Jump to content
3DCoat Forums


Advanced Member
  • Posts

  • Joined

  • Last visited

Everything posted by gbball

  1. In the works I believe. Probably a fairly big undertaking though.
  2. I've started testing out the latest build 2022.49 with multires sculpting and it is very promising! Not exactly what I expected as it seems we're still forced to work with a triangulated mesh. But beyond that, the sculpt feel and sense of control was great. Being able to step up and down through subdivision levels is fantastic and the results always felt predictable. Being able to sculpt on a predictable mesh and get more detail or less detail when necessary works great. Things I noticed Good: Cavity effects on shaders looks way better...is there some kind of smoothing algorithm going on? Or does it just look better on a subdiv mesh? Mind you, I still have to dial down the cavity settings as I think the default cavity and bulge values are too high on some of the default shaders. Sculpting feels really nice and stepping through levels is smooth Performance is great on high poly sculpts Smoothing worked effectively Detail work is a lot better now. Way more predictable and easy to control. But this might just be me working at higher polycounts that I typically do while testing. Bugs/Needs Work: I noticed some weird graphical glitches in the UI, corresponding with slight pauses in the application. Not sure what caused it, but I haven't experienced this in prior versions of the program. I experienced frequent times when tweaking shader settings or when using the render room that the program would pause for a long time to do simple adjustments. The base level sculpt layer's topology is important...If I switch from a voxel mesh to a surface mesh, the topology isn't that clean and it leads to artifacts in the sculpt at high subdivision levels when I smooth things out. I noticed that all the different smoothing options aren't available, which is understandable since the topology is not dynamic...but I miss anti-bump smoothing for the reason above My mesh got destroyed at one point stepping down below the base level. The destroyed mesh was carried through to the higher subd levels. Missing: The ability to bring in a quad or ngon mesh to the sculpt space while retaining vertex order and UVs The ability to work seamlessly between a model or retopo mesh and a sculpt room multires sculpt. Being able to add/delete faces, edgeloops, extrusions, etc. and have that reflected in the sculpt. Subdivision style options Straight subdivision Whatever you currently have (it's great!) Catmull-Clark Microvertex Of note, but not necessarily good or bad: It seems that each subdivision level stores it's own shader settings...I wasn't expecting this, I don't know how to feel about it to be honest. It could be good or bad depending on the circumstance. I haven't tested all the brushes yet, but I was able to get some really great results that feel very similar to what I would want to achieve in a program like Zbrush for detail work and generating skin textures and wrinkles with brush alphas or just building up forms.
  3. Yes, What you've said is currently an issue and more so for those coming to 3D Coat from other applications and hoping to use the same workflows that they've used previously. 3D Coat can probably do what you want, perhaps aside from the displacement baking accuracy, but you might have to approach it in different ways to do most of what you want and possibly enlist the use of other complimentary apps for displacement baking. 3D Coat could do a better job of integrating a more familiar workflow as inevitably more and more people will discover it coming from other applications and be puzzled about the workflow. Blender had a very similar issue pre 2.8 where it had it's own way of doing things and people coming from other applications just could make it jibe with their expectations. If 3D Coat added a more unified, non-destructive mesh/sculpt system, and dropped rooms and instead had contextual tools depending on modes and/or what kind of mesh you're working on, it would make a massive difference in terms of adoption. There is still a lot that is unique to the program and that is what it needs to be appreciated for, but this might not be what the typical 3d artists are looking for.
  4. Yes, I agree, I know you've suggesting rolling the tweak options into the modeling/retopo space. From my understanding it is on the TODO list and it is definitely needed. I think in some ways the thing we're looking for is tantalizingly close, but being danced around in some ways. I also think that 3D Coat is uniquely positioned to be able to do something better than regular subdivision sculpting. Some of the work being done with curves is moving in this direction...Like we could conceivably see a polygon style context for curves manipulation where they can be extruded in realtime to make more curves and to create a real-time voxel sculpt representation. Polygon Basemesh <-> Curves <-> Ngon or Quadbased Microvertex paint/sculpt (would need true 3 dimensional vector displacement and solid sculpting tools) In some ways Microvertex painting has a lot of unrealized sculpting potential and could serve as the bridge that many are looking for. It seems like 3D Coat is reaching a point where texture and sculpt details can be meshtype independent. Truthfully, even Zbrush isn't using true catmull-clarke subdivision sculpting, they have their own unique algorithm. I wonder if there is a way now to step up and down through subdivision levels on a Microvertex sculpt since it basically turns the polygons into patches that can be subdivided. Is there a way to store/cache texture data volumetrically in a non-destructive process and bake to UV mesh when necessary...or if some kind of Microvertex/Subdiv mesh fusion could be implemented that uses vector displacement for sculpting/painting? @AndrewShpagin Does what I'm proposing make sense? How big of an undertaking would it be?
