PDA

View Full Version : Particle Flow Discussion


Pages : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 [16] 17 18 19 20 21 22 23 24

SoLiTuDe
01-13-2007, 08:10 PM
so... lemme get it straight real quick. You're using the bottom plane to emit some particles, tell them to go into the vortex, and if they hit any buildings they spawn some more particles, correct? I think the only reason it's going so slow is because of the udeflector. It takes up a lot of processing power... and you're emitting from -100 so even at frame 0 it's got a lot of stuff going on. For starters, turning off real time playback will help you see what's going on a lot better... i don't think there is really a way to get rid of the udeflector / collision spawn in this case if you need the debris to break up on a collision with a building.

Depending on the viewing distance / detail needed you may be able to get away with not using a collision detection to spawn the particles on the buildings. ... maybe just birth particles on it, and have them spawn a few extra at the time one is born? ...this way it will still look similar, minus seeing something actually hit it. Or you could combine the two... cut down how many actually collide with the buildings, AND have some just spawning where a "collision" would be. If you don't get what I mean i'll post up an example

everlite
01-13-2007, 08:40 PM
plejboy

Cheers, though what do you mean by rough up the deflector? The Cache seems to work a little.

SoLiTuDe

Yep that's about it. I think i understand what you mean, though a little confused by; spawn a few extra at the time one is born? it already spawns on collision, is this not the same thing as what you mean?

Cheers guys, any more input very welcome.

- Dave.

Glacierise
01-13-2007, 08:56 PM
Hi Dave,

I have the same CPU, and only 1 Gb of RAM, and still was able to play the anim. The deflectors are the problem, yeah. And you could cache the flow, though you are limited by your RAM. In a chaotic system as yours, you could also create the movements you want with a small number of particles, then boost the counts at render time.

P.S.: What he meant by 'rough up' is creating multiple simple deflectors, that together approximate you complex shape ;)

plejboy
01-13-2007, 09:03 PM
Hey Dave.
I meant that you could build a very simple representation of your city with only deflectors. If you just make the ones you need it shouldnt be too much of a hassle.
Or just do what Glacierise said and lower the viewport count of the particle system.

Anyways, good luck. Time for a beer over here!

everlite
01-13-2007, 09:31 PM
I see, so effectively rebuild the facing geometry with deflectors, ok.
Strange when i reduce the viewport count this has no effect, does it still calculate the collision for all the particles even when the count is less than 30% which looking at my scene isn't that many.

Cheers again - Dave.

SoLiTuDe
01-13-2007, 09:54 PM
--Everlite, instead of having collision spawns ...Imagine a gun firing at a wall (from off camera), you don't necessarily see the bullet, you just see the effect of the bullet hitting the wall (sparks, and dust), so you don't have to create a bullet, you just have the bullet-hits, without calculating the bullet. So to do this, I'm suggesting you create another birth event, set to position object, then add a spawn test to that event, (set to spawn much like your collision spawn but without a collision) this way it will just create "explosions" or basically the same effect as the collision, just that you won't see a paticle come and hit it. Of course, this is ok for some of it, but i would also do what the others have said... create regular deflectors around parts of the buildings where you would actually see the collisions. And this makes more sense, especially because you won't be seeing the collisions behind the buildings. Same goes for the first part of my solution, use selected parts of the mesh, rather than the entire thing for birth since you won't see the back side.

everlite
01-13-2007, 10:17 PM
Cheers mate, i'm tuned into what you're saying now. Just curious, speaking theoretically, how does one handle a scene where you have large/dence geometry and you need to set up collisions, ie; if you had rain falling onto an animated character, and the rain itself spawns on impact, is it just a case of better hardware can handle the demand or is an alternative approach needed?

Cheers again - Dave.

Werewolf006
01-14-2007, 12:00 AM
Cheers mate, i'm tuned into what you're saying now. Just curious, speaking theoretically, how does one handle a scene where you have large/dence geometry and you need to set up collisions, ie; if you had rain falling onto an animated character, and the rain itself spawns on impact, is it just a case of better hardware can handle the demand or is an alternative approach needed?

Cheers again - Dave.

by making 2 systems one as the character (or a proxy preferably) an emmiter
and the other as the rainfall. like solitude said no collision needed here things are usually Happening too fast for the eye to track each raindrop.

NB this doesn't include cases like pixar's "a bug's life" huge raindrops and the super-punch in the matrix obviously.

everlite
01-14-2007, 12:10 AM
You do tend to see rain, be it a little blurred, but i see what you mean. though you would see the slash, which itself would require a collision to spawn? I think i saw a tutorial on Alan Mckays website showing how to do this.

SoLiTuDe
01-14-2007, 01:16 AM
--he uses a sphere deflector though which is much faster than a udeflector, and yes he uses a collision spawn, but for farther away shots you don't need a collision spawn to create splashes since you won't be able to see the drops hitting the object, you can just create spawns to create the splashes instead.

buller
01-23-2007, 02:56 AM
hy guys īn gals!

just did some (payed *yeah* !) research for a TV spot to be produced from january on.
the main character is going to "die" and resurrect again serveral times until a certain product sets him a final end...

therefore i started to learn Thinking Particles by heart. hereīs a first "T1000 shatter death":

http://core2core.de/spielwiese/TP_schimmel-fracture-test_1&2.wmv

(the biped is a free one from the net, nothing oblige here...)

hope you like it! C&C welcome. donīt be too harsh, itīs my first TP render :D :D

kind regards

anselm
hi
good fx
keep them coming
many people learn from you
( free rig is brad noble's,i think ) hehe

depleteD
01-23-2007, 03:09 AM
That was a really cool effect. Fragmentaion/RB solving is something I really wish PFlow did. I've used TP in C4D a long time a go and started to like it...then I quit. Yea man I hope u do more tests like this one psycho. Fragmenting is my 2nd favorite effect. Assembly is my first ;). Think you could make all those pieces reform? :P

U have 2 runs in here. I like the first one alot better. Its awesome. I really wanna see that on a badass character. The 2nd test has force applied off the object and it looks weird. Why are those pieces leaping off the character.

Keep up the good work man, looking forward to seeing more of your stuff.

-Andrew

buller
01-23-2007, 03:34 AM
there was double slash and space in the url.
corrected versions:

http://maxima8.abac.com/dekovv/tmp/pflow_dust_destructions.rar

http://maxima8.abac.com/dekovv/tmp/pflow_kiddaz_sphere.rar
hi
the links still dead?
BTW thnx
you guys rock!

loran
01-23-2007, 09:07 AM
Is there a particle flow tool box#1 update for 3dsmax 9??

PsychoSilence
01-23-2007, 09:47 AM
hi
good fx
keep them coming
many people learn from you
( free rig is brad noble's,i think ) hehe

thank you guys :) glad you like it.

both shots are basicly the same setup but with different fracture maps/fracture gradients...

the test character is taken from crowd it, a great crowd sim tool: http://crowdit.worldofpolygons.com/What_is.html

definetly worth a try! (please forgive me some minor product placement ;) )

the biped/mocap data is from somewhere on the net...forgot where i downloaded it...long time ago... :)

hopefully within the next days i can show more from "moldy" the mold monster :D (the commercial is going to be for a sanitary product...). seems like the sanitary industry like our stuff :D we did "clorox - mermaids" and "clorox - pirates" during summer 2006 :D

@ deleteD: yeah! liek TP more and more...even if the setup takes ages compared to flow :( in my particular case i connect the best of both worlds eventually. "moldy" is going to die by explosive-arrows in one shot so "winchester" by James Haywood is weapon of choice for the shooting. fragmentation is more a thing of TP stuff. so i dunno how weīre get stuff going for now but we WILL get that going :D

@ loran: havenīt heared of anything like an update for max9 yet :( maybe ask over there at orbaz.com

kind regards

anselm

oatz
01-24-2007, 10:29 PM
Is there a particle flow tool box#1 update for 3dsmax 9??
Probably. Box 3 is, although it doesn't have an installer. I'd shoot an email to support at orbaz dot com.

Khye

charleyc
01-24-2007, 11:04 PM
Probably. Box 3 is, although it doesn't have an installer. I'd shoot an email to support at orbaz dot com.

Khye

I got an e-mail from Turbosquid just the other day on an installer for both Box's (Box3 didn't need it though...that has been available for a few months now). However, I am having some problems with Box1 in R9-32. I would love to know if anyone else has the same problems. I have already sent files to Oleg, but it would still be good to know if anyone else is getting them.

PsychoSilence
01-25-2007, 12:04 AM
we havenīt changed yet...but tell me please when box#1 will be fully supported :)

does anyone know if or when thinking particles will be supported?

AB and FUME allready are...

kind regards

anselm

loran
01-25-2007, 08:25 AM
ih guys :)
I recieved the email to PF tools boxes updates. I installed it and it seems to work fine. What is your problem Charleyc ??

charleyc
01-25-2007, 05:37 PM
Box1 is causing some crashing (when using Group operators) and the Lock/Bond Test does not do what it is supposed to. If they are working fine for some, then I assume it tis something on my end and will try and remove the plugins and reinstall them with the new installers. Thanks.

