Particle Flow Discussion

#4915

Cool - actually what would be great was if there is something similar to the Time Scale setting inside Reactor. So that if you are pleased with the settings, but the director or supervisor found everything to be too slow, you could easily crank it up and keep on adding to the simulation afterwards.

#4916

Physx has veeeeery cool time scaling - best i’ve seen in a solver, check out the rayfire demos.

#4917

^ I like Rayfires time scale implementation as well, that thing is fun:)

Here are a couple of simple examples (tweaked Oleg’s idea a little, seemed like a good platform :slight_smile: ) of adding more gravity…their are a host of other tweaks you can perform but for the sake of simplicity.

Test 1 Gravity Default 980.066 - QT 16 mb

Test 2 Gravity 2980 - QT 13 mb

Test 3 Gravity 5980 - QT 15 mb

and for the heck of it 12168 particles could use a little tweaking to settle. sorry, its 500 frames and 83 meg. so you know.

#4918

that first one looks like there is some jointing going on. hows that created and how does it break?

#4919

^A little bit of Glue :slight_smile: with the ability to set binding strength/breakability plus a host of other options.

#4920

Great tests, keep them coming :smiley: Cool to hear that physx has a time scale feature, that should take care for everything. Is it animatable as well?

I’m bidding for a VFX-job now that calls for some major destruction and particle work. I’m crossing fingers that Box#2 will be released within a month or so, would be a cool addition to the stuff the client is looking for…

#4921

NO.
But anyways, it looks awesome must-have tool :thumbsup:

#4922

Deko is back, Deko is back!!!

don´t be a stranger no more around here :frowning:

kind regards,
Anselm

#4923

When importing meshes or objects from RealFlow, an Integration Step lower than 1 frame in ParticleFlow screws up the whole flow.
Has anyone else experienced this?

Cheers

#4924

Do you guys have seen the severals sample scenes about Pflow in the Max2010 SampleFiles DVD?
Awsome! many many good samples. Very good to work faster.
most of them are shaded, ready to render.

#4925

Hey Anselm, nice to see you :slight_smile:
Still hope so that you will show even more Box2 examples… :smiley:

#4926

I am currently crumbling an wall with PF. I want big chuncks to fall faster than small bricks.
Bobo has a tutorial about this (http://www.scriptspot.com/bobo/mxs5/pflow/pflow__force_by_mass.htm) but I use instanced shapes from the scene instead of basic particles (http://laurent.renaud.free.fr/tuto01-uk.html) and the Bobo 's setting point size on particle size. Is there a script line to measure each instance volume?

thanks

#4927

Assuming you’re adding the particle shape in the birth script, you can add the function for calculating the volume there.

Faster than calculating the volume would be to calculate the difference between the bounding box minimum and maximum using the nodeGetBoundingBox or nodeLocalBoundingBox methods.

For example, you can multiply the dimensions of the node’s bounding box by each other to get the volume of the bounding box:


bbox = nodeGetBoundingBox theObject theObject.transform
bboxVolume = ((bbox[2].x*2)*(bbox[2].y*2)*(bbox[2].z*2))/2
pCont.particleFloat = bboxVolume

Note that if the object has been scaled, the scale needs to be taken into consideration because the object’s transformation matrix itself is used for nodeGetBoundingBox.

If you want the true volume, try this bit from the maxscript reference:

There is currently no access to the Measure Utility from MAXScript, but you can use the following MAXScript function to get the same results:

SCRIPT


fn CalculateVolumeAndCenterOfMass obj = 
( 
local Volume= 0.0 
local Center= [0.0, 0.0, 0.0] 
local theMesh = snapshotasmesh obj
local numFaces = theMesh.numfaces 
for i = 1 to numFaces do 
( 
local Face= getFace theMesh i 
local vert2 = getVert theMesh Face.z 
local vert1 = getVert theMesh Face.y 
local vert0 = getVert theMesh Face.x 
local dV = Dot (Cross (vert1 - vert0) (vert2 - vert0)) vert0
Volume+= dV 
Center+= (vert0 + vert1 + vert2) * dV 
) 
delete theMesh
Volume /= 6 
Center /= 24 
Center /= Volume 
#(Volume,Center) 
)
#4928

Glad you liked them! Unfortunately when I was commissioned to create them I was given a lot of limitations like no volumetrics or plugins, so just straight pflow which meant creating fire or explosions a bit uninteresting, of course the particles are all in order just when it comes to rendering if you throw them through afterburn you’ll get some much nicer results! Other than that a lot of the more complicated setups I needed to keep playing in real time so it couldnt ever get too complicated, as well as various other limitations.

