ZBrushCentral

WIP Experiment

ZBrush Artworks
#1

This is an experiment in taking a model made in ZBrush and translating it into an animation model in another package.

When modeling for animation, you need to keep in mind the layout of the polygons. Doing this and still trying to be creative and keep an eye on the aesthetic look of your model is quite the strain on the brain pan. You can get good at this, but I find the back and forth process draining.

That’s why it’s good to have a design worked out on some other medium, such as paper or clay, before starting. Most big productions sculpt and scan the models to get the 3D data to build an animation model off of. Clay has fewer rules to work with than polygons.

Not all of us have digitizers or access to scanning equipment. I’ve been playing with different “poor man’s digitizing methods” over the years. This one takes advantage of ZBrush’s easy 3D sculpting features. This is a nice middle ground between clay sculpture route and building from scratch in a 3D program.

I had done something similar before, but this time I added a step that really helped using texture master.

  1. The original ZBrush model
  2. The wire frame, showing the denseness of the polygons. Not good for animation
  3. A new texture was applied to the model and painted on with texture master. The lines represented the spline guides for later in the other 3D package.
  4. The model and texture are exported and brought into Lightwave 3D. From here, splines are created by selecting the points along the painted guides created in ZB. These were first created loosely, and then cleaned up when I was ready to “patch” them. Different programs handle these entities in various ways.
  5. The reconstructed model, made from the spline patches. Creating this is where the more technical side of modeling came in. But that was all I had to think about since the character design was already done. Note the difference in polygons from the original.
  6. A shaded view of the new, “lighter” model in subdivided mode for added smoothness.

I was happy with the results since the new model matched the original pretty well. I’m not sure if I’ll do anything more with this model, but I thought I’d share the process even though it may not be everyone’s cup of tea here. If anyone has similar techniques, additions, questions, feel free to post or ask

#2

What a nice technique :slight_smile:

Thank you Ken for bringing this up - this seems to be a great workflow integration of ZB :+1:

Now i just have to find out how to do the splines in lightwave :wink: (if you happen to have some good tips on that matter i would be delighted to read them)

I’m eager to get to work asap hehe

Again thank you Ken

#3

Hey this is great Ken! The dense models ZBrush makes has always
bothered me as I’d love to take my ZB models into Max and animate
them, without breaking the back of the computer. But ZBrush was never
really intended to be a model exporter and painter for other apps, so
it requires creative methods or 3rd party applications to accomplish the
lowering of polygons to an acceptable level without losing the detail.

I’d be interested in knowing how long you spend on step 4?
And then after this do you create your textures with new UV’s
and paint them back in ZBrush?

Thanks for the post. This is the meat I really like! :+1:
Upham :slight_smile:

#4

Hey, I was just thinking about painting wires on a ZBrush mesh the other day, looks like the technique works well.
:cool:

#5

This is very interesting Ken ! But i have a few problems by translating your little tutorial into german for me so that i don’t understand everything from Step 4.
I know that you select the points from the model which are painted on the texture as guides. Okay, but what do you do then ? :confused: :confused:
Do you use the selected points to make splines of them ? Sorry, maybe you could explain it more detailed. That would be very nice. Thx
Dorian

#6

Steps 4-5, creating the splines and patching them, took about a day and a half. The patching part is where the bulk of the time is spent. I haven’t really used Lightwave splines until now. Drawing splines in blank 3D space makes it rather difficult to visualize the right form. The ZB guide model took out any guess work.

The first image shows a close-up of the selection of points for a spline. After they are selected–in order— Ctrl+P is pressed to make a spline.

To make a patch, you must select the splines in a clockwise or counterclockwise order–and be consistent in this order for the whole model. It is important that their corners share points. Keep the division of the patch low so that the model doesn’t end up as dense as the ZB model. It may take several attempts to arrive at numbers that capture the form and detail.

