Jump to content
3DCoat Forums

SUBD Topo?


Recommended Posts

  • Member

I'm a new user of 3DC with Lightwave 3D as the main app.

In a project I'm asked to reverse engineer (retopo) several scanned objects, trying to keep the difference between the two surfaces to 0.25mm or less.

To make my life easier I create a low poly version in 3DC and bring it back to LW for subdivision.

(BTW is there a subdivision method in 3DC that I haven't found yet?) But my problem now is that if I subdivide the model in Lightwave, where curvature exists, surfaces tend to shrink cause that is how the algorithm works. The SubD patch doesn't necessarily pass from the vertices that control its shape, however those vertices are the ones that 3DC uses to match the topo model to the original.

So after Subdivision I get a model that is slightly shrunk and may not be acceptable. Is there a way around this?

I definitely need to output a low poly from topo, cause I work on it further in Lightwave and It helps that its low resolution. So outputting a high resolution model from 3DC is out of the question.

I just wish there was a way to topologize using a subd surfaces mode. That would make my life so much easier.

Any ideas?

Link to comment
Share on other sites

  • Advanced Member
I just wish there was a way to topologize using a subd surfaces mode. That would make my life so much easier.

Any ideas?

I completely agree. There should be a Sub-D mode and the vertices that stick to the model should be the sub-D vertices not the flat polygon vertices.

The control points that the un-subdivided mesh represents can actually be outside or inside the subdivided mesh depending on if the surface is convex or concave.

And they can end up being close or quite far away depending on how much curvature the mesh has. Not good for making an accurate representation of the model you just spent

a bunch of time sculpting.

Maybe there is already a way, but I've been wondering about that myself.

Tom

Link to comment
Share on other sites

  • 2 months later...
  • 3 weeks later...
  • Advanced Member

Being able to sculpt a low-poly model while viewing the sub-d version would be very useful.

I asked on the Mudbox forum about this a few weeks ago, and they don't have it either. Several others jumped in and said that they wanted it.

This could be a feature (among all the other cool 3D Coat features), that would help it stand out from Mudbox and give users a reason to switch.

Of course, 3D Coat should also add Mudbox's additional tools to sculpt morphs (blendshapes) at the same time.

Link to comment
Share on other sites

  • Reputable Contributor

As for the Retopo SubD request, I'm not sure exactly what you're asking. You can subdvide the model in the Retopo Room, or you can have it subdivide upon export. If you check the Coarse option, it is supposed to help prevent the shrinking issue you are describing, although I've not used it yet...so I can't say how effectively it is. What I do is either add extra edgeloops where I want to preserve proportions or definition...even on a lower poly model. I then just export the low poly model and add the SubD there or in Max's Editable Poly rollout, I can select edges, apply NURMS SubD's, and weight the edges. I know Lightwave has edge weighting as well. Nevertheless, when you SubD the Retopo mesh in 3DC, every vert of that gets snapped, so I'm not clear as to what the problem is.

No there is not a SubD toggle (would be nice), but you can always UNDO (CTRL Z), if you don't like the result. I don't know that a toggle is really needed. You voxel object is the end result, so it's not like modeling in a 3D app and toggling the SubD to see how the model will look smoothed.

As for having the ability to model on a lower poly obj while seeing the high res update. I don't quite see the practical application here. Either sculpting on the hi-res is sufficently fast or it's not. All 3 sculpting apps can do so fairly well. And with that being the case, what would it buy you to be able to work on a lower poly and force the computer to have twice the calculations per stroke, by deforming both the low-poly AND the high poly too? This is actually one of the drawbacks from voxel volume sculpting. 3DC is having to do two calculations per stroke. One for determining which voxels to switch on and off, and then it has to resample the outer mesh to conform to the new changes...with EVERY single stroke.

With that in mind, it's actually quite a feat for Andrew to have brought voxel sculpting performance to it's current level. That's why I think he's spending more time in the area of Surface sculpting. He's on a more level playing field with ZBrush and Mudbox.

