I finally get thinking particles!

[Guest Tread]

word

[Guest parallax]

Burnout!

 

For real? good luck, i too am hanging on by a thread.

Edited June 1, 2006 by parallax

[Guest SermonOfMockery]

that spray paint looks fantastic- well done!

[Guest T02]

omg @ that spraypaint! Amazing!

[Guest palmer]

Is there a difference between cebas thinkingparticles(max) and cinema 4d thinkingparticles? are you both using cinema I guess is what I'm getting at?

 

The paint drips look dope man. I think there might need to be some divergence in the timing to make it a little more random feeling.... but otherwise it is a great effect and I look forward to seeing it in production. Definitely the most successful I've seen, I look forward to further iterations.

Edited June 3, 2006 by palmer

[Guest Chinaski]

Is there a difference between cebas thinkingparticles(max) and cinema 4d thinkingparticles? are you both using cinema I guess is what I'm getting at?

 

The paint drips look dope man. I think there might need to be some divergence in the timing to make it a little more random feeling.... but otherwise it is a great effect and I look forward to seeing it in production. Definitely the most successful I've seen, I look forward to further iterations.

 

 

All these tests are done with Cinema 4D. TP in Cinema is the same Cebas plug-in for Max, although they are not 100% the same. I haven't used the system in Max, but looking at Cebas' website, it appears the max version has a little more functionality and more/different nodes.

[Guest Srek]

TP for CINEMA is based on TP 1 for max. In the meantime TP 2 for max has been released.

TP for CINEMA however has some advantages to TP on max since it is fully integrated and uses the native node based expression system Xpresso. This allows for much more and tighter interaction between the scene and TP then in max. TP for max has a few more nodes then TP for CINEMA, however this is mostly due to it having to make up for what Xpresso can already do.

Cheers

Björn

[Guest Chinaski]

Not thinking particles, but since this thread was heading in a spray paint direction, I decided to do a small After Effects tutorial here:

 

http://mograph.net/board/index.php?showtopic=8119

 

Works pretty much like my C4D spray paint system, but with much more limited functionality.

[Guest Tiziano]

I just had a breakthrough.

 

I think I finally get Thinking Particles in Cinema 4D and how to get them to do what I need.

 

Thank 'effin god...

 

Good lord that took a long time.

Too bad that Vulcan Mind Meld thing never really worked, LOL.

Man do I have a mountain to climb in figuring this out. The pBlurp stuff is pretty simple but collisions are driving me nuts

[Guest Chinaski]

A couple more quick tests, using TP to drive the shape of an object.

 

These are based off of Srek's files, so any credit for the technique and the initial Xpresso set-up goes to him. I'd like to push this further by seeing if I can morph between shapes with all this crazy stuff going on in-between. That will be my next test.

 

Render 1

 

Render 2

 

Render 3

[Guest TimC]

cool stuff!

 

Was that using the particle position to fix the objects points to? I think Per Anders originally did this in his soft bodies example. Obviously Bjorns contribution is not in question here, he provides some awesome examples.

 

nice..

 

cheers

Tim

[Guest Chinaski]

cool stuff!

 

Was that using the particle position to fix the objects points to? I think Per Anders originally did this in his soft bodies example. Obviously Bjorns contribution is not in question here, he provides some awesome examples.

 

nice..

 

cheers

Tim

 

 

Yes, that is how its done. I don't know if the technique originated with Per Anders or Bjorn. I got the set-up from Bjorn's site and was not aware that Per Anders had done this as well.

[Guest C.Smith]

Yes, that is how its done. I don't know if the technique originated with Per Anders or Bjorn. I got the set-up from Bjorn's site and was not aware that Per Anders had done this as well.

 

Normally I'm preaching about how C4D uses much simpler steps to achieve a goal than Maya. But in this case, Maya is just "Convert to softbody" then add a "Newton" force to achieve this. Well, basically that.

 

Beautiful renders.

[Guest Chinaski]

Normally I'm preaching about how C4D uses much simpler steps to achieve a goal than Maya. But in this case, Maya is just "Convert to softbody" then add a "Newton" force to achieve this. Well, basically that.