  5. @Andrew Shpagin I think the biggest difference is that in Zbrush or Blender, is that once you have a finished sculpt, whether you used Dynamesh or Scuptris or something, you can still keep sculpting afterwards. So the Dynamesh/Scuptris is for the blockout...then once you have your finalized quad mesh, you can step up or down on the quad mesh to add more details and at that point, the linkage between the low poly mesh and the high poly mesh works both ways, where changes to the high poly mesh affect the low poly one, and vice versa. I know 3D Coat has it's own way of handling this with cached geometry and decimation, which is fine...but that doesn't change the fact that voxel and surface sculpts are still more geared towards blockouts/preliminary sculpts as opposed to finalized sculpts and meshes as I'll outline below. In 3D Coat the blocking out stage in voxels and surface sculpts is very good. But once you retopo or autotopo, it becomes cumbersome to make sculpt-like wholesale changes to the mesh because you're using the modeling or retopo room tools...further, there is only a one way linkage between the high poly sculpt and the optimized retopo mesh and it works from Sculpt -> Retopo mesh. In my opinion, being able to control high poly sculpts with low poly geometry is more valuable...and if you think about it, the main selling point of a sculpting application...being able to move millions of polygons with a brush stroke or using a move tool. Wouldn't it be incredibly useful to manipulate a sculpt mesh by moving a single low poly face, edge or vertex? Or even better dictating an edge loop and extrusion to create a ridge where you want it? This workflow doesn't currently exist because changes I make to a retopo mesh have zero bearing on what happens to the sculpt. One solution would be 3 types of sculpt meshes in the sculpt room. 1. Voxels - 2. Surface - 3. Quad/Ngon with non-destructive subdivision stepping for detail sculpting after retopology. Moving from a Voxel or Sculpt mesh to a Quad/Ngon mesh would trigger an autotopo command and moving from a Quad/Ngon mesh to voxel/surface mode would either be a destructive step or it would cache the Quad/Ngon mesh and run the conform to retopo For early blockouts Voxels, Surface sculpts, Dynamesh, Sculptris are all great options...but when you want to finalize something and eventually bake, having a clean retopoed mesh where you can dictate the edgeflow is more ideal. 3 shortcomings with Surface sculpting when finalizing a sculpt. 1. Not always able to work on a uniform mesh since it works best when dynamic remeshing is enabled. As a result some areas of the mesh are low poly and trianagulated, while others are highly detailed. Getting the whole mesh to be the same density is a destructive step that requires going back to Voxels and returning to surface mode... Doing this will inevitably erode any small details you might have had, or worse merge together overlapping geometry. It's the equivalent of Dynameshing in Zbrush...good for a block out, but not optimal for the final stages of a sculpt as compared to a subdiv sculpt where topology conformity is moreso forced and automatic when you add subdivisions. 2. Baking becomes problematic when you have self intersecting geometry - Baking from a high poly subdivision level to it's base level would solve this...it would also provide perfect displacement maps, which has also been cited as an issue by some before. 3. Bringing an existing mesh into this space is a destructive step, where vertex order, UV information, and topology are lost. So while I can bring my mesh into both the retopo space and the sculpt space, I will be limited in terms of having things like topology adjustments on the retopo mesh being reflected on the sculpt mesh. @AbnRanger I don't mean any disrespect, but I've noticed over the years, you've consistently touted the alternate methods 3D Coat has for sculpting and dealing with high poly/low polygon stepping during sculpting whenever this particular request is made. However, I've outlined this issues as I see them and I think it does a disservice to the development of the application to stand in opposition to requests for workflow improvements that would be so beneficial to so many. We all love 3D Coat here and we all respect the hard work and innovation from the dev team. So I am sensitive in terms of feature and workflow requests. However, for the furthering of the app, I think it would be helpful if you got behind these kinds of requests for progress, especially when it's been made consistently by so many for so long. Hopefully I've outlined why it would be beneficial.
