How would I go about applying a bump map towards the normalmap calculations? Is this even possible in Zbrush? I’m thinking of very fine detail which can’t be represented on the model unless it its (too) highly subdivided, but which I want to show up on the normalmap (skin is a good example).
I can think of workarounds like adding the detail to the displacement map (does Zbrush use the dispmap resolution when I create the normalmap, or is it looking at the mesh resolution?). But maybe there’s an easier way? Programs like Melody allow the user to add a bump during calculation, can Zbrush do something similar?
You could try using Nvidia’s normal map plugin for Photoshop and see if you can use that to blend a normal map plus a bump map. I’ve seen other people do that, and it came out nice.
Here’s a link to a fellow who did this, with excellent results.
http://www.cgchat.com/forum/showthread.php?s=&threadid=19350
He talks about using the plugin on page 5. And the you can see the final result here:
I forgot the link to the plugin.
http://developer.nvidia.com/object/nv_texture_tools.html
It only works on PC though.
Thanks. The normalmap plugin as a postprocess is one of the workarounds I’ve been considering, but its results aren’t as good as a bumpmap that is mapped to the object, THEN used in the normalmap calculations. If you overlay a planar map on top of a model with actual UVs (rotations etc.) the angles represented on the planar map don’t match up with the rest of the normalmap. If you map it to the model first all the directions in the normalmap will be correct.
For very fine detail the errors aren’t visible, but for bigger features they become apparent, making the model “not look quite right”.
I’m not sure I understand why you want to do this so this may not help much lol!
I assume you have a low res model, a medium/high res version of it that you’ve used to create the normal map, and a bump map?
Why not apply the bump map as a displacement map to the high res mesh in ZBrush and create a new normal map from that? _ Baz
Because the bump map might have such fine detail that the model won’t pick it up 100% when it’s applied as displacement (even Zbrush has its limits when it comes to subdividing models). That’s what my question in the initial post was aiming at: when creating the normalmap, does Zbrush look at the resolution of the mesh or the displacement map? If it looks at the dispmap I don’t have to worry about the mesh resolution, because all detail present on the map will be considered for the normalmap calculations. But if it only looks at the mesh I’m bound by the subdivision granularity, which is probably too low for what my company is presently doing.
OK, I understand now, sorry I can’t help! 
ZBrush looks at the actual high resolution mesh.
If you can’t get the mesh high enough in resolution to support the level of detail that the bump map is used for, try splitting your mesh into two or more parts:
- Hide half the model.
- Clone the model.
- Delete the hidden points.
- On the clone, invert visibility and then delete the hidden points.
- For each half, store a morph target so that you can use it when you come back to subdivision level 1 later to generate the normal maps.
You should now be able to divide the mesh halves to a point where the displacements generated by the bump map can be applied to the mesh. You’ll then be able to go to level 1, switch to the morph target, and create your normal maps. These maps can then be combined in Photoshop to create a single map for your entire model.
Thanks, I’ll give that worlflow a try.