PDA

View Full Version : Emitter and Collision Detection...



Samoht VII
01-07-2011, 09:49 PM
Ok here is what i am trying to do. Try and stay with me.

I have a character as seen in the image and he has a mask with eyes cut out creating a window into the head. Inside the head is a small circle that blocks viewing inside the head, so essential if you could imaging, this character could lie on his back and you could fill his eyes with water i guess is the best way to explain it. Anyways, i have created an emitter and i want the 'flames' to pour out his eyes like his eyes are smoking of sorts.

Check this out
http://www.youtube.com/watch?v=-D5wTYmcBX4
at 0:32 is exactly what i am trying to achieve.

So the way i see it I put an emitter in his head/eyes and make the particles move upward and then attach some collisions so the particles don't flow through the head but instead pour out the eyes.

My knowledge of how collision work and how to set them up isn't that flash hot though so i was wondering how it would be possible.

My character will also be moving around so i will be parenting it to the character.

Attached is what i have so far....

meshpig
01-07-2011, 10:47 PM
Hey the GonG! What you probably need to do is set up a simple version of the situation and start experimenting.

Try a type : object/mode : attract on the collision surface.

Also checkout http://www.newtek.com/forums/showthread.php?t=77002 there are several vids dealing with the basic settings.

Pkg the scene if you can.

prometheus
01-08-2011, 05:58 AM
If you can afford it I would suggest invest in turbulence 4d fluids..
http://www.jawset.com/gallery_t4d.php

Try the demo out if you havenīt, Its full featured but have watermarks.
http://www.jawset.com/downloads.php

Otherwise I would say you have two different particle voxel approaches to do this,

1. lower amount of particles and voxels/hypertextures fairly big

2. big amount of particles an very small particles without hypertextures, wind effectors needed to get vortex motions.

Right now you are using the first method and from what I can see you have to much density in there, I would tone down the whole thing and select your hypertextures carefully with a good tweaking of the hypertextures frequencies,contrast,scale and strenght.

you could use the second approach, and create geometry for the eye socket and add dynamic particles to that and set nozzle to vertices or surfaces so the emitter spawns particles from those to eye sockets.

I think I would create a high density spline or subdiv obj around the eye socket and set the emitter to object line, by doin so you can have the emitter spawning particles in a regular fixed way wich fools the eye to perceive it as an even density flow as
opposed to standard particle emitting wich has irregularities, and therefore more notable to the human eye.



Michael
http://vimeo.com/user680656/videos/page:2/sort:newest

UnCommonGrafx
01-08-2011, 06:26 AM
cheaper would be fire and smoke by pawel:
http://polas.net/smoke/index.php
Pretty old school; pretty usable at $60. The rolling vorti are pretty freakin cool.

Otherwise, using regular particles, I would use spherical collision balls in his eyebrow area in order to have the particles flow over that area. We don't talk enough about the goodness of collision balls as a mass for collision detection...

prometheus
01-08-2011, 06:35 AM
cheaper would be fire and smoke by pawel:
http://polas.net/smoke/index.php
Pretty old school; pretty usable at $60. The rolling vorti are pretty freakin cool.

Otherwise, using regular particles, I would use spherical collision balls in his eyebrow area in order to have the particles flow over that area. We don't talk enough about the goodness of collision balls as a mass for collision detection...

Personally... I wouldnīt invest in that old plugin, rather spend that extra money needed to fill in for a much future/feature friendly fluid plugin wich opens up many moor doors for effects in the future.

what the fire and smoke generator can do, that is something very reachable if you just tweak the particles carefully and using wind effectors properly within lightwaves native system, no plugin needed for that.

Granted..the fire and smoke had some nice stuff, weigh map emission..but that can be approached using images in birth rate or points placed where you want the emission.

Granted..fire and smoke has vorticles( particles themself as vortex fields)
thatīs nice an cool, but just add correct amount of winds and settings and your almost there.

Michael

UnCommonGrafx
01-08-2011, 06:43 AM
Ya know,
We will disagree on this. ;)

I bought it, and an extra copy of tree designer, because they work at what they do, and very very well inside of the "old" lw.

What you have described is everything that that $60 gets you past: the need to go to tweak hell in order to get a motion that resembles smoke and fire. It just works.
If I had need/money to throw away I would EASILY invest in Turbulence.

Personally, I did invest in that old plugin, after all these years, because it's priced right and does things differently than what's built in.

Fortunately, the poster of this thread gets to make their own decision as to what might work for them. And that is the point: gathering thoughts and opinion to make up your own damned mind. (a piece of dialog stolen from the matrix.)

prometheus
01-08-2011, 06:51 AM
Ya know,
We will disagree on this. ;)

I bought it, and an extra copy of tree designer, because they work at what they do, and very very well inside of the "old" lw.

What you have described is everything that that $60 gets you past: the need to go to tweak hell in order to get a motion that resembles smoke and fire. It just works.
If I had need/money to throw away I would EASILY invest in Turbulence.

Personally, I did invest in that old plugin, after all these years, because it's priced right and does things differently than what's built in.

Fortunately, the poster of this thread gets to make their own decision as to what might work for them. And that is the point: gathering thoughts and opinion to make up your own damned mind. (a piece of dialog stolen from the matrix.)

Amen to that, it also comes down to how familiar you are with winds and particles of course in order to do it nativly.

that quote..let me guess, the oracle?

Would be nice if we could..dodge the bulletīs or should I say, when youré so skilled in lightwave, we wouldnīt need any plugins:) but I guess thatīs only in genious programmers heads who can read and write the codes as they see fit.

Michael

UnCommonGrafx
01-08-2011, 07:02 AM
Creacon's talking about using a physix engine in LW and there is Core with Bullet: I await anything where it all just works together, without issues.

Old school can and does work. If this were my teacher account ([email protected]) it would be all about doing it natively. ;) But if it's about money, time is the issue more so than wishing and wanting natively great behavior of the app. We wish a lot with LW... Cheers to the ideal of Core.

Mr. Rid has shown that some really incredible things can come out of LW natively.

prometheus
01-08-2011, 02:16 PM
Creacon's talking about using a physix engine in LW and there is Core with Bullet: I await anything where it all just works together, without issues.

Old school can and does work. If this were my teacher account ([email protected]) it would be all about doing it natively. ;) But if it's about money, time is the issue more so than wishing and wanting natively great behavior of the app. We wish a lot with LW... Cheers to the ideal of Core.

Mr. Rid has shown that some really incredible things can come out of LW natively.

I donīt know if bullet is on itīs way to have intergrated fire and smoke fluids anytime soon, nor if physx will?

I would like to see some particle enhancements in Lightwave updates in the nearest cycle if possible, perhaps that could come through if they lay a foundation in core, or maybe it requires a completly different code in lw 10 specificly, I donīt know.

What I would like to see if possible, is to push the limits of particle renders
push it away from the 1 million mark to reach several millions or even a billion mark would be nice, perhaps some iteration renders in partitions is needed, something along krakatoa.

I would also like to see a subframe calculation of particles, and a distance between particles gradient on every voxel channel possible.

particle vortices( vorticles would be nice) and add some additional vortex
noise field to enhance the particle flow and make them work when particles follow wind paths , as it is today they donīt work together.
The wind path needs an overhaul too where you simple plot/draw out a spline curve for it to follow instead of having to add and copy null handles today.

Michael
http://vimeo.com/user680656/videos