You must keep in mind how each patch will meet with the other and adjust the Perpendicular and Parallel numbers accordingly. But there were times where I had mismatched patches–partly to terminate a level of detail along the model. Later, I would weld points to merge them together.

The spline cage I made here is actually rather messy. I could have cleaned it up more, but since I knew I was going to patch at low resolutions, the individual points on each spline wouldn’t come into play that much. I was also eager to make the model! Still, clean work at any stage is a good habit to get into. I just had to do a bit more point welding at the end.

AFter the model is pathced and cleaned up, you can assign the UV’s (in your 3D program or ZB) and texture it, or sculpt it further in ZB and so on.

#7

Very interesting work, Ken !

Didn’t you use the Auto Patcher MK at any stage ?

ivan

#8

Found this today:
http://www.splinegod.com/onlinetraining.htm

At the end there is a free spline patching tutorial. Should help with this technique.

#9

Excellent and informative as usual Ken :warning:
thanks for sharing :warning: :+1: :+1: :+1: :+1: :+1:

#10

Thanks for the information Ken. I tried to deal with creating a mesh for animation in another package with a different approach. I created the original mesh using Zspheres trying to have a good topology to start with. Then I took this mesh into Wings3d and started deleting and adding edges and vertices while trying to preserve the original concept. I ended up with a mesh that was much easier to manipulate.

#11

I did not use auto patcher. I wanted to have more control over each individual patch

Wings 3D is a great program, and I did do some experimenting in there as well. That method works better if the model you are importing doesn’t have a too high polygon count. (such as a model made with the adaptive skinning process)

#12

Hello Ken,
Firstly I think a thanks is required for the great work you have shown us and the time you have spent writing the useful tutorials on this forum.

I have been using this technique for quite some time now, coming to my attention after using Cyclice. I have not used lightwave before and I don’t know if it allows thesame features ,but cyclice is able to snap perfectly,idependant of the resulution ,the spline guids to the model. Further it is able to create ,nurbs,polyor subd, over the topolgy and further snap the hulls of the patches to the geoemtry.This seems to work extremly well and is very fast as it has the option to transfer and adapt thesame topology for different models of similar structure.

I have my experiment here , but the final piece is not quite finished.

thanks once again

take care

Kiril

#13

It looks just like the technique uesd to create Woody in Toy Story! Except there’s no messy clay! :wink:
That Cyslice looks good but it’s hardly a poor mans program!

#14

Excellente method Ken B.
this is the Cyslice of poor :slight_smile:

Thx for that experiment!

#15

This is a very interesting technique. Especially if you are wanting to create full body models. It seems to be much easier to sculpt your model from zspheres, then later create a low poly wireframe of that same model in another package, instead of trying to figure out how to do that before hand. However, you will loose certain details, because it would be difficult to map exactly how your Zbrush model looks from a different wireframe.

A few problems that I can think of from this is:

  1. You will be doing everything twice. Although this isn’t necessarily a bad thing, because practice makes perfect, but in a production environment, time is always a problem.

  2. Once you have the new model made, you will have to go back in and create the details all over again.

But, in retrospect, I can see a bunch of benefits to this too:

  1. Using the ease of Zspheres, you can create anything you want without the need to worry about wireframe layouts.

  2. Once you have the low poly model made, you will have a much cleaner model to bring back into Zbrush and add detail.

  3. You can UV map the low poly version much easier than the high poly version. This is a benefit only if you need to use UV layouts other than AUVtiles.

  4. This technique could work great for games, especially with the new normal mapping features.

  5. Once you have your low poly object, you can use displacement maps in your final renders, and you won’t have to use the high polygon models to animate with.

I’ll have to keep this in mind when I work on my next project. I can see this technique really improving my modeling process.

Thanks for sharing Ken.

#16

cheers ken :+1: this will be very useful. i will have to pull out some older sphere 3d model and try the conversion on them. thanks alot :slight_smile:

#17

hmmm clever method

it is extra work as wchamlet said, but that’s almost always the case anyway if you want reliable great results… lots of extra work need be done by hand