  6. For what it's worth, 10 years ago I wanted to start learning more 3d and I was looking for a digital sculpting application. I already used Blender a bit and I had tried zbrush and silo and Modo. A friend told me about 3D coat and I tried it and was hooked. I've progressed with the program over the years, but always had a feeling that I would eventually have to switch to zbrush to take the next step. 2 or 3 years ago I purchased zbrush to try an integrate it into my workflow, but it just never happened and I never use it. I haven't really found a situation where I needed it over 3D Coat. Now in the new versions of 3d coat the sculpting and possibilities just keep increasing, I can barely keep up. There are some things I am still missing. In particular, I'd like to see non-destructive subdivision stepping and the ability to sculpt on a UV'd non-forced triangulated mesh. This for me is the only thing really missing. The rest is polish and workflow enhancements. To go into more detail on why I want subdiv quad based sculpting...or at least something equivalent is that in 3D coat, there are ways to have a high poly mesh manipulate a low poly mesh, but not a really great way to work the other way around. It would be nice to have a way to manipulate millions of polygons using a simplified geometry. This seems tantalizingly close, but frustratingly not implemented. Another thing that would benefit from something like a subdiv sculpting approach would be baking. I think curvature, ao, normal maps and displacement maps would all become perfect if you could bake from your UV'd sculpt down to a lower level. But I'm onboard for the ride. There are things that can be done in 3D Coat that can't be done in any other program and it has an almost complete workflow. Sculpting gives me a sense of freedom that I haven't experienced in other tools.
  7. This might be your issue...you have to import the UV'd mesh into the Retopo room separately from the the sculpt room and then bake onto that UV'd mesh. If you bring the mesh into the sculpt room directly, it is used as a stamp or cookie cutter to make a sculptable mesh. You could also import in the first place directly to the Retopo room and then there is an option to copy the retopo room mesh into the sculpt space as a stamp, in which case a copy would remain in the Retopo room. To do it this way, you'll want to first bring the mesh into the Retopo space and then switch to the sculpt space and use the Import tool (pick from Retopo button) on the left to bring in the Retopo mesh as a stamp in the sculpt space. There will be no connection to these two meshes at this point...but there is a way to have the retopo room mesh 'follow' the shape changes you might make on the sculpt mesh if you check the 'Conform Retopo' mesh option at the top.
  8. This is a great idea! I've had very similar thoughts...check out this thread.
  9. Thanks! I still think it's a strange choice for a program like this...especially if it's turned on by default. There are a few edge cases where it becomes counter productive.
  10. @AndrewShpagin One thing I noticed in the new version is that when I am using certain dialogs, the 3d viewport dims...this is problematic because if I'm tweaking a material or doing hue/saturation/value adjustment, I can't see the result of my choices in realtime.
  11. Thanks for the responses. I think one of the cool things about 3D Coat is the innovative solutions to different problems. In some cases the solutions weren't copies of other tools, but instead a new and sometimes better way of achieving a similar goal. So I think rather than us asking for @AndrewShpagin and the team to copy this programs solution or this other program's way of doing things, we could instead make arguments for how and why a particular adjustment or tweak would align with certain principles that guide the development of the software. For example...this request Aligns with this principle. I expect that 3D Coat will keep adding innovative new features. But I also think that the tool is a mature enough point that certain principles can be implemented to make the overall program more cohesive and easier to work with. With the new Core API, I wonder how much of this can be done aside from the main development team. @haikallewas doing some cool work to optimize tools in the retopo room at one point. I wonder if the current scripting functionality exposes that same level of power to the user. If given enough customization capabilities, we could develop advanced tools and optimize workflows ourselves.
