PDA

View Full Version : 2 questions of TP fragment


paulselhi
04-06-2005, 06:15 PM
1. If i create a cube as a single particle with enough polys i can use the fragment node to "peel" away the cubes faces by animating the threshold and then use wind to blow them off. I can set up interparticle collision to keep the frags apart but when i set up interparticle collision to keep the frags from passing through what is left of the original particle it will not work

I think i know why..the orignal cube has its center fixed..well at it's center and thus the frags can only have repulse and bounce set to the distance from the whole cubes center, remaining parts are thus not calculated

Is there any way to get the frag particles to bounce of the remaining shell before it turns into frags itself ?

Confused ? err.. let me explain again. A cube is set as a particle and pfragment causes it to fragement from z to -z, to get the frags to come of in a stream (left to right) rather than all at once ( much like pblurp which i don't want to use here) i key frame the threshold value, now i have the frags peeling off but i cannot get the frags to collide with what remains of the cube, they just pass through it

I can of course simply have the speed of the particles set at 0 and remove the threshold keying then i have all my particles born at once and so don't have the issue with passing through the original cube remanents but then i loose that peeling effect, all particles are blown off at once

I have managed a work around using variations in particle mass and the wind set to take that into account..is this the only way ?

Anyone know if it is possible to solve this ?

2. On a similar vein i was unhappy with the way say a glass sphere would shatter on impact with a floor. If the animation was simulating slow motion then the whole sphere particle would explode at once on impact

OK i have worked out how to solve this with the peeling effect ( shockwave effect) of the threshold setting but it must be set to a particular axis. So if i have a rotating particle hitting the floor how would i be able to tell PFragment what alignment it has on impact so that the fragmentation direction is correct ? I don't think the alignment correction option is for this

What we somehow need is for the fragmentation to start at the impact point wherever that may be

Bloody hell know i know why they call it Thinking Particles !! It should carry a health warning " Whilst contemplating TP do not drive, operate heavy machinery of attempt neuro surgery"

Srek
04-06-2005, 08:11 PM
Hi,
you can try to use the original geometry as a collider, but of course collisions will then happen also to the now invisible since fragmented parts. Another option is to use particle / particle collision with a specificaly set distance or alternatively put the source particle and the fragments into different groups and do two particle/particle collisions with different settings.
The second problem is unsolveable currently, as you said it would need the ability to define the point of fragmentation on the particles geometry.
Cheers
Björn

edfenner
04-06-2005, 09:06 PM
Hey Paul,
This sounds like a cool little adventure you are on. I would love to see what you come up with. Sounds like a great bit of thinking will go into these particles!

ed

paulselhi
04-07-2005, 06:32 AM
Here is a quick divx of the idea..it is to be an explosion shockwave..just a bit of fun nothing essential i just am a TP addict, i get an idea then am obssesed with how to do it in TP


You know i am sure maxon put subliminal stuff in C4D it is so addictive !!!

http://www.black-and-white-to-color.com/stuff/tpswfrag.avi

I have 2 groups of particles from the hidden cube ( er..deformed cube !!)

Group is the main building and group 2 is the frags wind effects grp 2 only

There is coliision set between both groups and grp2-grp2

I have a deflector on the floor which is causing the usual problems of particles passing through, see the larger rotating parts.. i would love to know how to solve that but if it too difficult then some pyro dust will cover it

To get the building particle to say put i have set mass for each particle group so the building has a very high mass and so grp 2 bounce off it without moving it

I could try making the invisible building object a defector i suppose but then you would get collisions where there should be none as the structure should have broken away

The idea is to build a little set showing a blast wave travelling through it

Medicine Horse
04-07-2005, 03:44 PM
I think your effect looks great so far and some dust will definately cover up the floor collisions. I didn't even notice them at first. You could try setting the deflector to "geometry" although it maybe be jumpy because the particles are already so close. Better yet render them separately and fake it in post. Nice work.

paulselhi
04-07-2005, 05:34 PM
Yes post would be a good idea.. i think my problem is that i want too much from TP when i am only a total amateur when it comes to CG effects

In the real film world the priority is the shot and not the perfection of the TP or whatever tools used. Most movie cuts are no more than 5 secs and for action often shorter so i should be thinking about solving a lot of problems with "clever" placing of cameras and timing

We tend to be very "microscopic" when viewing our own effects but insitu within a whole movie effect i doubt whether Joe public is really checking out each fragment

Actually CG has spolied my enjoyment of many films as i keep looking at the CG SFX with C4D in mind, this happened a lot whilst watching LOTR

That said for a really spectacular slow mo shot of say the white house being blown away the collisions would have to be more complex and of course the geometry would be different

Maybe dust will cover a lot of problems and i also thought that the main particle fragments could be less numerous and more symetrical and perhaps have secondary parts ( smaller) fragmenting from them. It seems the smaller a fragment is the beter we can do the collisions as the center of mass is closer to the fragment boundaries

It is the long thin frags that cause the problems as the thin sides will bounce OK but the lengths will pass through floors and other objects

CGTalk Moderation
04-07-2005, 05:34 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.