Advanced Member popwfx Posted February 9, 2013 Advanced Member Report Share Posted February 9, 2013 I know you can Auto-retopo, and I know you can manually retopo. But what if you load a reference mesh and mostly manually retopo it, but want to auto-retopo some parts of it - say one material/surface of it. Is there a way to auto-retopo just that surface or section of the ref-mesh or do I have to load it separately and perform an auto-retopo on it? Is there a way to just draw some strokes on one surface for a particular Retopo Group and automatically retopo just that part, and then continue onwards and maybe manually stitch that auto-made section to your manually made groups? I can probably do this the tedious way, by splitting it up and loading separate meshes into separate 3DCoata projects - but it would be nice if there was a way to mix and match in the same project/object. thanks Quote Link to comment Share on other sites More sharing options...
Reputable Contributor digman Posted February 9, 2013 Reputable Contributor Report Share Posted February 9, 2013 Ok, I going to give you a possible workflow to test and see if it meets your requirements that you presented in your post. First, a little understanding about "Merging to Surface Mode" It is not resolution dependent. That means your model will be merged at it's true polygon amount except the model gets triangulated upon import. Also I believe smoothing groups are respected. When resizing the object during the merging process your polygon count of your model does not increase. Do not worry about what the estimated polygon is reading.. Do not subdivide the model. You do not lose any details of your model as it is polygon based and not voxel based. Workflow: Merge your model as the image shows, look carefully at the Merge tool selections and select the same ones I have selected in the merge tool dialog box. Resize your model for ease of working on. Press enter to merge and say yes to the scale question, 3DCoat will automatically create the layers for your sub objects. Your lightwave model should come in with all the separate sub objects in their own layer. I have not tested a lightwave file as it still crashes my linux version of 3DCoat. Now you can work on each layer independently of each other, running auto-retopo or working manually plus a combination of both on each sub-object etc etc... Also you can work on several layers at time or hide polygons using the hide tool. Lots of options here... In the imaged I uploaded I have merged a Saturn rocket ship obj model with all it's sub objects in their own layer. Appox layers around 30 or so... The model was appox. 16,000 polygons and after merging it was 32,000 with the triangulating with no lost of detail and the smoothing groups respected. Now, like any workflow, there are always things you run across, I do not use this method very much as my need for it is not as great as yours. You would have to use it often to find possible glitches in this method. In some ways this method is more powerful to use than merging a reference model into the paint room for retopoing as you have more options. Hope this helps some and if you have a question, contact me through Skype... The obj model is not mine, I just used it to show an example... For future questions I have a suggestion when you use the word " Load" could you be more specific in giving the method of loading, like are you Importing a reference mesh into the paint room, merging a model into surface mode or voxel mode etc as it will help on answering your questions 1 Quote Link to comment Share on other sites More sharing options...
Advanced Member popwfx Posted February 9, 2013 Author Advanced Member Report Share Posted February 9, 2013 thanks digman Sorry about not being clear. I was importing externally created models into 3DC as a reference mesh to retopo. It seems the most common workflow is to retopo your voxel mesh. 3DCoat is so awesome even if you are not voxel sculpting, it has such great retopo tools I often use it just for that. Seems all my questions are answered in terms of the sculpt room as that's what everybody mainly uses in 3DC I guess. but 3DC is just a part of my workflow and I often have 3D scans or externally created models to retopo,UV, and paint - all without voxelizing them. No skype on this client desktop at the moment so I'll try you later, but in the meantime, let me ask you a dumb question. Can you explain "merging" in 3DCoat terminology. The manual says: Merge into scene: Lets you merge a retopologized mesh into scene. All details from the reference mesh will be baked into a new object. Additional extrusion can also be applied. A UV-set is required, but if you have not assigned a UV-set it will be generated automatically. Merge hull in scene: Merges outer mesh over reference object into scene. This function works like Merge into scene but is optimized for a multi-object reference mesh. It prefers outer points while baking surface. Merge raw patch: Merge a raw patch into the scene. No baking will be performed. This command can be used to merge a new object without projecting details from the reference mesh But what does that mean? Is merging just for taking voxel sculpt things into polygon? The manual couldn't be more obtuse when using the same words for the definition "Merge into scene lets you merge into scene" lol. What is getting merged where? What is merging as far as 3D Coat is concerned? If I read what you wrote correctly, you are telling me how to take a reference mesh, voxelize it and then make it polygon again by merging it? In my situation I am not using voxels. thanks again for your reply! Quote Link to comment Share on other sites More sharing options...