PsychoSilence
01-25-2007, 06:15 PM
Box1 is causing some crashing (when using Group operators) and the Lock/Bond Test does not do what it is supposed to. If they are working fine for some, then I assume it tis something on my end and will try and remove the plugins and reinstall them with the new installers. Thanks.

in max9?

wanted to buy it today for the company i work atm. does anyone know if turbosquid has trouble with MASTERCARD (have no VISA available :( donīt wanna use my private credit card for obvious reasons)??? is there a way to buy it via paypal e.g.? i allready wrote to turbosquid help desk and orbaz.com sales...

kind regards

anselm

PsychoSilence
01-25-2007, 07:10 PM
On sunday we finished another project at soulpix (www.soulpix.de) so we got some cohibas and cuba libres :) was hell of a funny night :D

http://www.core2core.de/psychosilence_at_soulpix.jpg

DAAAAAAAAAAAAAAMN!!! I really look like sh*t...those were hard night shifts...unshaved, unstyled, tired to the bone and fat :D

PsychoSilence
01-26-2007, 06:28 PM
Just some quick and diry test for R&D:


http://www.core2core.de/spielwiese/fracture-moldy-test.jpg

viewport preview:
http://www.core2core.de/spielwiese/fracture-moldy-test.avi

work ahead:

- modeled debris
- better shape marks (blood splats e.g.)
- "spawn by travel distance" for blood spilling out of the big chunks...
- reactor involved for physics...


kind regards

anselm :)

superhypersam
01-26-2007, 07:34 PM
dude!

great test, reactor an p flow?

seems like more of a job for TP,
but maybe thats just cos I have bin in Tp world for the last month.

rdg
01-26-2007, 07:39 PM
hello
I'd like to post my experiment for this *effect*:

Splinter are scattered on a plane.
A magnet is floating above.
If the magnet comes close enough the splinters start trembling and align towards the magnet. If the magnet is too far away the splinters lay down.

Maybe there is an easy pflow for this - I haven't seen it.

So this is my setup:

http://proforma.preset.de/cgtalk/pFlow/magnet.jpg

rdg-asdf-particles_01.zip (http://proforma.preset.de/cgtalk/pFlow/rdg_asdf_particles_01.zip)

A SDeflector is hovering over the scene
Script Test 01 and Script Test 02 are testing if a particle is entering or leaving the SDeflector.
If the particle is inside the SDeflector Script Operator 01 is executed:
A tempLocation interpolates between the particle position and the center of the SDeflector and calculates a LookAt for the particle.
Script Test 01 also stores the actual distance Particle->Sdeflector in particleFloat, this way I get a 0-1-0 range.

There is some noise added to the tempLocation so the splinters can tremble.

Now I am confused how to get this happen with lying splinters.
I think I need to store the particleOrientation in particleVector and then align my tempLocation to the rotated lying particle. If the SDeflector - Particle distance shrinks I interpole to the center of SDeflector. If the distance grows the tempLocation needs to sink to particlePosition.z and the splinter is lying again, but looking at the SDeflector center.

I think I also need to update the particleVector with the new particleOrientation when the particle leaves the SDeflector.

So far so good.
But even watching Bobos Matrix over and over again.
This orientation transformation stuff drives me crazy.
:argh:

thank you for your patience.

Georg

PsychoSilence
01-26-2007, 08:04 PM
dude!

great test, reactor an p flow?

seems like more of a job for TP,
but maybe thats just cos I have bin in Tp world for the last month.

thanks bro! :)

iīm deep into TP research atm BUT the problem is that "big boss" had the great idea of switching to max9 :( no TP anymore no box#1 no nothing (at least AB and FUME allready work in max9)...iīll have to rework my particle pipe for the upcoming project :eek::twisted:

iīd love to imegrate to the TP world, believe me :(

btw: didnīt make it to assort the scripts for you but i will during weekend, i promise ;)


kind regards

anselm

rdg
01-27-2007, 09:54 AM
I found a workaround for my matrix-nightmare.

In the first event I place a Script Operator and for every particle I calculate a random point on a circle around .particlePosition. this is the first lookAt target for the Particle. this point is now stored in .particleVector and the next events can interpolate between this inital lookat and the magnet. (The inital lookat could also take place in a birth script)

rdg_asdf_particles_05.zip (http://proforma.preset.de/cgtalk/pFlow/rdg_asdf_particles_05.zip)

Next step will be:
Currently he particles are returning to their inital position if they are leaving the active range of the magnet. There should be a *switch* so those particles that are influenced enough will look at the magent until they leave the range. there new lookAt-position will be updated in .particleVector.

Georg

(http://proforma.preset.de/cgtalk/pFlow/rdg_asdf_particles_05.zip)

rdg
01-28-2007, 05:29 PM
magnet-preview.mov (http://www.preset.de/3dsMAX/magnetic/)

A preview of my magnet setup.

Georg

PsychoSilence
01-28-2007, 05:42 PM
neat!

i wish i was a better scripter :( but on the other hand you have a job that way around :D

plejboy
01-30-2007, 11:02 PM
Posted my reel over at:
http://forums.cgsociety.org/showthread.php?f=154&t=457919

I dont wanna push it on you guys but since its got a bunch of Pflow and the fact that I've picked up most of my knowledge around these forums I thought I might share.
Anyway, thanks for sharing guys!

PsychoSilence
01-31-2007, 09:21 AM
Gr8 reel mate! Already commented on it in your thread. :)

You should explain a little about the first shot and how it was done (the tree growth).
Maybe you can post a stripped assembly file?

cheers

anselm

DaForce
01-31-2007, 10:23 AM
Yeah what he said :)


wicked reel man.

Would love to see the guts behind the growth... ahh.. you know what i mean :D

SoLiTuDe
01-31-2007, 10:45 AM
:sad: and here I am reworking my demo reel... I just feel so ashamed now.

Great work on the reel!

PsychoSilence
02-01-2007, 01:28 AM
I worked a little on my fracturing test a few posts ago...
Originally i planed on doing the effect with Thinking Particles which is the right weapon of choice for VFX like that indeed...but thereīs still no TP for max 9...........so i have to get creative... the task is called "how to accomplish a VFX shot if someone took your balls" :D

http://core2core.de/spielwiese/frag-test-screens.jpg

hereīs a quick scanline render:

http://core2core.de/spielwiese/frag-test-basics-only.mov

donīt mind the colorful character. thats a special UVW-Check Map so i can see if everything is still in tact after passing it into new flow events etc.......

hereīs the flow:

http://core2core.de/spielwiese/frag-test.jpg

(really nothing you wouldnīt expect i suppose)

work ahead:

- STILL needs modeled debris
- better shape marks (blood splats e.g.)
- what i still have to add are PflowABs/Fume Events to pass some events over to AB / FUME
- (DROPED)"spawn by travel distance" for blood spilling out of the big chunks...
- (DROPED) reactor involved for physics...


kind regards

anselm

Now Winchester and/or Rayfire come into play :D

feldy
02-01-2007, 03:00 AM
its cool though question about that? was the man a particle the hole time he was walking?

PsychoSilence
02-01-2007, 09:44 AM
hy feldy!

no the char starts getting particle from frame 65 on. itīs pointcached file with animated visibility (frame 65 still visible, frame 66 zero visiblity, in the same frame the particles get visilbe with the same technique...). If you use thinking particles you can use bipeds and that kinda stuff easily. in flow itīs more of a challenge...

kind regards

anselm

PsychoSilence
02-01-2007, 11:16 AM
PFlow / Blobmesh bug in max9?

hy folks, just encountered a problem with flow and max9...maybe itīs reproducable. i would apprechaite if someone in here tests that issue with me on a separate mashine :)

do the following: create a simple flow of your choice with more then one event (maybe with some wind for chaos etc.) and add a blobmesh to the scene. then pick the flow as Blob Object. Now comes the buggy part: if you wanna select only a certain event for blob calcuation in the Particle Flow Parameters by hitting "Add" nothing happens... :(

is that error reproducable?!?

well i know blobmesh sucks but from time to time it does itīs job when thereīs no Glu at hand...

flowerpower! :arteest:

cheers

anselm

depleteD
02-02-2007, 04:46 AM
haha psycho, that was awesome. I think you have a weird force on the large fracture particles. Particles that large should go up and just fall.

great work nonetheless

PsychoSilence
02-02-2007, 05:17 AM
so true! the influence of the gravity is still default (1000,0 %) thatīs only a very rough preview stadium iīm in just to show the customer that iīm working on it right now...i had a preview deadline showing some significant effects of the commercial we do atm :(

thatīs one of those "customer calm down - previews"...in other words quick and dirty in other words it sucks :D

kind regards

anselm

depleteD
02-02-2007, 05:51 AM
AH I dig psycho, don't sweat it tho, you are kicking ass.


Well I have been chaseing after this for a long while now. I"m doing a tech test for a larger project right now that I will soon be completing.

With help from Brandon Davis I finally figured out how to relight atmospherics in post.

I've seen a couple people searching for how to do this aswell. If people would like I can do a video tut. My way not be the best and/or right way, but i think it works.

If anyone has any input into this process please tell me, i think my way is shady.

AtmosphericRelight (http://opticdump.tbxr.com/wp-content/uploads/2007/02/atmosphericrelight.mov)

Bercon
02-02-2007, 10:08 AM
PsychoSilence, yes its a bug. Blobmesh in max9 doesn't work with Pflow correctly.

http://forums.cgsociety.org/showthread.php?f=206&t=441415

rdg
02-02-2007, 10:18 AM
***
it looks like this problem is not pflow related.
***

please help finding the error why this pflow doesn't render.

I already added pCont.useTM, which was missing, but still the pflow doesn't update the particles at rendertime.

I combined my strange event-network into one birth script and one script operator.

update_question02.zip (http://proforma.preset.de/cgtalk/pFlow/magnet/update_question02.zip)

I tried to solve this without particles, but the interactive way is much more interesting.

Any suggestion would be highly apreciated!

