"non-linear" displacement maps

General discussion about Indigo Materials - material requests, material developement, feedback, etc..
Post Reply
9 posts • Page 1 of 1
User avatar
Frutiger
Posts: 122
Joined: Wed May 05, 2010 4:05 am
Location: Northern California

"non-linear" displacement maps

Post by Frutiger » Tue Nov 16, 2010 6:35 am

So here's the deal. I'm trying to create a method for creating a smooth curve for a displacement map. For example, if one was trying to create a corrugated metal siding texture or a cap molding for tiles like the attached file.
36VC00.gif
Desired Effect.
36VC00.gif (21.62 KiB) Viewed 3987 times
Creating linear gradients in photoshop doesn't quite cut it.

Here's as close as I've been able to get using photoshop.
CapDisplacement.jpg
Displacement Map for Cap Molding
And here's the rendered results. Ignore the ugly grout lines, that's just me being sloppy while I'm trying to figure this out.
Cap Molding Test.jpg
Cap Molding Displacement Test FAIL
As you can see, even when smoothed out it's not quite right. Ideally I'd be able to create rounded "parabolic" gradients instead of smoothed linear ones, but I haven't figured out how yet. Any ideas?

User avatar
Frutiger
Posts: 122
Joined: Wed May 05, 2010 4:05 am
Location: Northern California

Re: "non-linear" displacement maps

Post by Frutiger » Tue Nov 16, 2010 6:40 am

And if someone tells me to just use ISL, I'll come to your house and punch you.

StompinTom
Indigo 100
Posts: 1828
Joined: Mon Sep 04, 2006 3:33 pm

Re: "non-linear" displacement maps

Post by StompinTom » Tue Nov 16, 2010 7:43 am

Just use IS... um.

Make a linear gradient, then apply a different gamma curve to it. In theory, as far as I can tell, the shape of the gamma curve should correspond to the shape of the curve.

Photoshop -> Curves

This occurred to me one time when I was after something similar though I don't think I ever really seriously tested it. I'd be interested to see how it turns out, though...

User avatar
dag
Posts: 189
Joined: Tue Feb 02, 2010 2:28 am

Re: "non-linear" displacement maps

Post by dag » Tue Nov 16, 2010 8:12 am

The time it takes to create a map like that, you would be finished modelling it instead, and for displacement you need high geometry so it's not really worth it.

Black-white-black gradient, then fiddle around with curves and levels.
Attachments
d.png
r.jpg

User avatar
snorky
Indigo 100
Posts: 331
Joined: Fri Apr 17, 2009 8:54 pm
Location: Parma, Italy

Re: "non-linear" displacement maps

Post by snorky » Tue Nov 16, 2010 8:30 pm

tile:
tile.png
tile.png (9.09 KiB) Viewed 3938 times
roof:
roof.jpg

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Re: "non-linear" displacement maps

Post by Zom-B » Tue Nov 16, 2010 11:41 pm

Subdiv + smoothing is still broken since some time :(
It can create some strange UVM problems that also messes with the displacement.

Glare told that they'll have a look at it just recently, so we might get lucky
for some of the next releases :)
polygonmanufaktur.de

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: "non-linear" displacement maps

Post by CTZn » Wed Nov 17, 2010 8:22 am

Beside issues that were not demonstrated in this thread, meshes subdivision in Indigo comes with various options wich are managed differently depending on the exporter you use.

At some point it may become necessary to hand edit the <mesh> definition in the IGS. Options such as:

Code: Select all

	<normal_smoothing>true</normal_smoothing>
	<merge_vertices_with_same_pos_and_normal>true</merge_vertices_with_same_pos_and_normal>
	<wrap_u>true</wrap_u>
	<wrap_v>true</wrap_v>
	<max_num_subdivisions>10</max_num_subdivisions>
	<subdivision_smoothing>false</subdivision_smoothing>
	<displacement_error_threshold>0</displacement_error_threshold>
	<view_dependent_subdivision>false</view_dependent_subdivision>
to name them, have a great impact on the result.

Could you post the IGS mesh options for the tile ? (not the geometric data if present please).

PS: I'm seconding the tip on curves (aka texture gamma, aka texture exponent). Your issue may come from displacement_error_threshold being positive, perhaps also curvature dependent subdivisions. Set both to -1 if present.
obsolete asset

User avatar
Frutiger
Posts: 122
Joined: Wed May 05, 2010 4:05 am
Location: Northern California

Re: "non-linear" displacement maps

Post by Frutiger » Wed Nov 17, 2010 4:29 pm

I think Dag spaketh verily. Just modeling the tiles out took less time, but was a little less elegant of a solution.

Here's a little context of the application that prompted this thread.
TileContext.jpg
Context
It'd be nice to just have a good texture to throw on there seeing as putting all those in was a little tedious.

Are you guys saying that even if it was executed correctly, one of these curved displacement maps would still be non-functional?

User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

Re: "non-linear" displacement maps

Post by galinette » Fri Nov 19, 2010 9:37 pm

I Fully agree that modelling this instead of having it done with displacement will make it more precise and cleaner.
If you absolutely want to do it through displacement, it is better not to use subdivision. Make it enough subdivided in the modeller instead. Subdivision is not very accurate for regular precise patterns.
Eclat-Digital Research
http://www.eclat-digital.com

Post Reply
9 posts • Page 1 of 1

Who is online

Users browsing this forum: No registered users and 17 guests