CompuServe Thread

#R3 Spotlights

15 messages in this thread
#110142From: Paul LindJun 28, 1994 7:23 PM
I'd like to know what method of decay spotlights in 3DS R3 use. Do they: a.) have no decay b.) decrease in linear proportion to distance c.) decrease inversely proportionally to the square of the distance d.) another method Thanks A Ton, Paul
#110154From: Phillip L. MillerJun 28, 1994 9:01 PM
Spotlights (and omni lights) have no decay unless their Attunate attribute is on (in which case the spotlight will have linear falloff between the inner and outer range circles). – Phil
#110157From: Paul LindJun 28, 1994 9:11 PM
Phillip, Is that somewhere in the manual? I was curious because we're trying to match the lighting in R3 with lights in a film piece and another unmentionable program and after a very long discussion on light here, it seemed strange that the spotlights would not use an inverse of the square formula like spots do in life. Quite frankly, until now I've never had a need to concern myself with it. Paul
#110223From: M. G. BATCHELORJun 29, 1994 2:13 AM
Paul, << it seemed strange that the spotlights would not use an inverse of the square formula like spots do in life. >> I believe most 3D software does not use the inverse square method simply because it just doesn't "look" right, FWIW. BILL
#110224From: Paul LindJun 29, 1994 2:23 AM
Bill, Actually several packages use the technique as a default. Most high end ray tracers do. We happen to prefer the 3DS method, hence the questions. It's a huge pain in the butt trying to get other packages to respond like 3DS. Paul
#110230From: M. G. BATCHELORJun 29, 1994 2:59 AM
Actually, I was referring to some higher-priced renderers as well. They call it inverse square, but I *_believe_* they actually use the un-squared distance parameter. Seems like I've read that 'way back – I could be wrong. BILL
#110304From: Paul LindJun 29, 1994 10:56 AM
I was just repeating what I was told, two different SGI based companies sent me tech sheets with explanations of their spotlights, and they both are set up so that you can set decay to "decrease light intensity inversely proportional to the square of the distance." They also sent us example calculations to accompany the brief. As a note you can also set the spots to higher or lower ranges of decay by changing some parameters. The main options are, no decay, linear decay, inversely proportional to the square, inversely proportional to the cube. None of it makes alot of difference to us, we were just in need of matching some 3DS lit scenes to work in some other packages and we were having alot of trouble matching lighting. Hence the questions.
#110166From: John N. JordanJun 28, 1994 9:50 PM
>spotlights, decay Look at 7-158, and pages around it in the reference manual. Just looking up spotlights myself.. Look especially at attenuate and the stuff referred to in that paragraph JNJ
#110177From: Yost GroupJun 28, 1994 10:59 PM
You seem to be asking about penumbral decay of the spotlight, not attenuation over distance for spot and omni lights. Given that, are you sure you _really_ want to know? <g> Basically, spotlight decay uses a cubic spline with hermite interpolation that's shaped like an s, so that the spline is tangent to 1 and 0 (it has 0 first and second derivatives at the ends). The formula is: u*u*(3-2*u) …where u is a number which goes from 1-0 as you go from the hotspot to the falloff. (0 at the outer range and 1 at the inner range) That's for circular spotlights. The rectangular ones are a hybrid variant that's way too complex to explain in an email. (does anyone else do rectangular lights?) – G
#110181From: Jeff L. BlackwellJun 28, 1994 11:11 PM
Hey, I have no idea what formulea are involved, but I love rectangular spots. Jeff
#110202From: Yost GroupJun 29, 1994 12:45 AM
I'm glad to see that they're appreciated. They were a bear to implement. – G
#110985From: Andrew McClaryJul 3, 1994 10:21 AM
Gary, Your rectacular spot came in real real handy for re-creating the glow from some lines of neon on the outside of a building we recently did for a high profile client. (With some luck they will be introducing thier stuff to the public before SIGGRAPH so I can show some of the animations we did for them there.) I also want to thank you and sing praises for the projection lights. I haven't had time to get on the forum for a while. Is there going to be any meetings or get together of the 3DS users in Orlando during SIGGRAPH? Andrew McClary Graphic Alchemy Productions, Inc.
#110991From: Jonas Ruikis [ADESK]Jul 3, 1994 11:21 AM
Andrew, << Will there be any meetings or get together of the 3DS users in Orlando during SIGGRAPH? >> Autodesk is sponcering a party Wednesday nite at the Hilton Hotel, Walt Disney World. All registered 3DS owners will be receiving invites in the mail. Last year there was also a ASOFT'rs get together but I haven't heard anyone organizing ont this year. jonas[adesk]
#110997From: Yost GroupJul 3, 1994 12:48 PM
As Jonas said, Wed evening 7-10pm. C'mon down! <g> Glad to hear that you're grooving on the new light features. – G
#110949From: Bruce McGlassonJul 2, 1994 9:33 PM
Hmmm… lessee now… penumbral decay… a cubic spline with hermite interpolation… first and second derivatives at the ends… unh, hunh, unh, hunh… u*u*(3-2*u) … (I should've taken a little more arithmetic at college)… yup, I like them rectangular spots, too… a hybrid variant, eh? <blink, blink….>