Georg

oatz
02-02-2007, 07:15 PM
PsychoSilence, yes its a bug. Blobmesh in max9 doesn't work with Pflow correctly.

http://forums.cgsociety.org/showthread.php?f=206&t=441415

Psycho and Master - just wanted to let you know that while the Add button isn't functioning properly, you can still add events to the list using scripting. In the listener:

$Blobmesh01.pfEventList = #($'Event 01')

Event 01 is now in the list.

Khye

PsychoSilence
02-02-2007, 09:51 PM
Psycho and Master - just wanted to let you know that while the Add button isn't functioning properly, you can still add events to the list using scripting. In the listener:

$Blobmesh01.pfEventList = #($'Event 01')

Event 01 is now in the list.

Khye

great news! thatīs a workaround which simply shouldnīt be necessary upgrading to an update with that price policy but hey i deny myself from any further comment...i know why i stick to max8...

i should mention how much orbaz.com and oleg especially helped us to get a max9 pipe going. i wrote kinda like daily the last few days simply mourning and asking for stuff NOT working anymore...

iīm not much of a scripter but would like to help RDB with his issue simply becasue heīs a great TD and a friend of mine, maybe you other flowerpower guys could! i promise iīll do more scripting tuts :(

depleteD
02-04-2007, 05:19 AM
SO this is a test, im trying to replace the texture on particles in post. Unfortunatley, the uv data is occluded and I'm looking for a solution to this.

Can anyone help?

ALso, is there a script that can make the z face of the shape instance face the camera?

UVreplaceTest (http://opticdump.tbxr.com/wp-content/uploads/2007/02/uvtest.mov)

-Andrew

rdg
02-05-2007, 10:31 AM
Are you looking for the "Shape Facing" Operator?
This "creates a flat particle shape that is oriented towards the camera ..."

For the occlusion stuff:
I don't know exactly, but I would try differnet render passes.

Georg

loran
02-05-2007, 11:19 AM
depleteD
Is that what s you re looking for??


http://laurent.renaud.free.fr/divers/cgtalk/PF-evolution.jpg (http://laurent.renaud.free.fr/divers/cgtalk/PF-evolution.max)
max9 file (http://laurent.renaud.free.fr/divers/cgtalk/PF-evolution.max)

PsychoSilence
02-05-2007, 01:58 PM
nice one loran! :)

[shameless plug]
A new tutorial of mine was just released at CGArena.com
Project Setup/Management and Quality Assurance for Small Business (http://cgarena.com/freestuff/tutorials/max/projectSetup/index.html)

Maybe you might have a brief look...

thanks in advance

anselm
[/shameless plug]

ashrass99
02-05-2007, 05:39 PM
nice one loran! :)

[shameless plug]
A new tutorial of mine was just released at CGArena.com
Project Setup/Management and Quality Assurance for Small Business (http://cgarena.com/freestuff/tutorials/max/projectSetup/index.html)

[/shameless plug]

We posted this tutorial but this is coolest and informative tutorial available online after a long time. Anselm very well written.

rdg
02-06-2007, 08:39 AM
Regarding my "not rendering the pflow :cry:":
Bobo kindly reminded me to restudy the birthscript samples.

http://www.orbaz.com/forum/viewtopic.php?p=4957#4957

Georg

Glacierise
02-06-2007, 08:52 AM
Anselm this is awesome man! You're covering a very barren field in my education! Thanks!

PsychoSilence
02-06-2007, 11:39 AM
Anselm this is awesome man! You're covering a very barren field in my education! Thanks!

thanks for the flowers pal :)

PsychoSilence
02-06-2007, 09:04 PM
Just kinda stumbled over it:

"Sometimes" By PLEIX Films (http://www.pleix.net/movies/Sometimes.mov)
http://www.pleix.net/images/sometimes.jpg (http://www.pleix.net/movies/Sometimes.mov)

nice use of TP! (at least i guess so...)

cheers

anselm

SoLiTuDe
02-07-2007, 02:48 AM
http://forums.cgsociety.org/showthread.php?f=154&t=460755 <--Shameless Plug:):)

That "sometimes" video was way cool!

PsychoSilence
02-07-2007, 10:08 AM
http://forums.cgsociety.org/showthread.php?f=154&t=460755 <--Shameless Plug:):)

That "sometimes" video was way cool!

You rock, bro! Especially the robot and city scenes are kick 4$$!!!

Heiki
02-07-2007, 11:30 AM
Is it possible to scale particles by grayscale?

Darker areas = smaller particles, brighter = bigger.

theotheo
02-07-2007, 11:38 AM
Yeah it is, by using a speed by surface operator set to grayscale and a script operator,
Zhangy have made a scene with the script which you can find here : http://www.zhangy.com/main/index.php?module=documents&JAS_DocumentManager_op=viewDocument&JAS_Document_id=21

Cheers!

-the0

PsychoSilence
02-07-2007, 12:38 PM
Is it possible to scale particles by grayscale?

Darker areas = smaller particles, brighter = bigger.

read this thread in addition:
http://forums.cgsociety.org/showthread.php?f=206&t=455092

cheers

anselm

Heiki
02-08-2007, 10:14 AM
Thanx! I got it working!

Menesis
02-08-2007, 08:43 PM
Ok, here are the omni lights and particle flow example files.

http://support.discreet.com/webboard/wbpx.dll/~3dsmax/upload/PF%5FScripted%5FLights.mov (http://support.discreet.com/webboard/wbpx.dll/%7E3dsmax/upload/PF%5FScripted%5FLights.mov)

http://support.discreet.com/webboard/wbpx.dll/~3dsmax/upload/PF%5FScripted%20Lights%20Example%5F01.zip (http://support.discreet.com/webboard/wbpx.dll/%7E3dsmax/upload/PF%5FScripted%20Lights%20Example%5F01.zip)

CheersI was looking for a way to let particle flow create lights and I got to the post I quoted. Unfortunately the links are broken! Can somebody please help me?
I just want to be able to 'shape-instance' an omni instead of mesh-objects.

edit: I think I found a nice example somewhere else in this (HUGE!) topic :)

PsychoSilence
02-09-2007, 01:38 PM
I was looking for a way to let particle flow create lights and I got to the post I quoted. Unfortunately the links are broken! Can somebody please help me?
I just want to be able to 'shape-instance' an omni instead of mesh-objects.

edit: I think I found a nice example somewhere else in this (HUGE!) topic :)

controlling light position with flow...

http://www.orbaz.com/forum/viewtopic.php?t=631

http://core2core.de/spielwiese/research/PFlow_Fake_GI2.jpg (http://www.orbaz.com/forum/viewtopic.php?t=631)

cheers

anselm

Menesis
02-11-2007, 03:25 PM
Many thanks! I'm going to try it in a few days! :thumbsup:

rdg
02-16-2007, 07:15 PM
I don't know if something like this already posted in this thread.
This is a 'hommage' to the 'FoRest gump' challenge at odforce.net [1].
I stumbled across this and as my bindings to 3dsmax are stronger than I would like to admit I tried it at home.


http://proforma.preset.de/cgtalk/pFlow/controlledByMesh/controlledByMesh-sm.jpg

A low res mesh controlls the distribution and properties of objects.
Movie: controlledByMesh (http://proforma.preset.de/cgtalk/pFlow/controlledByMesh/)

Georg



[1] http://forums.odforce.net/index.php?showtopic=78

PsychoSilence
02-16-2007, 11:30 PM
great solution!

saw that allready in DEV ;)

hereīs some minor crappy updates in the disintegration stuff iīm involed atm :D...bloodstreams and stuff:
http://core2core.de/spielwiese/hollow-man-disintegration.mov

"hollow man movie" style "under the skin" dissolve stuff...inspired by http://www.sidefx.com/index.php?option=com_content&task=view&id=496&Itemid=68

kind regards

anselm

Anesthaesia
02-19-2007, 09:45 AM
Hi All,

I've recently started looking at PFlow, and have looked at Allan McKay's tutorials which are really impressive.

What I would like to know is how does he get so many particles in his tutorials? I seem to start having problems when I get to around 100k on screen...

His dispersion tutorial especially shows a very wispy particle flow - is this a trick or simply loads of particles? I have used caching but it eats through an incredible amount of memory.

The example I was referring to is the image below:

CapitanRed
02-19-2007, 09:49 AM
Use facing particles with self illumination material, and no light. this way you can render about a million particles ;)

cojoMan
02-19-2007, 10:56 AM
hi guys...I have been watching this thread for some time now...great work everyone...

I have a problem...i want to have some volumetric clouds that I can manipulate...like...I want to add a vortex space warp and I want it to affect the closest parts of the cloud...and like...if an object...say a comet passes through, it ripples through the clouds....making a hole in them...know what I mean ?

can u help ? 10x in advance...


I don't even know how to approacj this...PF ?...afterburn ? other ?...

PsychoSilence
02-19-2007, 11:31 AM
hi guys...I have been watching this thread for some time now...great work everyone...

I have a problem...i want to have some volumetric clouds that I can manipulate...like...I want to add a vortex space warp and I want it to affect the closest parts of the cloud...and like...if an object...say a comet passes through, it ripples through the clouds....making a hole in them...know what I mean ?

can u help ? 10x in advance...


I don't even know how to approacj this...PF ?...afterburn ? other ?...

check afterburnīs void demon :) thereīs a tutorial file shipping with it...


EDIT: Some quick update on my disintegration approach....
http://core2core.de/spielwiese/hollow-man-disintegration.mov
&
http://core2core.de/spielwiese/skeleton-disintegration.wmv

kind regards

anselm

a13xr3d
02-19-2007, 07:31 PM
Hey, for your disintigrating effect, do you just have all of the particles on the mesh itself?