In all three applications, during the entire sculpting process you can step up and down in resolution and have the changes done in the lower Sub D levels propagate upward in the stack. However, the main thing to keep in mind is to plan out your sculpt a bit. Don't dive into the higher levels of resolution until it is absolutely necessary. This is the case whether you are sculpting clay in an fine art setting or digitally. Even if you are working with Morph Targets/Blendshapes. Don't fret about the small detail until you've gotten the main expression nailed down. Even in Character Animation, you can make a huge mess for yourself if you try to dive into the smaller movements before you've nailed down your key poses.

Plus, we haven't even gotten an overhauled Poly Sculpt Room yet, so I think we should cross that bridge first.

Link to comment
Share on other sites

  • Member

If you check the Coarse option, it is supposed to help prevent the shrinking issue you are describing, although I've not used it yet...so I can't say how effectively it is.

Abnranger I dont see this option in 2.5.2. I see merge coarsed mesh under retopology dropdown, but nothing about exporting a coarsed mesh. Can you point us to where that option is?

Link to comment
Share on other sites

  • Reputable Contributor

Abnranger I dont see this option in 2.5.2. I see merge coarsed mesh under retopology dropdown, but nothing about exporting a coarsed mesh. Can you point us to where that option is?

The Coarse option resides in the the Retopo dropdown when merging from there, and when exporting from the Paint Room, the export dialogue will usually have the option for that as well. I don't see anything in the Manual covering this, so that is something that needs more coverage, for sure.
Link to comment
Share on other sites

  • Advanced Member

The problem is that laying polygons onto a hp mesh with no idea how it is going to smooth and retain volume really doesnt cut it. You need to be able to see a 100% Opaque smooth subdivided mesh as you work. Having to hide the high poly, subdivide the mesh and then undo, and unhide the high poly is alot of work just to see if your support edges are holding your shape.

I often have to export to silo so I can see how the mesh is working. Its a complete pain.

My suggestion would be to have live subd in the viewport and an opaque preview mode bound to a sticky key.

Link to comment
Share on other sites

  • Member

On a related note does anyone else notice when you subdivide your retopo mesh in 3d coat that some of the vertices become disconnected? This happens to me all the time in certain areas of my retopology mesh. The verticies often lay on top of each other so you cant tell they are disconnected until you export. Close holes command doesnt repair it either.

Link to comment
Share on other sites

  • Advanced Member

We all have our priorities, and Andrew can only work on so many, but I would like to see the ability to sculpt morphs on a low-poly mesh while viewing it sub-divided - as well as have all of the normal tools available in the sculpt room.

I believe that 3ds Max allows you to set up an editable poly with a turbosmooth on top to see the end result while you sculpt - so Mudbox can't be far behind in implementing this (famous last words).

As a lot of morph creation is done, I think?, using low poly meshes, that would have to be a useful feature?

One goal, besides providing new cool features, might be to provide features that the completing products don't have, to pull users from those products, generate more revenue, hire more great programmers.

This might be one of those features? A linear colorspace workflow might be another even more important feature - as the compositing apps that will use 3D Coat generated art are definitely working in linear color space.

Link to comment
Share on other sites

  • Advanced Member

This might be one of those features? A linear colorspace workflow might be another even more important feature - as the compositing apps that will use 3D Coat generated art are definitely working in linear color space.

Why would this be advantageous?

sRGB to Linear gives good results for textures. Whats the advantage for 3DC users. sounds more like something you'd only need in film

Link to comment
Share on other sites

  • Advanced Member

Why would this be advantageous?

sRGB to Linear gives good results for textures. Whats the advantage for 3DC users. sounds more like something you'd only need in film

Because everything works better in linear. If you're not using linear, you should be.

