CompuServe Thread

Global Ripple

22 messages in this thread
#37428From: Margaret CopelandMar 27, 1992 6:00 PM
Ok, In Steve Worley's Imagine book he makes the comment that one can "animate" a global picture. This puzzles me. Does this mean that I need to load a new global picture for each rendering in an animation? Apparently fx effects do not work on grounds – is there a creative way around this problem. I'd like to be able to ripple etc. the ground. Any suggestions?
#37489From: Charles BlaquiereMar 29, 1992 5:21 PM
Margaret, if F/Xs don't work on a ground, how about creating a huge, two-triangle plane? Blaq!
#37509From: Margaret CopelandMar 30, 1992 11:03 AM
Will that change the way the spots look?
#37519From: Charles BlaquiereMar 30, 1992 6:16 PM
Margaret, try rendering the same plane with the texture axes rotated. The spots will certainly look different, I'm sure. Now if our friend added a texture to an object, the new default axis placement may have gone unnoticed; if, on the other hand, he simply re-used the same objects without modification, then I'm clueless as to why the spots aren't the same size. Blaq!
#37524From: Margaret CopelandMar 30, 1992 7:59 PM
Obviously the only way to check this out is to try it. Still some of these processes take so long to "check" that I'll take some advice. I think I'll work on ripple first instead of memory hog wave. However, since ripple displaces the surface, a two triangle plane would not ripple much.
#37817From: Joseph C. PizzicaroliApr 5, 1992 7:55 PM
Margaret – The ripple effect is not constrained by the polygon count of the object being rippled. I've gotten great looking wave propagation across a 2-triangle plane (20-30 waves from edge to edge). Just give Imagine a wavelength and it'll do its thing – it is perfectly happy to create surface perturbations inbetween polygon edges. – Joe
#37822From: Margaret CopelandApr 5, 1992 10:02 PM
Oh great. I've been experimenting with a 20X20 polygon plane. I've tried various travel distances which has interesting possibilities.
#38029From: Joseph C. PizzicaroliApr 11, 1992 10:46 PM
Margaret, Yeah, there's no need to use more than a 2-triangle plane unless of course there are specific details or feaures you need to incorporate. I've done wave anims where I set the travel distance to be exactly 1 (or a multiple of 1) wavelength, so that I can loop it. This can be extended – I did one where the scene was static (except for waves rolling along), then objects rise out of the water (great reflection effects!), then remain static with new objects visible and new reflections in rolling waves – all timed using multiples of wavelengths so that each segment is seamless to the next. Really nice effect, but VERY RAM hungry. – Joe
#38044From: Margaret CopelandApr 12, 1992 6:34 PM
This I will have to play with more. I want to start experimenting with global pictures and using 2-D morphs from Image Professional. It is too bad it takes so long to test out some of these effects.
#38614From: Joseph C. PizzicaroliApr 27, 1992 9:20 PM
I sadly agree. There are many, many ideas I would like to try out, but the realities of rendering times (even on a 33 MHz '030 machine) dictate that many, many of them get left by the wayside. I remember after first getting Imagine V0.9 I needed to figure out just how the disturbed texture worked and how each parameter affected the results. The manual was no help, and dear old Mike H. at Impulse said "just experiment, then you'll understand." Yes, many hours and dozens of traces later I understood exactly what a half-page of clear instructions would have revealed. The Imagine universe is so vast that I am grateful for the efforts of those like Steve Worley who are sharing their hard-won knowledge so that the rest of us can save some precious time and put the power to work for us. Good luck with your explorings. Let us know what you find. – Joe
#37570From: Charles BlaquiereMar 31, 1992 10:39 PM
Margaret, oops — I guess I had my threads crossed. You were the one complaining about F/X and grounds, while someone else had problems with dot size. Hmmm. Now I'm more confused than ever. Blaq!
#37591From: Margaret CopelandApr 1, 1992 3:10 PM
It's all interrelated.
#37550From: Peter GreciMar 31, 1992 12:26 PM
FX DO work on the Ground I've gotten Rippling Water using the Ground Primitive ya simply need 2 copies of the ground with different FX settings and Morph and Voila'
#37560From: Margaret CopelandMar 31, 1992 6:11 PM
Maybe I'm a little dense here but you say two copies of the ground with diff FX settings? So you are applying a texture to ground #1 and then the same texture to ground #2 and then morphing…. or are you using FX settings in the Action editor?
#37738From: Peter GreciApr 4, 1992 3:06 AM
using FX in action Editor and morphing or ya can use Textures with 2 different Settings 1 for each model version and morph to attain nice rippling water
#37498From: Joseph C. PizzicaroliMar 29, 1992 7:06 PM
Margaret – I don't know about grounds, but I have had great success applying the waves texture (Imagine 2.0) to a plane primitive. You just have to make sure that the boundaries of the plane are outside your field of view. The result was truly spectacular, but incredibly memory hungry (12 MB of RAM for a 120 frame HAM sequence). – Joe
#37510From: Margaret CopelandMar 30, 1992 11:08 AM
Well one thing the Imagine 2.0 manual has that Worley's book does not is a ballpark estimate of ram hogginess. Apparently "wave" eats your memory. Unfortunately neither manual talks about FX memory requirements unless I missed it. I'll try both wave and the ripple FX and see if I have enough ram to do it. 12megs eh?
#37818From: Joseph C. PizzicaroliApr 5, 1992 8:02 PM
I don't have a feeling for how much RAM the wave texture eats while actually rendering. The reason for the finished animation taking 12 MB is mostly a function of the large number of pixels (almost all of them) changing from frame to frame, so that Delta compression is essentially worthless. – Joe
#37823From: Margaret CopelandApr 5, 1992 10:03 PM
What size rendering were you working with?
#38030From: Joseph C. PizzicaroliApr 11, 1992 10:54 PM
Margaret – As I recall, rendering was HAM overscan at 352 X 440 using Imagine. The problem is simply that when looking at rippling water, particularly if the lighting and reflections are significant (full ray trace), then the odds are that a very high percentage of pixels is going to change from frame to frame. Presto, delta compression is highly ineffective (and can even result in delta files LARGER than the IFF frame files themselves) and the anim becomes huge. I should point out that in the stuff I rendered, the rippling water was framed so as to completely fill the screen (looking sorta down into it). If, on the other hand, your scene had water in only a portion of it, and something else like sky in the rest, then the size of the anim will be reduced accordingly. Still, in general I would say it's a rather expensive effect in terms of RAM consumption, but it's really a nice one, too. – Joe
#38045From: Margaret CopelandApr 12, 1992 6:38 PM
I've pretty much given up on full ray traces unless the scene really warrants it. You can get a lot of interesting pictures without it. I did do one that I really like, a 320X400 that is in Amigaarts called Finnies.jpg < I think I jpegged it. Anyway the trace created reflections that I never would have expected and I have to look at it a long time to figure out what it really is reflecting. As I recall it "cooked" all day.
#38615From: Joseph C. PizzicaroliApr 27, 1992 9:25 PM
I agree – I dread the effect of tracing on rendering time. And for some applications it is indeed overkill. But for some of the stuff I'm working on, the full impact of the anim just won't come across without it. I am however interested in experimenting with reflection mapping. I've never tried it, but from what I understand it just might convincingly fake it in some cases. So little time to experiment…but here's a case where a little experimentation might reap great dividends in reducing rendering times. – Joe \exit