What would be a the best way to have that same effect, but depth wise, like the veins going inside the head, with a fractal-tree-vein-type thing?

PsychoSilence
02-19-2007, 09:14 PM
well at least i guess that would be easy to achieve...

create a smaller object inside the bigger objectīs center and then spawn particles on the inner core leading the spawned by travel distance particles into a find target. within the find target make sure mesh select is checked and then pick the bigger object...havenīt tried that but should work...

kind regards

anselm

nitrocom
02-20-2007, 12:21 AM
I m speechless right now... I cant believe that i missed those kind of source... Hi guys :)

Its starting to be boring that hanging around the same thing all the time now i can see something else :)

Well great, Im sticked here :)
:thumbsup:

erilaz
02-20-2007, 01:25 AM
Hi guys. This is a tonne of useful information, with the slight draw back of it being all in one huge thread. What I would like to suggest is we start to break the useful tidbits into the CGWiki Particle Flow section (http://wiki.cgsociety.org/index.php/3ds_Max/Particle_Flow). I'm suggesting the same thing for the mental ray thread as it's getting cumbersome also.

PsychoSilence
02-20-2007, 02:06 AM
@ nitrocom: welcome aboard! :)

@ erilaz: youīre right, itīs just a huge pool of unsorted information. BUT itīs such comfortable to have everything in one place...if iīm in desperate need of something i use the search on this thread > here we go! as simple as 1.2.3...do re mi...simply because this thread is an institution over here...
Iīd be always down with a wiki on pflow and would feed it of cause but here i feel comfortable...

But thatīs just my personal opinion...

kind regards

anselm

erilaz
02-20-2007, 02:19 AM
@ erilaz: youīre right, itīs just a huge pool of unsorted information. BUT itīs such comfortable to have everything in one place...if iīm in desperate need of something i use the search on this thread > here we go! as simple as 1.2.3...do re mi...simply because this thread is an institution over here...
Iīd be always down with a wiki on pflow and would feed it of cause but here i feel comfortable...


Oh absolutely, and I wouldn't dream of wanting the awesome level of content discussion in these threads to stop! I just thought a bit of nugget-digging might be nice for the head-scratchers out there as a sort of parallel "digest" so to speak.:)

nitrocom
02-20-2007, 11:46 AM
Hmm let me say something about a script...

"

on ChannelsUsed pCont do
(
pCont.useTM = true
)

on Init pCont do
(
global Lampen = $lampe*
My_Lights.pos = [0,0,0]

)

on Proceed pCont do
(
pCount=pCont.NumParticles()
count=amin #(pCount,Lampen.count)
for i in 1 to count do
(
pCont.particleIndex = i
Lampen[i].transform=pCont.ParticleTM

)
)

on Release pCont do
(

)

"

This is the script that written by PsychoSilence however it gave me error like "The script has errors. Close the script"

So is there anything that i made wrong?

PsychoSilence
02-20-2007, 12:29 PM
Dear nitrocom,
i think i know why itīs creating errors >> Because iīm fuc***g stupid from time to time :D i allready changed the script in the Orbaz.com Thread...


that should work:


on ChannelsUsed pCont do
(
pCont.useTM = true
)

on Init pCont do
(
global Lampen = $Lampe*


Lampen.pos = [0,0,0]

)

on Proceed pCont do
(
pCount=pCont.NumParticles()
count=amin #(pCount,Lampen.count)
for i in 1 to count do
(
pCont.particleIndex = i
Lampen[i].transform=pCont.ParticleTM

)
)

on Release pCont do
(

)

Note that you can replace "Lampe*" with the name of your lights and "Lampen" with a Global you prefer...

see my setup at the attachment...thatīs hell of a difficult flow ;)

nice would be a script that creates as much lights as there are particles. at the moment you have to spawn as much particles as there are lights. maybe a scripter could automate that process...

kind regards

PsychoSilence
02-20-2007, 10:54 PM
http://core2core.de/spielwiese/stadium_PFlow.jpg

I just found an older image back from july 2006 when we did one of the ASIA GAMES spots at BlackMoutain VFX in Stuttgart/Germany. Thatīs the Al-Sadd-Stadion in Doha/Katar modeled by Markus Hund who should lurk around here as "CERBERUS" from time to time. I post it because itīs the basic reason why i did my stadium tutorial over there at CGArena.com. EVERY SINGLE chair in the stadium is PFlow...over 44,000 in total if i remeber right...

notice that the flow builds the flag of Katar http://www.ipicture.de/4images/data/thumbnails/101/flagge_katar.gif in the upper right corner of the image. i achieved that with a few group selects and different materials...6 different types of chairs and shaders in total. all one single PFlow system........

kind regards

anselm

Bobo
02-20-2007, 11:26 PM
Note that you can replace "Lampe*" with the name of your lights and "Lampen" with a Global you prefer...


You can just put in $Omni* because most people with no scripting knowledge just create a bunch of Omnis and then wonder why your scripts does not work ;)

PsychoSilence
02-21-2007, 08:46 AM
So true!
i did that in the example attachment...i wish some GOOD scripter so thats excluding me :D would automate that by writing a script that automaticly spawns as many particles as there are lights. maybe with a little spinner as MiniUI how many and which type of ligts are desired. i should suggets that as a scripting challenge maybe? "Particle Light & Magic" :D

kind regards

anselm

rdg
02-21-2007, 11:03 AM
So true!
automate that by writing a script that automaticly spawns as many particles as there are lights.

I was a huge fan of cebas' proOptic suite (?) which had a particleLigh option. nowadays this is part of the lumaObjects (?) - who knows.

well, on my endless quest for becomming a TD (http://www.youtube.com/watch?v=nti85714PVQ) I tried this:

I added an custom attribute to the pflow-source (using this script: light_ca.ms (http://proforma.preset.de/cgtalk/pFlow/particleLight/light_ca.ms))
You can pick an omni and the ca generates the desired number of copies. The lights are stored in a nodeTab - and will be saved with the scene.
In the flow the is a Script Operator "Let there be light". I switches the active lights on and off and places them at the particle positions.
If there are more particles than lights, the lights are attached to the first particles that are born.

Download particleLights_04.zip (http://proforma.preset.de/cgtalk/pFlow/particleLight/particleLights_04.zip)

***Note this is just an quick and dirty R&D.
I guess you can add any operator before (!) the "let there be light"-operator.
I don't think that the lights will follow to other events as particleIDs are changing between events. No problem if all lights are the same, but if you add variations based on life time and other stuff ...
I guess one needs box#3 to get a setup like this really bullet proof.
You'll find my address on my homepage if you want to send me a copy ;)

Stuff learned today:
[1] customattributes as pFlow interfaces
[2] nodeTabs aren't that mysterious
[3] max sends error reports after closing this scene ... ???

Georg

PsychoSilence
02-21-2007, 11:37 AM
I was a huge fan of cebas' proOptic suite (?) which had a particleLigh option. nowadays this is part of the lumaObjects (?) - who knows.

well, on my endless quest for becomming a TD (http://www.youtube.com/watch?v=nti85714PVQ) I tried this:

I added an custom attribute to the pflow-source (using this script: light_ca.ms (http://proforma.preset.de/cgtalk/pFlow/particleLight/light_ca.ms))
You can pick an omni and the ca generates the desired number of copies. The lights are stored in a nodeTab - and will be saved with the scene.
In the flow the is a Script Operator "Let there be light". I switches the active lights on and off and places them at the particle positions.
If there are more particles than lights, the lights are attached to the first particles that are born.

Download particleLights_04.zip (http://proforma.preset.de/cgtalk/pFlow/particleLight/particleLights_04.zip)

***Note this is just an quick and dirty R&D.
I guess you can add any operator before (!) the "let there be light"-operator.
I don't think that the lights will follow to other events as particleIDs are changing between events. No problem if all lights are the same, but if you add variations based on life time and other stuff ...
I guess one needs box#3 to get a setup like this really bullet proof.
You'll find my address on my homepage if you want to send me a copy ;)

Stuff learned today:
[1] customattributes as pFlow interfaces
[2] nodeTabs aren't that mysterious
[3] max sends error reports after closing this scene ... ???

Georg

*instantly gives Georg a hug in a not that gay manner*

theotheo
02-21-2007, 01:05 PM
It would be really cool of you could create UI controls for the pflow scripts and expose certain parameters from the script to the script operator, so that you may create a interface for the script, kinda like expose in box:3. I guess this could be done with customattributes to scene objects, but it would be cool to have it as an option to do it on pflow operators aswell. Any thoughts ? Or just a bad idea?

-theo

rdg
02-21-2007, 07:30 PM
Re: Particle Light & Magic
The particle light setup can transfer the particles to other events and the lights seem to follow. Here are some stills of my test animation:

http://proforma.preset.de/cgtalk/pFlow/particleLight/fairies01-.jpg
http://proforma.preset.de/cgtalk/pFlow/particleLight/fairies03-.jpg
http://proforma.preset.de/cgtalk/pFlow/particleLight/fairies04-.jpg

The last one is a fail frame where the lights get scaled with the pixie dust ...

Georg

PexElroy
02-22-2007, 07:13 PM
PsychoSilence >>

Was wondering if you may go into more depth on how this R&D was created http://www.core2core.de/spielwiese/fracture-moldy-test.avi) and how you got PFlow to work with reactor in this way.

rdg
02-22-2007, 08:26 PM
I like my coffee black just like my metal.
Anyway I have chosen a more 'dancy' sountrack for the pflow light sample implementation:

http://www.count-as-one.org/


It looks like a


theLightStore[pCont.particleID].transform = pCont.particleTM


produces more stable results as relying on pCont.particleIndex.

