PDA

View Full Version : Particle Line fx, how did we do this?



Igu4n4
05-03-2011, 07:43 PM
About a year ago, (perhaps a little longer)... there was a thread either here or on Spinquad (I've searched on both, and can't seem to find it), where we were discussing how to create the effect of organic lines growing from an emitter. Essentially using fixed position particle emissions with an object-line nozzle, so that successive particles would emit out of the same position. Keeping them close enough and with enough particles would result in line organic lines growing from the emitter.

I can't seem to find the discussion, most likely because I can't think of the most appropriate search term. I am attaching a sample of what I'm talking about from a rendering I did. I just can't remember the exact technique, and others were doing the same technique, hence the discussion. I can't find the source files to look at what we were doing, to help reconstruct it...

Sooooo... I'm hoping that someone who remembers, or was part of it, can point the way back to it ;)

Image attached.

94834

thanks!

Steve.

manasa
05-03-2011, 11:52 PM
maybe...
http://www.spinquad.com/forums/showthread.php?21294-Fine-particle-streams!

check it

Marcelo

prometheus
05-04-2011, 12:50 AM
Yeah I remember those:)

I did that with a flat box as obj emitter set to object line, and a wind with procedural textures on the vectors in the direction mode that is set for the wind.
Hv..tiny sprites is applied on the particles.

hereīs a variation..not the same thou, this is using vortex winds
http://vimeo.com/1745145

as for the image example, it might look good as an image, but getting a good animated clip might be trickier.

Michael

Igu4n4
05-04-2011, 07:18 AM
Hey Michael, Marcelo,...

Ya this is the stuff... I remember that first post was what got me interested in it. I had some great results.. even animated, I just couldn't remember the settings. I'll look through the thread again and give it a go. Just wish I could find the scene files as I did a bunch of R&D on the technique, and now can't remember any of it lol.

Thanks again :) !

Steve.

Igu4n4
05-04-2011, 07:39 AM
Michael, I read through the old post on SQ again... definately got things going again... I know it was a long time ago, you had mentioned you were uploading a scene file, you don't happen to have that initial stream scene file hanging around anywhere do you? Having a bit of trouble emulating your initial setup.

Steve.

prometheus
05-04-2011, 08:04 AM
Michael, I read through the old post on SQ again... definately got things going again... I know it was a long time ago, you had mentioned you were uploading a scene file, you don't happen to have that initial stream scene file hanging around anywhere do you? Having a bit of trouble emulating your initial setup.

Steve.

Uhhm..I donīt think I ever posted a scenefile..or did I:question:

have to find it myself or do another setup in such case.

Michael

Igu4n4
05-04-2011, 08:12 AM
yea, I don't think you did.. you just might have said you were going to.

The thing I'm having difficulty with is the large spread you had going on in the initial example. Everything streaming away nicely. I end up getting a lot of streaming away, but also a lot of inward turning back into the square, and I can't seem to get anywhere near the length of strands that you had on the initial example. So was just looking for some insight.

So obviously there's a lot of things that effect everything else, including the wind settings.. a default wind at what size, power etc... with the turbulent vectors.

Anyhoo.. any other thoughts to get towards that initial example would be great. (I'm planning on doing a short animation about time, and would love to use this effect).

Steve.

prometheus
05-04-2011, 08:15 AM
yea, I don't think you did.. you just might have said you were going to.

The thing I'm having difficulty with is the large spread you had going on in the initial example. Everything streaming away nicely. I end up getting a lot of streaming away, but also a lot of inward turning back into the square, and I can't seem to get anywhere near the length of strands that you had on the initial example. So was just looking for some insight.

So obviously there's a lot of things that effect everything else, including the wind settings.. a default wind at what size, power etc... with the turbulent vectors.

Anyhoo.. any other thoughts to get towards that initial example would be great. (I'm planning on doing a short animation about time, and would love to use this effect).

Steve.

Are you using turbulence or vortex winds, or do you actually use the directional wind with procedural texture on the vector channels?
thatīs what I used, but yes it can be tricky to choose right settings within that anyway.

Michael

Igu4n4
05-04-2011, 08:42 AM
yea, am using a directional wind with a turbulence proceedural in the vector channel, trying with different powers, complexities etc.

Your original example had such a nice plasma sorta of organic feel, with very little moving inwards.. mostly moving outwards... mine seems to get stuck up around the wind vectors so the various arms start to spiral around them, and some move back into the square.