Psmith Posted February 9, 2013 Report Share Posted February 9, 2013 Merging is really just a combination of transferring your created topology (automatic or manual), and its associated UV coordinates - along with any sculpted detail, if that exists (which is usually of much higher resolution than your Retopo Mesh) - into the Paint Room and Tweak Room for adjustments and texturing. So, there is this aspect of "Baking" which takes into consideration any existing Voxel or Surface Mesh detail - for projection as real time Normal (Per-Pixel) or Displacement (Micro-Vertex) data - which you can add to by Painting in the Paint Room. It's probably one of the deeper concepts in 3DC to get a grasp of. If you are not using a Voxel or Surface mesh as a "base" for Retopology, then no "Baking" takes place - and you only have the option of "Painting" this detail from within the Paint Room. Greg Smith Quote Link to comment Share on other sites More sharing options...
Advanced Member popwfx Posted February 9, 2013 Author Advanced Member Report Share Posted February 9, 2013 Merging is really just a combination of transferring your created topology (automatic or manual), and its associated UV coordinates - along with any sculpted detail, if that exists (which is usually of much higher resolution than your Retopo Mesh) - into the Paint Room and Tweak Room for adjustments and texturing. So, there is this aspect of "Baking" which takes into consideration any existing Voxel or Surface Mesh detail - for projection as real time Normal (Per-Pixel) or Displacement (Micro-Vertex) data - which you can add to by Painting in the Paint Room. It's probably one of the deeper concepts in 3DC to get a grasp of. If you are not using a Voxel or Surface mesh as a "base" for Retopology, then no "Baking" takes place - and you only have the option of "Painting" this detail from within the Paint Room. Greg Smith Thanks Greg, that clears things up a bit. So just a to clarify, there are 2 kinds of baking - texture baking with the texture baking tool - and this merging/baking of taking voxel or surface meshes from the Sculpt room and prepping them for painting? You also mention "and its associated UV coordinates" - but how can a Voxel/Surface mesh from the Sculpt Room have UVs already - doesn't it need to be "baked"/merged to polygons so that you can create seams and unwrap in the UV room. Sorry - just trying to make sense of this. I gather the most typical workflow is to use the high-res sculpting as your initial modeling steps - it gets a bit confusing when you skip the sculpt step and bring your own external content in for retopo/uv/painting. thanks though! Quote Link to comment Share on other sites More sharing options...
Reputable Contributor digman Posted February 9, 2013 Reputable Contributor Report Share Posted February 9, 2013 You are misreading my statement. This is due I would think because you normally do not use the voxel room. Once you get on Skype again this is much easier shown than typed in a post. There are two types of meshes in the voxel room Voxel mode--- 3D square cube pixels. Surface mode--- Real polygons. Merging a model as a reference model for retopoing into surface mode is really no different than importing a reference model into the paint room for retopoing but with a lot more options for retopoing methods. Remember surface mode in the voxel room is real polygons. You are not trying to sculpt here, I know that. Do not just think of surface mode in the voxel room as purely for sculpting. That is an important concept to understand. In light of the above read below... First you are not voxelizing the mesh. Voxels are 3D square cube pixels. You are not first converting the polygon mesh to voxels but directing merging the polygon mesh into a surface polygon mesh with no lost of details. It would be better to rename surface mode to polygon mode as surface mode is somewhat unclear... A scanned polygon mesh stays polygons in surface mode and you do not have to increase resolution to capture the details of the model This is really no different than importing a reference mesh into the paint room but with a lot more options for methods of retopoing. Reread my first post and I think you will understand. I do understand than you just want to retopo and not do any sculpting. A one half hour skype session when you have time will clear this up Quote Link to comment Share on other sites More sharing options...
Psmith Posted February 9, 2013 Report Share Posted February 9, 2013 The second kind of "Baking" requires the existence of a Voxel or Surface mode sculpt to act as a "base" form for your Retopology work. If either of these exist, you would still need a Retopo Mesh to act as the recipient of the Normal, Bump or Displacement data - which becomes editable in the Paint Room, upon "Merging" there. Nothing is "Baked" unless there exists a Retopo Mesh to bake into - and Voxel or Surface data to "Bake" from. In your case, where you simply want to use a "Reference" mesh as the "base" for your Retopology work - there exists no high frequency detail information to be "Baked" - so no "Baking" takes place - only "Merging" of the Retopo Mesh and its UV coordinates into the Paint Room (where you can add high frequency detail with the texturing toolset that exists there). Perhaps we use the term "Baking" to describe too many processes, over here at 3DC central. It is not the most descriptive word. Greg Smith Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.