If the 'Let there be light' Operator is copied to other events, you have to comment/remove these lines:

for i = 1 to theLightStore.count do (
theLightStore[i].enabled = false
)
or the lights will go dark for one frame.

I used a script from allan mckay's rigidbody tutorial to bake the lights after sim.

Georg

PsychoSilence
02-23-2007, 11:15 AM
PsychoSilence >>

Was wondering if you may go into more depth on how this R&D was created http://www.core2core.de/spielwiese/fracture-moldy-test.avi) and how you got PFlow to work with reactor in this way.

Dear Robert,

this particular preview is still without reactor. it was an early stadium of this test: http://core2core.de/spielwiese/frag-test-basics-only.mov hereīs the flow: http://core2core.de/spielwiese/frag-test.jpg
the flow you refer on is the one on the left with different fragments only. i decided to split ground splats and fractures to have more control over everything...
Donīt mind the strange behavior of the fragement during mid-air. this is all done without ANY tweaking. still default forces and stuff thatīs why they fall so fast...i was a lazy bone i admit! this was a "customer calm down - preview" as we call it here to show that weīre working on everything and bla......

Allan and Loran have nice tutorials regarding the fracture topic as well:
http://www.intrinsia.net/amckay/tuts/test.swf
http://laurent.renaud.free.fr/tuto01-uk.html

i would have looooooved to use thinking particles for that simply because itīs fracturing and collision solvers would fit perfect for that kinda shots but we donīt have it here so had to go into deep R&D to make the magic happen without a magic wand

but we dropped the reactor approach already due to a lag of control. Now we use XSI6 soft bodies for the explosion shot because the solvers are faster and way better to control. Now we have nice soft bodies which react like real meat hitting the ground (nice wiggling on collision and stuff ).

Our TD Kai Wolter ( http://www.kai-wolter.de ) wrote a Cache format (quite similar to the Point Oven for XSI) that connects all the three worlds: XSI, MAYA & MAX in any direction. So we can use what ever getīs the job done and pass the scene afterwards...The animators work in maya, pass any dynamic related work to the XSI guy for solving and then i get cached geometries for any particle and volumetric work (Flow, FumeFX, maya particles & fluids eventually, etc).

I would like to point out again how fast and capable the orbaz.com support helped me out with max 9 issues i already ran into. oleg will get an extra fluffy cloud in heaven :D

hope that cleared the fog maybe.

kind regards

anselm

PexElroy
02-23-2007, 07:10 PM
PsychoSilence, thanks a lot for the details and links. Looking for a stable and accurate dynamics particle solution in max, but it seems we'll want to wait for Box #2 or get TP 2.5, but prior work-arounds are good to know about.

PsychoSilence
02-23-2007, 07:31 PM
thinking particles offers really great fracturing and collision solutions! best for dynamics. you should check it definitely out! it a real allrounder.

rigid bodies and flow is more of a hassle :( because you have to wrap them into a mesher compound which is dead slooow when you pass it over to reactor rigids :(

cheers

anselm

CapitanRed
02-25-2007, 12:47 PM
Does somebody know if the find target operator(if the target position is controled by script) also can affect particleOrientation?
I Used the script from allanmckays tutorial, and tried to add particle Orientation, but as I am not a scripter and have no Idea if it should or even could work, I ask you guys ;)
the script I added is green here, because it didn't work with it.



on ChannelsUsed pCont do
(
pCont.useVector = true
pCont.usePosition = true
-- pCont.useOrientation=true
)

on Init pCont do
(

)
on Proceed pCont do
(
count = pCont.NumParticles()
Global PRB = $Fragment_* as array
for i in 1 to count do
(

pCont.particleIndex = i
if currenttime <= 1 then pCont.ParticlePosition = PRB[i].position

pCont.particleVector = pCont.ParticlePosition
print pCont.particleVector as string
-- print pCont.particleOrientation as a string
)
)
on Release pCont do
(
)

rdg
02-25-2007, 04:38 PM
captainRed,

after enabling the pCont.particleOrientation = true, you need to set the orientation in the onProceed part to some value.

If I read your script right, you just printed the orientation?

I had best results with pCont.particleTM where I used a matrixFromNormal to set the orientation for the particle.

pcont.particleTM = matrixFromNormal (normalize (Postion1 - Position2))
where Postion1 and Position2 are Positions to get the desired particle Orientation,.
after that you have to recopy the pCont.particlePosition onto the paricle als matrixFromNormal moves the particle to the Origin [0, 0, 0]

maybe you can post a stripped down szene and I can take a look.

Georg

CapitanRed
02-25-2007, 07:01 PM
Hey, very nice you take the time to help me. this is a testfile for me, just to understand a little the scripting in max. but i really have difficulties, because i have no idea about it. (all script I know is Actionscript, and there also, i have not a really good understanding ;) )

ok, the Idea is this:

the sphere falls to the ground at the left side and gets fractured. then, at frame 60, every fragment gets a particle (script FRAGMENT_POSITION). and then, in the next event the particle starts to controle the fragments(script MOVE_THE_FRAGMENT).
this script, which makes the particles drag the fragments are turned off now. please, take a look inside before activating them. the idea is that the sphere gets rebuild.

as soon as you turn on all these (scripts MOVE_THE_FRAGMENT), and you drag the timeslider, the fragments will move to the right position, BUT if you go back to fram 0, the animation will not work anymore. then you have to open the file again ;)
for this, I will bake the animation, as soon it will be as I want.

What I want is to get the rotation of the particles to the Fragments, so the sphere is rebuilt at the end.

so, enough talked now....hre is the file :D

EDIT:---------

there are some little mistakes in the scene, which could confuse you, but they aren't really important. one fragment is not doing the animation. this is the reason why one particle is a the right side, and does not do the right thing. but this is my mistake, and has nothing to do with the P-System itself. and another thing is, that the fragments are not lying on the floor. I think the collision tolerance is the reason for that.

rdg
02-26-2007, 07:43 AM
CapitanRed,

I must admit that I don't understand the file.
But I would start simplifiing it and then add step after step.

If you store the whole transformation matrix in the particleMatrix you have the final position and orientation stored in the particle. I haven't found a way to set a rotation/orientation via 'go to rotation' and a script vector. but I know about script wiring only a few days.

I might end up in scripting this from start till finish, as I often don't know native pflow tricks. As usual I guess box#3 is the smartest answer :) I guess I will have to get it, as I often try to fake its bahavior with scripts and get mad trying it.

Georg

CapitanRed
02-26-2007, 08:29 AM
Yea, the file isn't really "the yellow of the egg" ;) lol

I'm just looking for a way doing this without any plugin. I think I was trying to do it on a complicated way, as the solution coul'd be simplier. the goToRotation Operator is a good input. never tought about it :p
I'll try it, and if it works, I'll have to bang my head :banghead:

PsychoSilence
02-26-2007, 10:44 AM
http://core2core.de/spielwiese/schimmel-disintegration.jpg (http://core2core.de/spielwiese/schimmel-disintegration.mov)

some disintegration test i did today... (http://core2core.de/spielwiese/schimmel-disintegration.mov)

this is kind of outer competition again...just did some tests for our current project today...please view that in super slowmo scrubbing through the mov to see all the details. i stuffed all the disintegration into 100 frames due to render time...

the final effect will be rendered with at least 4 layers of particles with differend random seed on the position object so there will be no see-throughs anymore :)

hell! i wish i already had mocap data for further testings :(

PS: The skeleton "explosion" will be accomplished in XSI so that is only a "replacement effect". take no further notice on that :wise:

kind regards

anselm

Glacierise
03-01-2007, 09:44 PM
First, nice disintegration :) And the bones look cool here, too. The ash could use some turbulence, though :D

Second: I'm back!!! Hi again to the coolest band 'round here :) Been away for a while, some project with Vue (stomps ass), but I'm back to sweet pflow again :D Contributed some stuff in the main forum, go see it ;)

PsychoSilence
03-01-2007, 10:19 PM
I'm back to sweet pflow again :D

Good to have you back! :) :bounce::buttrock:

yeah! some turbulence and of cause more speed / gravity variaition...but itīs not smart to go on further tests without the final camera and animatin data...i donīt even know if we see the ash hitting the ground...maybe itīs gonig to be some "western" camera angle...

kind regards

anselm

CapitanRed
03-03-2007, 11:58 PM
Some success in breaking and rebuilding the sphere...

www.megalomania.ch/capitanred/movies/rebuilt.avi (http://www.megalomania.ch/capitanred/movies/rebuilt.avi)

but as you see, there is some weird motion. don't know why. I hope i'll find a way to get that right.

also the rebuilting is little faked. the fragments turn into right position at last frame.( as soon as the particles reach their target.
Does somebody know where from the find_target operator takes the positions? i mean, how does the script look like. i could only find the functions, which are allready in the operator.
what i am looking for is how the whole find target is built up.(especially the part, where you put your targets inside. if the operator is building an array, and sending the particles to these positions, then it should be possible to send them also to the target transformation matrices....i hope ;) )

EDIT___________
mmmhhh....my host is down!!! damn! will upload is as soo as possible ;)

SoLiTuDe
03-04-2007, 12:13 AM
Captain Red... allan mckay has a tutorial on his site dealing with how the find target "pops" into place, and uses a really simple way to overcome this...

http://www.allanmckay.com/tut/tut_ObjAB.avi

depleteD
03-04-2007, 01:14 AM
CapitanRed Hey cap, the trick you need is to find target by script vector. I think you need box 3 to store multiple vector channels tho. I'm not positive. But you need 1 vector for position and another vector for rotation. Maybe you could convert the scale channel in the rotation channel through a script.