Listen to this podcast, it explains in detail.

http://www.fxguide.com/fxpodcasts/Compositing_in_Linear/

There's a video that shows a bit about linear workflow advantages (but it has a number of errors) here:

http://www.pixsim.co.uk/video_tutorials/The_Beginners_Explanation_of_Gamma_Correction_and_Linear_Workflow.zip

and a pdf to go with it:

http://www.pixsim.co.uk/downloads/The_Beginners_Explanation_of_Gamma_Correction_and_Linear_Workflow.pdf

But the podcast actually does a much better job of explaining why you want to do any texturing (or anything that uses colors or light - which means .... um ... everything) in linear, as I think the video just shows how much easier it is to do lighting.

BTW - the fxGuide podcasts are really useful to listen to each week.

Link to comment
Share on other sites

  • Advanced Member

sRGB(Gamma 2.2) textures present no problems to a linear workflow providing the renderer is correctly setup to interpret them.

aside from this exporting your textures in linear space just makes them less intuitive to work with in external applications.

Link to comment
Share on other sites

  • Member

The problem is that laying polygons onto a hp mesh with no idea how it is going to smooth and retain volume really doesnt cut it. You need to be able to see a 100% Opaque smooth subdivided mesh as you work. Having to hide the high poly, subdivide the mesh and then undo, and unhide the high poly is alot of work just to see if your support edges are holding your shape.

I often have to export to silo so I can see how the mesh is working. Its a complete pain.

My suggestion would be to have live subd in the viewport and an opaque preview mode bound to a sticky key.

This is exactly what we need. While it is fine that you can bake down your voxel details to your retopo mesh and retain all details, it simply does not work so

easy for any other retopo work. Don't forget that you do not only retopo voxels but imported meshes aswell.

I once tried retopo on an imported car model - it was impossible because I only could guess how well it

would hold once subdivided. I don't understand that this has not been requested more often considering how important it is.

Link to comment
Share on other sites

  • Advanced Member

sRGB(Gamma 2.2) textures present no problems to a linear workflow providing the renderer is correctly setup to interpret them.

aside from this exporting your textures in linear space just makes them less intuitive to work with in external applications.

Did you listen to the entire podcast?

Link to comment
Share on other sites

  • Advanced Member

I did his course on it actually.

I dont understand what your suggesting. Just say it and stop telling me to list to the podcast.

Ok.

Creating textures with layers and blending, etc. works better in linear space just as outlined in the podcast.

It outlines simple tests you can perform to demonstrate this.

3D Coat was originally about 'coating' objects with textures, so ... this would be better.

Link to comment
Share on other sites

  • Advanced Member

Ok.

Creating textures with layers and blending, etc. works better in linear space just as outlined in the podcast.

It outlines simple tests you can perform to demonstrate this.

3D Coat was originally about 'coating' objects with textures, so ... this would be better.

:D thank you

I totally agree, Its an interesting point I hadn't considered. your suggesting correcting all the colour inputs right?

Link to comment
Share on other sites

  • Advanced Member

:D thank you

I totally agree, Its an interesting point I hadn't considered. your suggesting correcting all the colour inputs right?

I'm no expert on linear, in fact it's a bit fuzzy as it's been a while since I really focused on it (but I do use it and prefer to work with exr's), but I think that if we could do all of our color texturing work in linear, while viewing it with a gamma applied (so it will look 'correct'), that this would be a good thing, for almost any purpose.

Otherwise you get all sorts of crazy banding when blending layers - as all the math should definitely be done in linear.

I haven't tested 3D Coat, so Andrew could already be doing all the math in linear. That podcast outlined some simple tests to show the effects. I really need to listen to it again myself - as I've forgotten exactly how to do them!!!

Whether people use it or not is their choice, but ... it could be a selling point to get more interest in 3D Coat (if the other competing solutions don't offer it), just like ptex was a big feather in Andrew's cap when he had that working before almost anyone else.

I can clearly see the groundswell though, and everything will be linear at some point for professional graphic tools.

Link to comment
Share on other sites

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.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...