Have been experimenting all morning, and get some lovely shapes, but am trying to break it back to your original, and having a bit of a hard time.

It's fun though non? :)

Steve.

prometheus
05-04-2011, 10:14 AM
if you start with a simple flat box 1m with and 1m height and 0 depth.

use that as the object emitter,set the nozzle as object line, increase particle limit to something like 20 000, the first thing you could add in the particle tab would be an explosion force around 10, the particle emitters birthrate could be something like 370..now the dynamic wind needs to be quite strong ..something like 2000, but that depends on if you choose to change particle resistance or weight, but by default settings ..try 2000.
It could also depend on if you choose to use a stronger value in the procedural texture instead..like the layer opacity or texture value.

further more we need to stretch our hv sprites in the velocity channel to avoid particle gaps or to Fake that the particles blends in together.

check your inbox.

Michael

prometheus
05-04-2011, 10:42 AM
another sample...other procedurals or some size changes of the procedurals instead of 1m..try10 meter, It yields different results.

And Important..is the stretching in velocity of hvs...in the attached image you can see how it looks without..and with.

I got to go back to the tele now, some hockey world championship and swedes against the u.s.a
I believe itīs a winner for the us today thou.

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

Igu4n4
05-04-2011, 11:22 AM
ah good on ya for the hockey game... living in Canada, I can appreciate the priorities. Thanks for posting the SS's... useful.

I was mucking around a bit more today, and decided to play around with using surface instead of sprites ... can get some great abstract stuff, and it animates nicely too.

94879

cheers.

Steve.

prometheus
05-04-2011, 12:20 PM
surfaces can be interesting, depends on the look you want, I still prefer sprites, since where particle density is a lot or if particles is close or overlapping eachother it will give a nice density increasement and a contrast to where itīs low. and you need to set a low opacity to have that effect where light can pass trough a little, this is not possible with surfaces, sure you can get similar with light incidence angles and other, but not a truly light interaction particle interaction shading based on particle density and relation ship.

This would look so extremly nice if the partice size is very small, and if you could have millions and millions of particles, now these cool images are still to be acheived with lightwave when newtek removes that d.a.m 1000 000 limit of particles and push it A lot more, along with a distance between particle gradient on all channels possible, but especially the dissolve or density channel, that is to help fade out particles where they are away from eachother.

subframe particle calculation might help too.
All this would be a sort of trailing on to those krakatoa particle renders for max.

There you go, now I did it again newtek..nag..nag:)

Michael

JBT27
05-04-2011, 02:17 PM
So the curving lines are strings of HV sprites? Yes, I've read the thread and messed about, but I'm not getting individual streams curving like the fine examples here ... I'm remembering something about motion blur with very high Blur Lengths, on an old thread as well ... but this is not that technique so it seems ... ?

Julian.

Igu4n4
05-04-2011, 02:33 PM
Yea. The reason I was mostly isolating things down with Michael, is that his original example, is what started a few of us down the road to playing around with this a couple of years ago. I had forgotten some of the steps (getting old).

My original example at the start of this thread, is something that was built with text in mind, and the lines wrapping around/engulfing it, but his his original example from the original thread has some really nice characteristics which take a bit of work to get to, hence the converstation :). But if you follow the steps he outlined earlier, you can arrive at at the particle lines.

Essesntially, it's an emitter using a fixed random order for particles to be emitted (so successive particles originate from the same spot as the orignal set does). They are released close enough to each other that when you shade them through HV, they appear to be touching, and thus, appear as a line. You then can do a lot of things to the HV's to alter the lines appearance, (adjust colour with speed/age, use the size as a gradient of age, for example). And then alter the overall structures appearance by changing the wind parameters, or the weight, resistance, of the emitter (as well as other settings).

There is no need for motionblur to achieve this, (except if your animating and say want the camera to feel richer). The line have to do more with the quantity of the particles and how long they remain (ie: don't die).

If you need any help, just ask, that was the purpose of this thread to revisit the technique. But as I said, Michael, almost wrote it out step by step in an earlier post in this thread (about 3 back).

Steve.

JBT27
05-04-2011, 02:44 PM
Great - thank you. Yes, Michael's outlined steps are clear enough and I've been messing about with them (when I should have been doing something else :D)

On and off, I have experimented with partigons and motion blur, since that original thread, and that works well, but like yourself I guess I've had other things and gradually I forgot some of the detail ... also blaming age on that one :D

Julian.