Just an idea, hopefully it helps you man.

-Andrew

CapitanRed
03-04-2007, 06:26 AM
solitude:

allans script is allready in use ;), but it saves only the position. and find target is only reading vector for position i think.

but now, in my scene, all chucks get to the right pos and rotation. it is just no smooth motion :p
check out the link. it should work now :)

depleteD:

can you explain further? how do you mean use scale for rotation?

depleteD
03-06-2007, 06:28 PM
Cap- Well its just useing the scale as a place holder for your variables. Its just theory tho, I'm not to sure, assembly is a quite interesting problem.

I think the key is to store the initial rotation of the particles, then have a go to rotation test, and call the initial rotation value. Just like calling a position through a find target by script vector. Particle FLow only has a set number of data channels, such as age(integer), position(vector), scale(vector) ect

I don't know how to use BOX 3 but i'm sure that it will give you the ability to do this. I remember reading about inifinte data channels for your particles.

I'm very unsure about this. I might be completly wrong. So take my advice with a grain of salt.

I have a project coming up that I will have to use assembly. I will have to do some R/D to figure it out, I will share my results with you.

-Andrew

CapitanRed
03-07-2007, 07:17 AM
I also tought about the go to rotation op. but coudn't find a method how to assign scripted coordinates to it.
also don't know about box 3 tools :blush:

the only way I see now is to calculte the difference between start and end TM of the particles. then, divide that trough the frame count of the animation, and then rotate the particles for that value each frame.

the problems are that the particles will rotate linear. there will be no easing. and i'm not sure, if you can apply a find target operator, and let them spin the way you want, till they find the target.

rdg
03-07-2007, 08:51 AM
The particles will rotate linear. there will be no easing

While you are on the scripting road - here are some easing functions:
http://www.gizma.com/easing/

Georg

PsychoSilence
03-08-2007, 06:27 PM
slightly updated my website after a long time if anyone is interested...

www.3Delicious.de

cheers

anselm

CapitanRed
03-08-2007, 07:06 PM
Nice psdMaterial thing there. did you write it? is it only for your usage? :D

rdg
03-09-2007, 09:03 AM
Nice psdMaterial thing there. did you write it? is it only for your usage? :D

I wrote it and Anselm is counselor and the-one-and-only-beta-tester (Is he? Maybe he established 0-day-ddl-script archive with my rdg::AUS-scripts - we will have to google this ...)
It is based on psd2material by Wahooney but I don't think there is one original line of code contained. So it is more a "hommage to psd2material".

It features the generation of materials based on psd files for both scanline and vray with optional autodetection - MR support would require some tweaking.
You can configure the layernames with a textfile to localise it and/or adapt it to your pipeline.
The script extracts the layers of the psd to png/tif/whatever and keeps optional backups of changed layers.

There are one/two/? Photoshop scripts that gernerate layered templates for scanline/vray setups. The reverse way material2PSD would also available if somebody were using PS CS2. But it looks like everybody is still using CS1 and/ 7.

It is 99% finished - it is just missing the documentation ... In the meantime I learned a lot and propably would like to add xml support for the config files - what is totally useless as it works fine the way it is.

I was thinking about realeasing it as gift/postcard ware, but don't have the time to support it or answer questions like "why doesn't behave it like I would do it if I could do it".

So it will either be my dowry to the shop that hires me, or ananother-useless-script release.

Georg

PsychoSilence
03-09-2007, 09:20 AM
It is 99% finished - it is just missing the documentation ...

Georg

*instantly keeps on writing the doku...* ;)

Phibmobil
03-12-2007, 06:14 PM
Hi all.

Iīm hitting memory limits when trying to put one animated material with different timing offsets on many (20+) facing particles. It seems max starts caching the moment i check the 'sync frames to particle age' box. Are there any workarounds for this ?

Any help please =)

!phib

Howard Day
03-12-2007, 06:19 PM
To the best of my knowledge it's unavoidable. However, you can minimize it by having your frame slider at 0 when you check that box, and by also keeping your material editor closed when you're scrubbing thru frames.
That's about the extent of the damage control I've been able to figure out on this one.

Phibmobil
03-13-2007, 10:42 AM
thats a shame.. Iīll use normal planes then, and script an offset mat. maker.. TP also seems to render rather slow when i comes to this kind of setup.

Thanks for the tips .. and keep this thread alive !!

PsychoSilence
03-13-2007, 11:43 AM
.. and keep this thread alive !!

we will!!! :) last thing iīd do is to neglect my favorite thread and community! My alibi is that i have some spare time since two weeks and i DONīT spend that in front of any screen...iīm more down with a beer,bitches,barbecue-attitude :D

but i will hang in here when iīm back on track, i promise ;)

Mir-Vadim updated his RayFire tool along with RayExplosion! "off the hook" is my only comment on that update...
http://forums.cgsociety.org/showthread.php?p=4253971#post4253971

cheers

anselm

relief7
03-14-2007, 02:56 AM
Hey guys sorry for jumping right into the discussion. I have a small question which involves a little bit of math, I guess.

I want to use the orientation of an object in a scripted operator. I would like to gradually rotate my particles to a final orientation value. Now I have the choice: TM channel or orientation channel. If I use the TM channel (+go to rotation operator) everything works fine but I always need to temporarily save the original particle position and write it back after applying the target object's TM to the particle, which is kind of ugly. It also makes problems when using scale as well so i would like to treat everything as separate: translation, orientation and scale and control it with the corresponding channel.

Does anybody know how to generate an orientation vector out of an object's matrix so I could use the orientation channel directly ?

And... how do I properly rotate an object about its local axises ? Is it by multiplying its matrix with the inverse, the rotationMatrix and then the original Matrix ? I could not get it to work so far as my math is a little while back. Hehe ! Would be really thankful if anyone could help out. :)

ahmedsheeraz
03-17-2007, 05:20 AM
Hi all, I am watching this thread right from the day when it was started and I must say that I learned a lot. btw this is my first post here.

my question is regarding pflow Box1 which had a 'particle position +' operator. but now I installed the new demo on 3ds max 9 n it don't hv this operator. did they change this operator with some other operator?

wat I want to do is to place particle on each vertex on my mesh which is not possible through built in position operator.

charleyc
03-17-2007, 05:43 AM
ahmedsheeraz - I believe PositionObject+ was actually part of the freebies (or at least a free dll somewhere on the Orbaz forums) and not actually part of Box1. That being said, there is no PositionObject+ for R9 as of yet. There is a thread on the Orbaz forums about this operator, basically it will be released as an upgrade to the standard PositionObject op (adding the + functionality). However, you can get a particle per vertex if you birth exactly the number of particles to the number of verts and use the Separation threshold of a standard PositionObject operator. The key to this method is making sure that the Separation Distance is less than the smallest edge length of the mesh you are using. Generally a setting of .1 or .01 will suffice.

PsychoSilence
03-17-2007, 11:14 AM
Hi all, I am watching this thread right from the day when it was started and I must say that I learned a lot. btw this is my first post here.

my question is regarding pflow Box1 which had a 'particle position +' operator. but now I installed the new demo on 3ds max 9 n it don't hv this operator. did they change this operator with some other operator?

wat I want to do is to place particle on each vertex on my mesh which is not possible through built in position operator.

what you are looking for is a birthscript that spawns and places a particle per vertex:

on ChannelsUsed pCont do
(
pCont.useposition = true
)

on Init pCont do
(
global pflowguide = $guide
)

on Proceed pCont do
(
if pflowguide != undefined do
(
t = pCont.getTimeStart() as float
NumVerts = getNumVerts pflowguide

if t < 0 do
(
for i = 1 to NumVerts do
(
pCont.AddParticle()
pCont.particleIndex = pCont.NumParticles()
pCont.particleposition = polyop.getVert pflowguide i
)
)
)

)

on Release pCont do ()

note that the line " global pflowguide = $guide" sets the geometry you want do spawn on. just rename "$guide" to the name of your geometry or name the geometry "guide" :)

kind regards

anselm

PsychoSilence
03-17-2007, 08:26 PM
Finaly encouraged by my team mates i dared to post my reel here :D Hope you enjoy my work...

http://core2core.de/flash/showreel.jpg (http://core2core.de/flash/gallery/3Delicious_Showreel_2006-2.mov)


http://forums.cgsociety.org/showthread.php?f=154&t=475184

ahmedsheeraz
03-18-2007, 09:31 AM
thanx, charleyc and PsychoSilence. So scripting is da answer of all probs. :) . Actually I know this script already but why write a script when u hv an operator for that task right in ur shelf.anywayz thanx and I will go with scripting. PsychoSilence, it's good to c ur showreel packaging. ;)

amckay
03-21-2007, 12:46 PM
Hi mate, great work!! very cool reel keep up the great work

Neejoh
03-28-2007, 06:24 PM
Damn, it's been a while! Great to see the thread is still alive. Haven't had time to check everything out since when I was last been here, but damn PsychoSilence! That shit looks great mate :)

PsychoSilence
03-28-2007, 07:29 PM
thanks bros! glad you like my style. hopefully i can find some time to create a "vfx only reel" along with the shading related stuff. ffx,flow,box#3,AB,krakatoa...stuff like that...

havenīt had time to be active alot the last few weeks but i will work on that ;) i was asked to write a tutorial on my fracturing thingy a few pages earlier. maybe iīll do so next week!

keep in touch guys! flower power!!!

kind regards

anselm

did i mention that i bought box#3 a few days ago?!? it currently reveals to me in all itīs greatness BUT i still donīt have a clue :D so get ready for some dumb ass questions by me in the near future ;)...iīll watch boboīs matrix explained dvd first to go more deeper into scripting and dealing with vectors and stuff.

