ominousplay opened this issue on Jul 14, 2006 · 17 posts
ominousplay posted Fri, 14 July 2006 at 3:37 AM
Never Give Up!
ominousplay posted Fri, 14 July 2006 at 3:40 AM
I found one of my answers. You can tell particles to not calculate - check off. And now it no longer drops my changes to the spline object. This still needs a lot of work. If I was a bird, I wouldn't get close to this monster fountain.
Never Give Up!
danamo posted Fri, 14 July 2006 at 11:01 AM
Thanks,that's good to know. I'm building a steam locomotive for a Carrara animation and on top of boning and rigging the valve gear(maybe I'll have to buy the "Cognito" plug) I want to have copious amounts of smoke trailing behind as it goes down the track. I'm just starting to get into particles and I love their potential.
ominousplay posted Fri, 14 July 2006 at 12:23 PM
Happy to share. I ended up adding water to the fountain instead of trying to fill it with particles. It would be nice to have the "water" in the fountain react to the particles - better for animation purposes. Post your work on the steam locomotive if possible. I've had my eye on cognito as well - looks like it would be great for machines.
Never Give Up!
danamo posted Fri, 14 July 2006 at 4:50 PM
It would be great to be able to have the "water" react to the particles. Hopefully fluid dynamics will be added in a future ver. of Carrara. I have had some luck using an AVI file to drive displacement of a mesh in a water animation. Any chance that something similar might work for your animation? I'm almost finished building my loco, so I plan to download and purchase Cognito tonight . I'll let you know how that goes and I'll certainly post a link to the animation. Thanks again.
ominousplay posted Fri, 14 July 2006 at 5:29 PM
Never Give Up!
ominousplay posted Fri, 14 July 2006 at 7:27 PM
Never Give Up!
danamo posted Fri, 14 July 2006 at 10:37 PM
The fountain spray looks very good, even without any post. I have no bias against postwork,LOL. Whatever it takes to make a great picture. I really like the pirate render! I've always gotten a kick out of High seas adventure and pirate flicks. I haven't tried importing a Poser scene yet, but this looks really good so I'll have to try it.
bluetone posted Tue, 01 August 2006 at 8:50 AM
If you have C5, then there is an option in the particle modeling room for setting the particles to start BEFORE your timeline. So, if you want to fill the fountain, set it to start earlier, then you can leave your timeline set at Frame '0' and not worry about accidentily animating an object by adjusting while at the wrong frame. (A problem I have struggled with more then once. :( I was SO glad when they added a Frame '0' to be able to correct for such Operator Errors!)
nomuse posted Tue, 01 August 2006 at 1:06 PM
Two things (learned after much experimentation, confirmed by tech support and other users); Particles never save their simulation. You can not "scrub" through a particle simulation, or go to a particular frame to save sime; in all cases the particles will run the complete simulation up to the present frame. Particles "leak" from solids. Apparently their surface detection is not that accurate. You can not currently fill a bowl from a particle fountain (well, you can, but the particles will be constantly leaking out the bottom and need to be replenished.) According to an Eovia tech this was considered a bug and they were working on it. I, too, was hoping to do some interesting things with particle pools -- creating ripples by dropping objects into the pool, for one thing. At the current state of the program that does not work very well. On the other hand, metaball particles work really nicely for water trickling down a rocky slope...
Ringo posted Tue, 08 August 2006 at 3:46 PM
The particle engine is NOT a fluid or softbody engine do not get the mix they do not work the same way.
You cannot create ripples just by dropping objects into a pool. The pool water would have to be a softbody dynamic and IT ISN"T at this time. Learn the features and limitation of the particle engine and read the manual it helps.
nomuse posted Tue, 08 August 2006 at 4:23 PM
Write a manual and I'll read it. What you have doesn't give ANY of the assumptions in the particle engine, the render engine, fresnel terms...shall I go on? All of the underlying algorithms have to be worked out through experimentation, like old-time atomic structure experimenters firing electrons in and seeing which direction they bounce. The "manual" is simply a laundry-list of the names of buttons, a slightly tutorial-like slant where you are told which button to press for a specific effect....oh, and maybe 30% by weight cross-references (including cross references that point from the bottom of a paragraph back to the top!) ((If I thought anyone cared I'd send in a page of typos, misspellings, and broken cross-references.)) Oddly enough, you CAN create something a lot like a ripple. I did it and made an animation. Actually, it was a bit more like a lumpy splash, but the idea was there. All you have to do is add the object you are dropping to the collisions. My render was not high enough granularity, however, to see if I was actually pushing particles, or if my "dropping" object was acting as a deflector to the still-operating particle fountain. And thus I have a very specific question; Do particles self-collide, and if so, what determines the collision radius?
nomuse posted Tue, 08 August 2006 at 4:29 PM
Re; page 118 of the C5 online help pdf; "Particles Push Each Other makes the particles that are too close to push each other and..."
nruddock posted Tue, 08 August 2006 at 5:05 PM
You need a more complex interaction not just (elastic) colisions between particles if your going to be able to simulate a liquid (with a free surface).
The "Equation of state" that these particles follow makes them most like a ideal gas.
nomuse posted Tue, 08 August 2006 at 5:18 PM
...Which for some cases is close enough. A single "ripple" can be simulated as an expanding circle of displaced material. Obviously you can't get the damping oscillation of a true surface out of that! A splash, also, can be as a first order approximated as a fountain aiming upward and outwards. If you can manage collisions so particles are sent upwards and outwards, do they not resemble a splash? Remember, metaball surfacing is allowed here, meaning you can smooth a "skin" over any group of particles not spread too thinly. Obviously you are no more simulating a liquid phase than you are a solid phase. But 3d is all about fakery, neh? At this point in time, which would be the best option for a Carrara animator wanting to do a simple droplet effect? A) Dump Carrara and buy Lightwave. B) Put the project on hol and wait in hopes that the next version will have the function you want. C) Paint in everything by hand frame by frame. or D) Fake up the best approximation you can with animated metaballs, particle effects, transmaps, formulae objects, and anything else that you can dream up.
nruddock posted Tue, 08 August 2006 at 5:41 PM
E) Do the fluid simulation in Blender, and import a series of objects that get shown one per frame.
nomuse posted Tue, 08 August 2006 at 6:17 PM
Heh. Pretty much the same thing as going to another software. Not in itself a bad solution! (I still use UVmapper and BBedit to tidy up .obj files I export from Carrara)... Heck, I could run a ripple one frame at a time on an old Mac plug-in I had for Bryce, convert that to a displacement map... The question is...is it worth trying to fake it in Carrara? Especially since the manual is a little unclear about what controls behavior on particles, at least two people thought it was worth at least giving it a try and seeing what actually happens.