ZBrushCentral

Recommended amount of active points for uv export

ZBrush Usage Questions
#1

Ok, so I’ve had two unanswered questions, let’s go for a third! lol.

I have a model (my first ever model) that is giving me grief when trying to unwrap it.

At it’s lowest subdivs, it’s 40k points, 4 subdiv levels later (at the maximum) it’s about 2 million.

Is the mesh too complex at 40k? It does look fairly dense and causes a few crashes when trying to copy and paste UVs.

Please please someone talk to me, this is incredibly hard to learn.

Thanks.

#2

You might aim for less than that but 40K for a base mesh shouldn’t cause too many problems. But it depends a bit on your system and what it can handle.

How did you create the base mesh? And what do you intend to do with it?

#3

Hi, thanks for the reply.

I created it, i think, firstly with dynamesh and then subdivided it. It’s for use with a still render inside 3ds max (see detailed sculpt below).

I actually managed to unwrap it, although it seems a bit dense.

Here is a screen grab:

flatten.jpg

I’m just wondering if, in future, it’s wise to start with a very low level? or what the benefits/differences are.

Here is the scultp at full subdiv:

knocker05.jpg

Cheers :slight_smile:

#4

Great sculpt. :slight_smile:

If you plan to export a model from ZBrush then it’s a good idea to plan a little bit so that you can meet the constraints of the other program. For example, if you are using DynaMesh to create your sculpt then make sure you are using a fairly low resolution. DynaMesh is intended as a way of roughing out your initial concept, not for producing much in the way of detail. For detailing, you would add subdivision levels. That way you will have a low res base mesh that you can easily UV map and export with displacement maps etc.

That’s not the only option though. ZBrush provides a number of working methods that can help get your model into another program. For example, you can create completely new topology for the base mesh using one of the re-topology methods and then project over the detail from the original sculpt. In your case, as you do not want to animate the mesh, the exact topology doesn’t matter. Decimation Master provides a good way of creating a new base mesh that can be easily UVed. Briefly the method would be:

  1. Go to the lowest subdivision level of your detailed sculpt (if it has one).
  2. Press the Tool>Make PolyMesh3D button. What this does is make a copy of the mesh without its higher subdivision levels. (The Tool>Clone button would include the levels.)
  3. With this new mesh selected, use Decimation Master to reduce the poly count to something suitable for a base mesh. This will depend on the mesh - you don’t want it to end up looking like an amorphous blob!
  4. Assign UVs to the decimated mesh using UV Master or the Tool>UV Map options.
  5. Subdivide the decimated mesh so that it has roughly the same number of polygons as the highest subdivision level of you original sculpt. You need the polygons to make sure all of the detail is transferred.
  6. Append the decimated mesh to your original detailed mesh.
  7. Use the Project All function to transfer the detail from the original mesh to the new mesh.
  8. Generate the maps you require and export from ZBrush.

You’ll need to read up on those various things but the method should work well.

HTH,

#5

Wow, that’s an awesome point.

It’s great coz it reaffirms my recent pondering that I should be working from a lower-res base mesh. I started out with a high ploy dynamesh and then went further with subdividing. After that, when unwrapping it, I thought, hmm, surely this is too dense?

I think I saw a tutorial on your process so it makes basic sense - as long as the detail really does transfer over.

Right now my issue is getting into GoZ or at least learning how to get:

low poly into 3ds max
correct exported textures into vray
replicating the detail I get in zbrush into my vray render

Currently I just see the low res mesh, poor detail, some slight bloat from the displacement… and that’s it. But once I figure out where I’ve been going wrong with a) the export options and b) the vray material options, I’ll be away.

Thanks for the reply!

#6

You may find this video series useful. See the bottom of the page, ‘building game resolution mesh’:
http://www.pixologic.com/zclassroom/homeroom/lesson/military-character-assets-with-joseph-drust/#building-game-resolution-mesh-part-1

#7

Thanks, I’ll check those out. Although I won’t be exporting to game mesh, but it’s probably the principle that counts.

But I’m starting to have my doubts about the limitations of zbrush.

Here is a comparison I did of the zbrush render and surface qualtiy, versus the export and subsequent 3ds max render:

Comparison: on the left is the low res mesh by itself, on its right is the low res mesh + normal map. Then there’s an image of the zbrush model. The finish is far superior in the zbrush render and the 3ds max render just has that rounded, stretched pixels kind of feel about it.

Is this kind of comparison standard or should I expect better results?

The normal map was 2048px. There seems almost no point in using displace as it just slightly bloats the mesh at render time.

Thanks.

#8

I don’t see any images. But a normal map will never look as good as a ZBrush hi res model because it doesn’t alter the geometry. A displacement map, when properly applied, should produce results that are much closer to the ZBrush model.

The map size also makes a lot of difference. You need to be sure that the size you use is sufficient for the detail you are trying to capture. As a rough rule of thumb, a 2048x2048 map can capture a maximum of 4 million polygons but, depending on the type of UV mapping, will actually capture rather less, perhaps only 2 million polygons.