PsychoSilence
03-29-2007, 05:24 PM
Fracturing taken further with box#3?

After i checked out the collision fake tutorial file i was asking myself the following: if we roughly know how big a particle is and where it's neighbour is in worldspace right now wouldn't it be possible to use this information for taking fracturing further?

I don't wanna deal with the per-fractring process itself that much because there are serveral really cool scripts around already. My personal point of interest is the handling of fractures and fracturing within flow because there is no real fracture module yet. "Let's work that out2 as RIF from CGFluids/CafeFX might say :) maybe we can ease some pain...


I would like to invite all box#3 users to go on a little virtual off-site with me to see if it can be done.


Here's a rough list of what's on my mind right now:

- if we would know what diameter a chunk has would it be possible to control if it's small enough to fracture? My idea is that only small enough pieces fracture on collision. Big chunks stay in shape

- If a piece is small enough to fracture on collision it shouldn't break into bigger parts then itself because this looks unrealistic!

- If we know thanks to box#3 how far each particle is from the next one would it be possible to create some interpenetration avoidance module?

- Bobo has a tutorial on his website dealing with particle size as mass. Would it be possible to build that in? So that smaller bits and pieces are influenced by more by forces then heavy boulder like pieces.

Maybe box#3 artists and scripting artists could work something out together? First we check out what's possible based on our knownledge about pflow/maxscript then we get it going!


I collected some usefull information already to have everThing in one post:

- allan's PRB tutorial where he links geometry to the particle position via maxscript to pipe that geo into reactor:
http://www.intrinsia.net/amckay/tuts/_amckay_PRB.avi (http://www.charleycarlat.com/amckay/tuts/amckay_PRB.avi)


- bobo's size as mass tutorial
http://www.scriptspot.com/bobo/mxs5/pflow/pflow__force_by_mass.htm

- orbaz.com box#3 discussion
http://www.orbaz.com/forum/viewforum.php?f=32&sid=5682bb57cb929ffd572add8ed35c7a45

Anyone interested in such R&D?


Kind regards

anselm

oatz
03-29-2007, 05:36 PM
F
- Bobo has a tutorial on his website dealing with particle size as mass. Would it be possible to build that in? So that smaller bits and pieces are influenced by more by forces then heavy boulder like pieces.

Box 3 has Input Standard : Scale and Output Standard Script: Float. Were you looking to just create an operator that would replace the script?

I think you're idea of fracturing smaller pieces than the parent piece is a good one and should be doable with Shape: Volume and Shape: Surface.

Khye

PsychoSilence
03-29-2007, 05:56 PM
i thought even bigger...maybe we all can throw a new tool set together making things easier!

charleyc
03-29-2007, 07:02 PM
This would be a good topic for Oleg to get in on as I imagine Box2 might deal with some of this :D

First off I would like to say that Box3 does not have a great toolset for dealing with particle mesh directly. For something like inter-particle collision, you need to work off the instanced geometry and apply that to the particles in question. Doable, but not straight forward.

1 - You can control particles by size quite easily with box3, and in fact I would take this a bit further and deal with force of collision (speed) to determine whether or not fracture occurs. This seems pretty simple for Box3

2- This is as simple as reducing scale (which could be accurately calced so all the parts equal the whole) at the time of the split (most likely done with a spawn)

3- See above. I think it is possible, but not straight forward for accurate collisions. However, I think you could easily get a lot further ahead than a Keep Apart Op gets you.

4- As was already mentioned, this is extremely simple to do with Box3. Using the Script output, you can setup Script data of whatever you need and send that out to a Force Op.

One final note, I no longer have that particular tut on my site. I needed to free up some space, so the larger ones had to go. Sorry.

OlegB
03-29-2007, 07:14 PM
I'm omnipresent and always listening :)

Thanks,
Oleg B.

JohnnyRandom
03-29-2007, 08:01 PM
Nice reel Anselm :D

PsychoSilence
03-29-2007, 09:10 PM
This would be a good topic for Oleg to get in on as I imagine Box2 might deal with some of this :D

First off I would like to say that Box3 does not have a great toolset for dealing with particle mesh directly. For something like inter-particle collision, you need to work off the instanced geometry and apply that to the particles in question. Doable, but not straight forward.

1 - You can control particles by size quite easily with box3, and in fact I would take this a bit further and deal with force of collision (speed) to determine whether or not fracture occurs. This seems pretty simple for Box3

2- This is as simple as reducing scale (which could be accurately calced so all the parts equal the whole) at the time of the split (most likely done with a spawn)

3- See above. I think it is possible, but not straight forward for accurate collisions. However, I think you could easily get a lot further ahead than a Keep Apart Op gets you.

4- As was already mentioned, this is extremely simple to do with Box3. Using the Script output, you can setup Script data of whatever you need and send that out to a Force Op.

One final note, I no longer have that particular tut on my site. I needed to free up some space, so the larger ones had to go. Sorry.

Dear Charley,
thank your for this dedicated answers. yeah including speed/force to the calculation would be great! same goes for the rest of your suggestions. Letīs start creating such a tool. maybe we end up with a few new operators artists can use. I see the development of tools from both sides always. A TD wants to have the entire data operator so he ca change stuff quickly. An artists only want to USE the tool in production. He wantīs an operator which deals with his task. I would like to see us creating a tool which automates most of the known issues and tasks with fracturing we posted here and in other topics. We have the tools, so why donīt we throw some ideas together?

btw. i changed the link to allanīs tut to the intrisia mirror.

and thank you very much JohnnyRandom! :)

kind regards

anselm

PexElroy
03-30-2007, 07:31 PM
Very kick butt show reel PsychoSilence, fantastic work.

I also wait for Pflow Tool Box #2, for AGEIA PhysX with particles, as TP 2.5 has this now, and it's very useful. The only one I know that can do non-pre-fracturing for a 3D mesh model is Ron Fedkiw and who he works with, as he does R&D work at Stanford. He is using super computers also to do this amazing work. The idea of fragmentation on-the-fly is a few years off, but it can work, as the problem is in how should the mesh material fracture, and LucasArts is looking into this too, with the next-gen games with “The Force Unleashed”, and it requires a data set of materials to shatter correctly and pre-material planning, or it would not fracture correctly or look game level [not Film level], whereas DMM is using data sets with Havoc physics to know how it may fracture in real-time.

http://graphics.stanford.edu/~fedkiw/

http://www.pixeluxentertainment.com/tech.htm

http://www.lucasarts.com/games/theforceunleashed/gameinfo/news/technology.html


-- What I would do to use that Digital Molecular Matter (DMM) by Pixelux Entertainment

ruster
03-30-2007, 07:52 PM
I'm afraid even to ask about characteristics of those super comps :)
But looks like we've almost got to a new step in CG!

PsychoSilence
03-30-2007, 10:03 PM
Very kick butt show reel PsychoSilence, fantastic work.

thanks, bro :)

I also wait for Pflow Tool Box #2, for AGEIA PhysX with particles, as TP 2.5 has this now, and it's very useful.

yeah, i know and already work with TP for some shots i did...but iīm curious about how far you could take it with box#3. iīm pretty sure from the things i listed earlier al teast 2/3 could be done easily to provide one or two new operators or presets. iīm still new to box#3 which is a damn pitty but wanna find out where the borders are, if there are any real ones. iīd love to see us sticking the heads together on that.

The only one I know that can do non-pre-fracturing for a 3D mesh model is Ron Fedkiw and who he works with, as he does R&D work at Stanford. He is using super computers also to do this amazing work. The idea of fragmentation on-the-fly is a few years off, but it can work, as the problem is in how should the mesh material fracture, and LucasArts is looking into this too, with the next-gen games with “The Force Unleashed”, and it requires a data set of materials to shatter correctly and pre-material planning, or it would not fracture correctly, and looks very game-engine, like the sometimes-fake fractures in AGEIA, where it is just a face break off, and is fake, whereas DMM is not fake.

http://graphics.stanford.edu/~fedkiw/ (http://graphics.stanford.edu/%7Efedkiw/)

http://www.lucasarts.com/games/thef...technology.html (http://www.lucasarts.com/games/theforceunleashed/gameinfo/news/technology.html)

i know his website. great R&D stuf over there! downloaded the movies already for instiration :D

-- What I would do to use that Digital Molecular Matter (DMM) by Pixelux Entertainment havenīt heared of that until now. currently reading the website :)

thank you for your input!

kind regards

anselm

Glacierise
04-05-2007, 09:16 AM
First off, great stuff, Anselm! (as usual ;) )

On the fracturing - imagine we have a box with a deep shell. We put a particle on each vertex. Yes, I am still totally lame on box 3 an scripting, but I know that we can make groups, we can detect distances to nearby particles, and measure their speed. We can count the particles in a group, and if the vertices in the object are more or less evenly spaced, the number of particles in a group could be used to approximate the group's mass. I don't know how a group could be transformed (transl/rot) as a rigid body, but I suppose it could be done. I also suppose that a particle group could be meshed, a vertex put on each particle. This can even be used for a simulation of a fracture, withour pre-fracturing. But I guess I'm getting quite far out here :D

What do you think of this? I think I could delve in the research, if some of the gurus say it's possible :)

BrandonD
04-05-2007, 06:25 PM
I'm a firm believer in using Voronoi Diagrams in 3d as a method of fracturing. Using particles as sample points and the Voronoi method to re-mesh the geometry contrained to it's original topology would result in some nice looking fragments. This is similar to how Particle Studio did fragmentation, though it was constrained to triangles back in those days. The denser packed your sample points, the smaller your fractures.