  12. As a longtime user of 3D Coat. I have been thrilled with the growth of the program and I'd like to use this thread as a way to share my thoughts and ideas on future development and overall 3D Coat design philosophies. Proposed Guiding Principles for 3D Coat. Optimized UI. Consistency of design. Organization of tools/rooms/menu options. Customizability. Extensibility. Stability. Performance. Innovation. Non-Destructive Workflow. Easy Management & Control of Complexity with Granularity. Optimized UI. What is the cost of an action in mouse movement and button presses? Wherever possible, repetitive tasks should be simplified into as few button presses as possible. Minimize clicks Reduce the need for popup menus where possible When this isn't possible, find a way to move quickly through the options and confirm choices. An object should own it's modifiers... For example...a curve should own it's modifiers, so if I duplicate a curve, there should be the a new set of modifiers on the duplicated curve. Consistency of design. What works in one room, should work in other rooms/windows For example, right mouse pop up options, Properties dialog Text and number field interfaces should all be the same...if some have a value scrubbing option, they all should. Organization of tools/rooms/menu options. Outliner with filters for each of different object types...Sculpt/curve/models/paint objects/etc...also the ability to parent one object type under another...making a curve the parent of a sculpt object. Customizability. Good - Really amazing navigation editor Bad - No hotkey editor window. Extensibility. Good - Core API addition Stability. Good - Keep fixing bugs Bad - Certain edge cases where tools don't behave as intended Bad - Some things break in new builds Performance. Good - The performance is already really solid Performance can always be improved. Move and Smooth on high poly sculpts with large brush size would be great. Innovation. Good - 3D Coat is one of the most innovative 3D tools. Non-Destructive Workflow. As much as possible, having mesh effects be tied to sculpt layers and modifiers would be optimal. Deeper integration of nodes would be good to see Easy Management & Control of Complexity with Granularity. Good - Being able to manipulate millions of verts using a brush...3D Coat does this and it's good. Bad - we don't have soft selection in retopo/model mode. Bad - we don't have comprehensive subdivision stepping...we kind of do, but the implementation is akward, because you have to switch between rooms - hence the Optimized UI principle. What does everyone else think? I think it can be easy to get lost in the trees and I think many of us use the program in different ways. But I'm sure there are some guiding principles that the community and developers can all get behind.
  13. It might be a bug or an edge case that wasn't accounted for. Have you considered mapping your 2nd stylus button to MMB and then using Blender style CTRL + MMB for zoom, SHIFT + MMB for Pan and MMB alone for rotation? That seems like it could work because there are no conflicts with the alt button.
  14. With some tools you might need to also hold alt with the other buttons you mentioned for navigation. That’s how it works with the default navigation style.
  15. I agree with everything on this list. Unnecessary clicks are a workflow killer. Also real-time previews are always welcome! Even better, non-destructive workflows. Soft selection and transformation space/pivot location controls are absolute musts. A lot of these things have been mentioned before by many people. Some are likely more complex than others, but overall these are important quality of life improvements that will make other areas of 3DCoat as enjoyable to use as the sculpting and the painting. I'd also add repeat last action to the list.
  16. Haikalle, had made some modifications to the Points and Faces tool in the retopo room a while back that was indicative of what would be possible with an advanced hotkeying system. It'd be nice if you could give the community to the tools to set things up in ways like this to really optimize workflows.
  17. Can we get a hotkey manager Andrew? A central place to review all available hotkeyable options and current keybindings. Nvil has the best setup I've ever seen. The way you handle camera navigation controls is on the right track IMO. I'd be happy to go over what I'm talking about with you. But in general, NVIL is powerful because it allows you make complex functionalities by doing things like using any key as a pseudo modifier key along with LMB, MMB, RMB to make unique tool combinations. They call it streamline tools. They also have customizable pie menus.
  18. @Andrew Shpagin Good change with the default IBL. Also glad to hear of Jama's involvement/consultation. With the shader advancements your making, any chance of getting some kind of better looking post process effect on the cavity setting to make them look smoother? Also, has the voxel smooth seen a performance enhancement? I didn't notice anything, but just checking if it's been done. Or maybe planned for. Thanks!
  19. Yeah, I know there is a bunch of stuff in development. I wonder what the rollout will look like. But I've heard mention of an updated rending engine, and nodes are there, but no formal announcements about it really, so I imagine it's still in development. I believe I've heard talk of paint room integration, but I don't know what that will look like ultimately...whether it will be tied to smart materials, or something more robust, or both. But it looks like there are plans for the use of nodes in several areas of the program.
  20. gbball


    This is incredible.
  21. It would good if the Layer in the paint room could 'own' the node graph and be used for sculpt layers, pbr procedural textures and adjustments such as levels/hsl/colorbalance/blur/etc
  22. This is actually a great idea...and could be integrated with the nodes implementation however that lands...in which case it would a nodes based adjustment or set of adjustments as opposed to just a layer. So to clarify to the OP, I don't think it exists currently, but it seems like there might be a pathway for a similar non-destructive workflow.
  23. Yeah, good point. I'm hoping we'll see smoothing get some of the new brush engine treatment.
  24. You can assign to hotkeys to different rooms. You can even stack the hotkey if you prefer that. So on mine I hit f5 to toggle between sculpt and render.
  • Create New...