Hi,
Happy holidays…
I’m having troubles bringing Zbrush 3.5r3 sculpts via displacement into 3ds max 2009, and I’m wondering why there still isn’t a fool-proof, more or less official way to do so. I mean, this pipeline, Zbrush -> 3ds max, should be one of the most frequently used, right?
Anyway, I browsed the forums intensively and also looked at a Gnomonology tutorial which is, unfortunately for me, aimed at Maya. The cleanest solution seems to be 32 bit displacement maps, but I can’t replicate those Alpha Gain and Alpha offset values in max 2009.
So, first question: How can I replicate that fool-proof setting from Maya?
Falling back to 16 bit maps, there are two problems:
1)
The displacement amount is guesswork. I’m using ADF * mesh scale which seems to be in the right neighborhood, but not good enough for the sofa I need to visualize.
Something doesn’t fit to my model in max, I assume it’s because TurboSmooth subdivides with another “curvature” than Zbrush. This is most notable at softly rounded edges.
In the Gnomon tutorial, subdivision is solely applied at render time by mental ray. This seems to be the only correct fit, since the displacement map is rendered off the faceted base mesh in Zbrush. It seems logical that TurboSmoothing that base is a distortion. Not TurboSmoothing it, but only subdividing via MR doesn’t give me the result shown in the Gnomonology tut.
If I use level 2 to base off the displacement, I can reduce the deviation in max, but not eliminate it.
So, second question: What’s the correct way to render and apply 16 bit displacement maps from Zbrush to max?
Here are the threads I’ve looked at (and tried out, of course):
http://www.zbrushcentral.com/showthread.php?p=471915
http://www.zbrushcentral.com/showthread.php?t=20310&page=32&pp=15
http://www.zbrushcentral.com/showthread.php?t=32020&page=3&pp=15&highlight=exr+max
http://www.zbrushcentral.com/showthread.php?t=68023
http://www.neilblevins.com/cg_education/mudbox_2009_to_max/mudbox_2009_to_max.htm#DisplaceModifier
About a year ago, I myself posted a workflow that worked sufficiently, but it turns out to produce small inaccuracies and other mesh artifacts, which is absolute not acceptable for my commercial project at hand.
edit:
In case you’re wondering why I don’t simply use an optimized mesh (e.g. by Decimation Master) to render: The sofa is flexible and I need to animate the angle-dependent folds and wrinkles around the joints.
So, please share your experience and solutions. I’d be most thankful for any constructive inputs. (scanline or mental ray, latter preferred)
