View Full Version : bump map "moire" stuff.. ?
12-05-2006, 09:58 PM
maybe someone has a short and comprehensive answer to this...
I'm dealing lately with some bump mapping issues, and if you'd take a little look at the picture i have attached the problem would be self obvious... why am i getting these "patterned" look bump map? ...
Are these related to the filtering algorithms employed (box 1/1 :).. )? or is it related to the 3d texture setup? I've used box 1/1 with samples min0 max1 because the resolution is somewhat high, and because of that I can't risk to spend the rest of my natural life waiting for this render to finish by making use of more "sophisticated" filter types...
thank you for your thoughts and ideas
all the best
Some 3d textures within Maya are not "looping" good when the place3dTexture is small in scene; in fact they are not supposed to loop at all but they do, maybe because they are based on pseudo-random numbers to fake chaos. What I do in this case is either:
- if the map has a "depth" or "octaves" or "lacunarity" attribute, I set it to compute high frequency details so I can scale the place3dTexture up, pushing the pattern "away" from my object,
- or I use another texture's alpha plugged into one of the guilty's attribute in order to perturb it's base pattern. This one works good and allow creation of even more sophisticated textures.
Hope it helps !
12-05-2006, 10:20 PM
you've pointed out an interesting approach... thank you :)
I shall give it a try if i'll get a chance, and surely I'll report back
12-07-2006, 06:55 PM
good evening everyone...
Exo7 - i actually solved it pretty simple by using higher values in the randomness parameter of the texture... though, i'm not sure you can actually control the dispersion of the 'cells' for example in the textures with the alpha channel of another... my previous experiences with this tehnique lead to distortions in the final map...
I have another problem though: how can I keep render times low when using bump maps with mental ray?
My render times actually have increased from 1h:30m to 9h !!!
If that's not by overriding the Sample Limits in the object's shape, then the topic is too advanced for me I'm affraid :) Maybe just try Higher values in the contrast threshold of the AA, for instance if you don't plan to generate an alpha you can leave its value at 1.
12-08-2006, 02:23 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.