CompuServe Thread

#Bump map render glitch

17 messages in this thread
#115154From: Richard CianoJul 29, 1994 10:45 AM
I have created an Asphalt material, default with an image of road aspalt as texture and bump map entries. I am panning around this road surface, and in certain frames, rendering in any resolution and palette combination, dithering and other settings in any combination produces wild "flashes", jaggy torn blocks of lighter or darker coloration as the camera pans around this "bumpy" surface. I am at a loss: I understand that I might have expected some interference patterns if these bumps were being treated as actual surface features (as with a CD-ROM surface) but these jaggy flashes don't really seem to conform to any pattern, or order, nor do they seem to change consistently as the road and camera change angles with respect to one another: it just seems that certain frames insist on having great blotches of dark or bright. I am at a loss.. if anyone has a thought, I would be delighted to be rescued.
#115191From: M. G. BATCHELORJul 29, 1994 1:02 PM
There really isn't enough information in your description to pin this down, but let's start here: Your reference to pallette implies to me that you're rendering a flic. If so, Are you using LOW pallette mode ? If so, change to MEDIUM and re-render a test. BILL
#115330From: Richard CianoJul 30, 1994 10:42 PM
Thanks for the reply. Sadly, this glitch is occurring when rendering individual GIF files, and in all resolutions and palette configurations.. the individual GIF frames show this glitch. Any further thoughts?
#115369From: James Coulter[Mindscape]Jul 31, 1994 12:52 PM
How does the image look when rendering to a 24 bit targa file? You may be experiencing a side-effect of reducing the data from 24bit to 8bit. — James — MAP —
#115713From: Richard CianoAug 2, 1994 9:52 AM
This "artifact" persists no matter what the output format or other settings. Further, it seems to appear and move rapidly between frames, as the camera pans across the road surface. However, it propogates from the top of the road surface toward the bottom in a raster-like fashion, regardless of the camera's orientation relative to the asphalt map (for me ruling out some sort of virtual interference pattern caused by the pseudo-bumps in the material, which I assume wouldn't happen anyway). At this rate, I will have soon torn out all my hair, and will have to start on my co-workers'. Further, can you please suggest where to browse for IPAS routines of various sorts, or obtain catalogues/promotional materials? I must make a case for the expense to my employer. -R
#115742From: Brian Rudolph [Adesk]Aug 2, 1994 12:12 PM
<<Further, can you please suggest where to browse for IPAS routines of various sorts, or obtain catalogues/promotional materials? I must make a case for the expense to my employer.>> Download IPAS.TXT and MMRG5-.TXT
#115875From: James Coulter[Mindscape]Aug 2, 1994 10:59 PM
Brian has your answer for for the IPAS/Promo stuff. As for your "artifact" problem . . . How big is the surface you are putting this map on? I have occasionally had problems with large planes made from long, skinny polys. For example, a camera placed close to a huge, flat circle with sometimes have problems. If your map is applied to a large object, you might try tessellating it to have more polys. Otherwise, there is a possibility that your 3DS.EXE is corrupted. You might try reinstalling it. Many people backup a clean copy of the 3DS.EXE file to a floppy so that when it gets corrupted they do not have to reinstall the whole thing. Something else just occurred to me. If the surface has any shinyness, you may be seeing the reflection of the light source propogating across the surface as the camera moves. However, I have never seen this cause a raster-like effect (unless anti-aliasing was OFF and pixel size was set to 1.0). It is really hard to "see" what is going on from where I am. <g> How big are the associated files? Too big to upload?? If not, I would be glad to look at it for you. — James — MAP —
#116364From: Richard CianoAug 4, 1994 5:06 PM
The frames of my corrupted animation are only 30K or so apiece, and a very few of them would be required for you to see the effect of which I am speaking. I shall endeavour to send them to you. I think my surface does not have any shine to it; I shall check. I will also try to play with the number of faces: the road is a simple oval track with the road surface itself a single face. Thanks for your robust reply!
#116465From: James Coulter[Mindscape]Aug 5, 1994 1:16 AM
Glad to help. My mailbox door is open . . . — James — MAP —
#116536From: Yost GroupAug 5, 1994 11:02 AM
You can find out if your problem is caused by surface shine by simply moving your lights and re-rendering the problem frames. – Jack
#116542From: Richard CianoAug 5, 1994 11:49 AM
Yes, I have tried experimenting, and still it persists. I will now spend the weekend banging my head on my monitor, to see if that helps g. R
#116693From: James Coulter[Mindscape]Aug 6, 1994 10:43 AM
Can you send me the project file and assocaited meshes via e-mail? Or are they too big?
#116987From: James Coulter[Mindscape]Aug 8, 1994 10:30 AM
Richard, I looked at the PRJ file. So far, I do not see the problem you mentioned, but there could be several reasons why: First, I have only rendered out a couple frames, so the problem may not be apparent on those frames. If you could suggest a frame number to look at I will check it out. Second, I am missing a few texture maps: WINMAP.BMP, SHRUBG.TGA, and STUCCO.TGA. I cannot say whether having these maps would make a difference. Third, you may be using different settings in your 3DS.SET file. Anyway, I wish I could be more help. I will try to render out a FLC file later to see if I can see it there. — James — MAP —
#117026From: Richard CianoAug 8, 1994 1:32 PM
I'm not certain I've ever seen those texture maps.. there should only be one material in the PRJ, Asphalt, which only needs ASPHALT.CEL. I will double check. The effect seems to occur especially toward the end of the first forty frames or so (which I threw in, in the most compact format I could), especially frame 35. Do you see the effect in the rendered frames I provide? Thanks again, btw, for your time. Richard
#115984From: Bruce MeikleAug 3, 1994 9:53 AM
I've had a similar problem just recently. I've just uploaded an example of it called PRBLEM.GIF in the images library, if you're interested (wait for the sysop to post it). I'm wondering if it's related to a glitch with ray-traced shadow casting spotlights which go a bit screwy if they hit an object that is self intersecting. The sea in the image had a bump map to simulate waves which rendered ok in other views but went nuts in this one. The material is on a flat plane but many of the objects I modelled penetrated the plane, rather than sat on top of it. The heavy raster lines were introduced at rendering and I'm wondering if this is related. Sound similar? Any ideas, anyone else? BPM
#116265From: David RhotenAug 4, 1994 11:37 AM
I haven't seen your image, but I had a problem with ray-traced shadows having jagged edges on the ground plane. I tesselated it a bit, so that instead of a couple hugh triangles, it was made of many, and that cleared up the ragged edges. Dave
#116365From: Richard CianoAug 4, 1994 5:08 PM
The waves glitch sounds -eerily- similar. I wonder. -R