 

Beautiful renders.

 

 

You can do softbodies in Cinema, as well, without going through all the fuss it took to set this up, but I don't think you'll get the same result. There are softbodies in the dynamics package, or very simply, you can even just put a cloth tag on your object. I haven't used them much, but I believe softbodies has a newtonian gravity force in there...

 

The idea with doing it with particles is that you can use rules to determine very specifically how the particles move and interact with the scene. I want to get them to reform into another object. I'm not sure if that is possible with softbodies or cloth. Basically, to set this up, you have to assign the particle index to the geomety points' index. To get them to reform, you'd have to make sure that they get the new object's point index and go to that location. I can get them to reform into the right shape, but not with the right point id's, so the shape is all twisted. I'm still working on that...

[Guest C.Smith]

You can do softbodies in Cinema, as well, without going through all the fuss it took to set this up,

 

The idea with doing it with particles is that you can use rules to determine very specifically how the particles move and interact with the scene.

 

I have Dynamics (w/ Studio bundle) but haven't played with it much yet. In Maya, Converting to softbodies basically turns all vertices of the mesh into particles. Then any control over normal particles applies equally to the softbody. Does the C4D softbody allow TP control over it's softbody as well?

[Guest Chinaski]

I have Dynamics (w/ Studio bundle) but haven't played with it much yet. In Maya, Converting to softbodies basically turns all vertices of the mesh into particles. Then any control over normal particles applies equally to the softbody. Does the C4D softbody allow TP control over it's softbody as well?

 

That's a good question. I'm not really sure about it... The dynamics module uses its own system that is a bit different than the regular Cinema 4D stuff. (It has its own forces, etc.) I'd have to play around a bit and see what kinds of interaction are possible between TP and softbodies. Through Xpresso, there probably is a way. Just might not be that easy.

[Guest Srek]

Nice stuff Chinaski

As for the question if it was Per or me, we did so many things overlapping, similar but differnt and we often exchanged our work so it doesn't realy matter

 

Cheers

Björn

[Guest BillD222]

wOW, beautiful stuff.

 

Chinaski, have you ever used particle flow in 3dsmax? If you have, how does the C4D Thinking Particles compare to it?

 

 

 

 

A couple more quick tests, using TP to drive the shape of an object.

 

These are based off of Srek's files, so any credit for the technique and the initial Xpresso set-up goes to him. I'd like to push this further by seeing if I can morph between shapes with all this crazy stuff going on in-between. That will be my next test.

 

Render 1

 

Render 2

 

Render 3

[Guest Chinaski]

wOW, beautiful stuff.

 

Chinaski, have you ever used particle flow in 3dsmax? If you have, how does the C4D Thinking Particles compare to it?

 

Yes, I use particle flow in 3dsmax all the time. I really like it. I like both systems. To be honest, I find particle flow much easier to use than thinking particles. The set-up time for a particle system is much quicker in pflow, and the nodes have a lot of options built into them, whereas with TP, you have to add a lot of nodes to get the same kinds of options. TP, also, does not have subframe sampling in it, which is a big problem for a lot of things I want it to do. I'd like to see TP be able to do a bit more, such as create wetmaps, have true particle collision detection (rather than bounding spheres), have a splitter node for easily separating particles into groups, etc (I could name more, but this thread is about what TP can do, not what it can't). Hopefully, those will come in a future release.

 

I think the real strengths of TP are how closely integrated with Xpresso it is. Being able to easily pass scene info back and forth between all kinds of objects and particle systems is an extremely powerful tool.

[Guest TimC]

TP, also, does not have subframe sampling in it, which is a big problem for a lot of things I want it to do.

 

Well it does actually. It is however pretty limited. You can link the position velocity of the emitter object with the emitter velocity port on PStorm.

 

See this file HERE (right click and download)

 

In the Xpresso unlink the connection into emitter velocity to see the difference.

It is still not enough though! Just thought I'd point it out incase you were not aware.

