workspace in the way it can apply Freeze by Picking Parts.

A cool feature of the Pose tool is that in Mode=Object, you can easily select Parts of a Sculpt layer. Is there a method to then move the selected Parts to another or new Sculpt layer? I'm aware of the Split Tool, and Cut & Clone, where I can use Lasso strokes to do that outside of Pose, but I just liked the Object (aka Part) quick select method in Pose...

2202.47 I'm finding that Shift works just fine to invert what gets removed when you are using a Lasso stroke with the Cut Off tool with Symmetry OFF, but with it ON, everything on the layer gets deleted. If that should work, let me know if you can reproduce. If yes, let me know how I will report it if you want. If it works for you, I'd like to try to figure out what other setting is causing my attempts to fail.

Thanks for the bit of feedback so far. Would still like to hear from pro game or vfx artists who can use whatever they want to, but have chosen 3DC for realistic texturing. Interesting aspects include baking capabilities, especially tangent space normal maps, ability for smart materials to use mesh maps, support for MatID masks, ability to procedurally generate and visualize a ray projection cage, anything else about the process which is non-destructive. I'm not as concerned about mesh map baking because I'm considering purchasing the very highly regarded Marmoset Toolbag specifically for baking, so it really boils down to the the texturing part. @Skye You mention generators. Does 3DC not have tileable procedural noise generators? Can't 3DC Smart Materials leverage mesh maps like position, cavity, curvature, AO?

After much not very helpful/informative Google'ing, and reading Polycount etc. posts, I haven't yet read any really good, objective comparisons of using 3D-Coat's texturing and baking toolsets as opposed to taking the model into Substance Painter. I was hoping (Don, Carlosan, others) might have some useful comments, or links to info you think helps to shed light on this. About as far as many related posts go is to say "3D Coat is great for hand painted texturing", but that isn't very helpful in understanding how the two actually compare for doing realistic hard surface model texturing. My work is almost all in hard surface, though also includes other "environmental artist" more organic objectives like foliage, rocks, etc. etc. I'm hoping there are some forum frequenters here who do hard surface realistic texturing work at a professional level, are familiar with Painter, but have chosen to do the work in 3DC instead. The news that 3DC is incorporating support for Megascans is great, and is certainly a differentiator from Substance, especially for those of us aiming our work in an Unreal Engine pipeline. I should point out I'm already a 3DC license holder, and plan to upgrade, so I don't need to be sold on the feature set outside of texturing. TIA

I've been going through the YouTube videos related to retopology, and for several of the Add Geometry tools, and the Edit Geometry Move tool (with component mode=Vertices), there is reference to welding behavior when RDRAG'ing verts over verts. The videos show a red dot showing when the verts are close enough to snap and weld. I'm not seeing that red dot in any context, and don't have the sense the welds are occurring. Using 4.7.06. I'm defaulting to the OGL version, but tried the DX one as well, but same thing. Tried with and without Auto Snap ON. Is this a known issue, or might I have some setting "wrong" which is causing the failure?