So I think all in all there are a lot of set ups that will save people time, as well as a lot of more technical set ups that will show others how to do some more tricky types of effects, rather than the same ol stuff. But just on the visual side its a bit bare without krakatoa or afterburn to render the FX out!

How are you Loren? Cool FX post btw!

On a side note, its my birthday tomorrow so I’m off to get drunk :wink:

Allan McKay

#4929

How can I attach omni lights to PFlow particles? I would like to make a fireworks animation, where there is a light effect on the buildings / flat surfaces just like in real world.

I have tried to attach spheres and apply vray light material, but vray didn’t give me smooth results for this 200+ small vraylight objects, but I think with simple omnis, the results will be smooth.

Do I need to use maxscripts for that?

#4930

yep you’ll need to use maxscript
I’d write it up here but I’m in severe crunch mode right now - perhaps someone else can, but its relatively easy you just need to build 50 lights for 50 particles (you can get it to generate them but it doesnt like deleting and creating on the fly so gets a bit unstable at times) and link their position

so the idea is something like this, but obviously you’ll need to flesh it out a bit more.
first create 50 lights if you’re using 50 particles. then…

in the creation section:
mylights = lights as array

runtime:
for i = 1 to pcont.numparticles() do
(
pcont.particleindex = i – makes current particle in the loop active
mylights[i].position = pcont.particleposition
)

something like that, hope it helps

-AM

PS. you could tell them initially to all locate to [9999,999,9999] in world space, or turn off on the first frame, and then turn on or relocate as they get assigned, this way they arent in the end position of the last frame when you cycle back through.

PPS. As mentioned about about the pflow assets in max 2010 I believe I created a scene file that ships with max that does this if you have 2010.

#4931

edit never mind, Allan covered it. Bobo has the same method covered in detail as well:

http://www.scriptspot.com/bobo/mxs5/pflow/pflow__MovingObjects.htm

#4932

Thank you for the detailed answer. I will see this maxscript way soon, but before I get into it I think it’s better to ask.

As I realized, there is the lighting and there is the visual effects about a fireworks, and these two are totally independent. In the real world they are not, but in 3D, I can do anything I want with the lighting and do something else with the rendering of the particles, and it’s just about personal taste which makes it look good. I mean the best way is not always physical correct simulation and GI, but the one which is fast and looks good to human eye.

What is closer to physical simulation would be the sphere geometry + vray light way both for rendering and for GI, but as it seems it’s just not efficient for animation. The technically better way would be not vraylight but omnis instanced using maxscript. The problem is that there are at least 2000 particles (think about a real fireworks), which could be a problem for maxscript (I will see soon).

The third way would be faking, by first rendering the scene (in linear color space), measuring the average luminosity of each frame (using some utility) and keying for a vray sphere light’s size and multiplier from this data. I think this needs the most work, but it produces the smoothest rendering results (there is just 1 dynamic light).

Do I see things correctly? Which way would you suggest me? I think it’s more important to take care about the rendering effects than about GI lighting, I think even a simple curve manipulation in composition could do the trick for the “wow” effect of fireworks.

#4933

allan, could you please reset your server download bandwidth or something? :beer:

#4934

I’ve got a completely new server, its lightning fast, unlimited bandwidth, you can download 5gb in under 2 hours…
However I’m just totally swamped trying to finish up 2012. Sometime in May I’m going to switch allanmckay.com over to the new server, at this time there’ll be a new website launch, LOTS of new free training content and I will finally get rid of the bandwidth issues I’ve been plagued with the the last half a dozen years finalllyy
so give me a little it of time, probably end of the month if time permits, I wrap this film may 14th so this is one of the many things I have on my agenda as soon as its finished.

There’s lots of new stuff for max 2010 I want to get into, and ICE and real flow too

I am also speaking at FMX next week if anyone is attending in Stuttgart, I’m suprised it seems like everyones attending this event so it’s going to be pretty swank!