 

cheers

Tim

[Guest Chinaski]

Well it does actually. It is however pretty limited. You can link the position velocity of the emitter object with the emitter velocity port on PStorm.

 

See this file HERE (right click and download)

 

In the Xpresso unlink the connection into emitter velocity to see the difference.

It is still not enough though! Just thought I'd point it out incase you were not aware.

 

cheers

Tim

 

Thanks Tim!

 

Yes, I was aware of that and used it in my spraypaint file to get an even distribution of paint particles. But, it seems like you know what I'm talking about here. I've been trying to create an easy splitter system to split percentages of particles into groups, but it just won't work right because of the subframe sampling issue. Basically, I'm comparing a random node to a constant and sending out the result to a pgroup node. The percentage of splitting works correctly, but the particles come out in clumps (due to the fact that random numbers are sampled on frames, not subframes). Switching out a noise node for random works, but I have not found an easy way to get noise to output consistent values between 0 and 1 that you can easily convert to percentages. (I've used the range mapper. You can try and see if you can get it to work correctly. I haven't been able to get it right.)

 

Besides that, it would help with collision detection and a bunch of other functions.

[Guest C.Smith]

Switching out a noise node for random works, but I have not found an easy way to get noise to output consistent values between 0 and 1 that you can easily convert to percentages. (I've used the range mapper. You can try and see if you can get it to work correctly. I haven't been able to get it right.)

 

Besides that, it would help with collision detection and a bunch of other functions.

 

I'm not in front of C4D right now, But I wonder if you used a gaussian noise that did it's bell curve peaked at .5 between 0 and 1 then used a clamp at 0 and 1 to catch any spill.

 

just a thought.

[Guest Chinaski]

I'm not in front of C4D right now, But I wonder if you used a gaussian noise that did it's bell curve peaked at .5 between 0 and 1 then used a clamp at 0 and 1 to catch any spill.

 

just a thought.

 

 

Yeah, that's kind of what I've tried. In theory, it works, but in practice, it doesn't give a good distribution of random numbers, even if I change the frequency and scale parameters. The numbers tend to clump together or keep repeating the same or similar values. Compare that to the type of distribution that a random node gives and you'll really see the difference.

 

But, maybe my set-up is not quite right. Let me know if you can get it to work well. I'd love to figure this little problem out.

[Guest ChrisC]

I've been trying to create an easy splitter system to split percentages of particles into groups, but it just won't work right because of the subframe sampling issue. Basically, I'm comparing a random node to a constant and sending out the result to a pgroup node. The percentage of splitting works correctly, but the particles come out in clumps (due to the fact that random numbers are sampled on frames, not subframes)...

 

Hi Chinaski - don't know if you saw this thread, but it might help you in the right direction, there's a couple of workarounds.

 

Nice thread all, I'll jump in when I can!

 

C

[Guest Chinaski]

Hi Chinaski - don't know if you saw this thread, but it might help you in the right direction, there's a couple of workarounds.

 

Nice thread all, I'll jump in when I can!

 

C

 

 

Thanks for chiming in, Chris!

 

Actually, I was aware of that thread. Bandini is my username on cgtalk (Chinaski was taken, so I chose another literary character). You were a big help there.

 

It took me a little while to get the hang of it, but I completely understand the how and why of that technique for splitting particles into groups. Just, the randomness never quite worked out correctly... I do think that creating multiple emitters would work fine for this, just takes a bit more time to set up.

 

If you want, you can see a file that I've got set up for testing this. There's a noise node in there that can be swapped out for the random node. Doesn't seem to make too much of a difference, though...

 

Multiple targets with one emitter

 

I'm not really looking for a solution to this in this thread, although I'd love it if there was one. I know there are some workarounds. Personally, I really would like to see a splitter node added in a future release of TP...

 

Anyway, I really wasn't trying to complain about TP, because I think it is very good, and the point of this thread is to start showing what can be done with it, not bitch about why it doesn't do this or doesn't do that. But, comparing it to PFlow, which is what BillD222 asked, I think some things are done much easier in PFlow, this being one of them.

Edited July 6, 2006 by Chinaski