Here's an example applet to play with. Click in the applet to place sample points and watch how it builds a fragmented array...
http://www.cs.cornell.edu/Info/People/chew/Delaunay.html

Glacierise
04-05-2007, 07:34 PM
This voronoi tessalation looks really nice! There is even a decent wikipedia article on it :D

http://en.wikipedia.org/wiki/Voronoi_diagram

Thanks, Brandon! Now we're holding our breath, waiting for Oleg :D

loran
04-06-2007, 08:48 AM
Voronoi Diagrams looks great! Can we hope a plug or maxscript using it??
An alternativ to fragment mesh is POLYBOOST (http://www.polyboost.com/). you can create amazing topology on any mesh in one click!
http://www.polyboost.com/Images/BBexamples2.jpg

http://forums.cgsociety.org/showthread.php?t=192942

PsychoSilence
04-07-2007, 03:47 PM
Happy easter to all of you :) ! keep the flower power alive!

kind regards

anselm

loran
04-10-2007, 01:28 PM
Hi folks, here is just some fun I just played around :)
I shared to boost that thread!
http://laurent.renaud.free.fr/divers/cgtalk/desintegrate001.jpg (http://laurent.renaud.free.fr/divers/cgtalk/desintegrate001.mov)
max9 file (http://laurent.renaud.free.fr/divers/cgtalk/desintegrate001.max)

PsychoSilence
04-10-2007, 01:52 PM
very nice as usual!!! :) and noble to share the file as well :D

loran
04-10-2007, 02:09 PM
thx Anselm,
I think we are the two lone guys who read that thread :))
I was just playing around with massiv particle renders with self shadowing after viewing KRAKATOA tests. How do you become betatester on krakatoa??

Glacierise
04-10-2007, 02:21 PM
No way you two alone :) I am still waiting for a development on the fracturing stuff, what do you guys think? C'mon :) And though it's too late, Happy Easter :D

mustan9
04-10-2007, 02:36 PM
That is very cool.

Thank you for sharing.

loran
04-10-2007, 03:00 PM
Glacierise, I don't use PFtb.3.
I would prefer to try to achieve a described effects better than a described method.
I m not clear sorry :)
tell me :" I want a glass fall and break on the floor"
Ok I will try to achieve this by any possible way.

don t tell me: " I want shatter of a glass in Pftoolbox3 with measur of each fragment and return values in a animation controller to animate a dog walk ..." :))

PsychoSilence
04-10-2007, 03:03 PM
hy loran! i emailed frantic software if i could get involved...wasnīt much of a story behind all that.

yeah, the fracturing. i hope we can develop something fresh indeed!!!...but i still learn box#3 :sad: i feel bloody n00bish *blushes*

oleg and i discussed shape collision in this thread: http://www.orbaz.com/forum/viewtopic.php?t=1108&start=0&postdays=0&postorder=asc&highlight=

hereīs our custom operator:
http://www.core2core.de/spielwiese/Shape-Collision-Fake.zip

http://www.core2core.de/spielwiese/particle-per-vertex_active.gif http://www.core2core.de/spielwiese/particle-size-by-grayscale_.gif
particle per mesh vertex & particle size by gray scale

http://www.core2core.de/spielwiese/2demo-scenes-exposed.zip

kind regards

anselm

nitrocom
04-10-2007, 07:40 PM
Good job, now i'm just takin look on it to understand material better, cause fadin out of box is very realistic, how did you achive that?

PsychoSilence
04-10-2007, 10:34 PM
tell me :" I want a glass fall and break on the floor"
Ok I will try to achieve this by any possible way.

don t tell me: " I want shatter of a glass in Pftoolbox3 with measur of each fragment and return values in a animation controller to animate a dog walk ..." :))

haha! thatīs great and itīs even possible with TPīs biped support :D

no, to be honest youīre totally right. but a few posts earlier i rose the idea of a box#3 driven (fake)fragmentation tool using the box#3 given new possibilities to gain a better look. thatīs what we talked about...

kind regards

anselm

depleteD
04-10-2007, 11:01 PM
Psycho- I tried doing that the otherday, but the problem i had was being able to access the mesh "id" in the array.

This is the problem it seams. Do you have a solution for this.

What I'm trying to do is replicate bobos script in pfbt3

PsychoSilence
04-10-2007, 11:16 PM
hm, i have no idea at the moment but itīs past midnight here :D

maybe oleg, bobo or allan are reading along with us?

kind regards

anselm

Wicked
04-11-2007, 09:18 AM
thx Anselm,
I think we are the two lone guys who read that thread :))
[...]

No way you guys are the only one! :p

I've been following this thread a long time, but almost never reply since my level of knowledge of particles ain't so high as the most of the frequent visitors of this thread. :p:)

Anyway, currently we are working on a 2d/3d shortfilm and I am assigned to do the particles and stuff. So I might drop by from time to time to ask for your help/advise.


Regards,

PsychoSilence
04-11-2007, 09:35 AM
Anyway, currently we are working on a 2d/3d shortfilm and I am assigned to do the particles and stuff. So I might drop by from time to time to ask for your help/advise.

yeah! do so. we are here and glad to help :)

cheers

anselm

PsychoSilence
04-11-2007, 03:14 PM
currently working on some minor vfx shots for an educational film for truckers...not too fancy and quite boring but at least can do some wreckage :D

http://www.core2core.de/spielwiese/cheap-truck-wreckage.avi

:eek::bounce::applause::shrug:

some scripting for random weight of each piece involved (inspired by a tutorial by guess who? yes, allan...). and YES the trucks looks cheeeeeeesy! itīs a reactor sim-dummy which is enough for my task atm to deliver demo files on how it could be done ;)

next step would be ground deforming and more truck wreckage like distortion...

kind regards

anselm

loran
04-11-2007, 03:23 PM
very nice! I hope you ll show the final shot

PsychoSilence
04-11-2007, 03:29 PM
thanks loran! coming from you that means much :)

well to me it still looks to chesy but with the right kinda animation by the anim.dept. and not my cheap reactor stuff this might turn out pretty cool...iīm currently setting up some krakatoa sand and fume smoke for postpro layering revealing that i still suck at fume :D

loran
04-11-2007, 04:51 PM
krakatoa sand? what do you mean? sand pushed by the sliding truck??

what s happened to the truck in the script? is it smashed by another one?

PsychoSilence
04-11-2007, 05:12 PM
you wish :D no it just turns over the cliff and slides down due to a mistake by the driver...not fancy at all but booorrriiing! i planed on doing some krak dispersed dust but that part was already dropped by the director a few minutes ago :D so my task is the glas shattering and some smaller rocks coming of the cliff as the trucks rushes down. maybe some left overs from the cargo if thereīs time...basicly the scene was intended to be more dramatic but we had a meeting today where the director pointed out that itīs still an educational film for truckers working at mines e.g. and no action sequence...*sight* i was about to burst it into flames and stuff :D :D :D

cheers

anselm

loran
04-11-2007, 05:24 PM
Ok, I m waiting to watch the achieved fx. So why Krakatoa on this?

PsychoSilence
04-11-2007, 05:43 PM
krak does nice sand twisters etc because of the high amount of particles it can handle after saving them into a file sequence. thatīs an awesome functional option!!! you build a huuuuge pflow handling like 10million particles. then you save it as krakatoa file sequence and delete the pflow! as you can imagine there is no 10minutes taking particle flow calculating every frame. instead krakatoa renders with 8x mblur in less then a minute per frame alowing me to render most of my stuff local... :) itīs faaaaaaaaaaast, thats why i love it...

cheers

anselm

EDIT: some left over gravel or tiny stones on the cargo area...dunno if it will be used in any kind of form later...just had a sec to throw it together...still lotta tweaking here and there:
http://www.core2core.de/spielwiese/leftover-cargo.avi

PsychoSilence
04-12-2007, 07:08 PM
hy guys...just wrote the cheapest macro in my life: the "StartFrame-PFlowViewToogle"...
it goes to the start frame of your animation BEFORE it toggles the pflow view. the advantage is that everything goes pretty fast now...how often i worked on a quite complex and high densed flow when i was somewhere in the middle of my timeline and when i hit "6" and edited the first value inside of an event flow started to simulate until that frame or you hit ESC to cut that out and turn everything off...further more it clears any selections. thats nice to me at least because sometimes iīm inside a modifier like EditPoly e.g. where 6 is assigned with another hotkey......dunno if this makes sense to you. if you think it suxx donīt download it or delete the line starting with "clearSelection()..." :D

macroScript StartFramePFlowViewToogle
category:"3Delicious"
tooltip:"3Delicious - StartFrame-PFlowViewToogle"
buttontext:"StartFrame-PFlowViewToogle"
(
clearSelection() -- Clears all Selections just in case you hit 6 (Default PFlow View Toggle Shortcut) while you are still in a sub-operator of a modifier using 6 as shortcut as well...
actionMan.executeAction 0 "50019" -- Time: Go to Start Frame
actionMan.executeAction 135018554 "32771" -- Particle Flow: Particle View Toggle
-- Particle View Open/Close Toggle
)

btw i assigned the hotkey "7" to it...so the old way of opening the pflow view still works...

kind regards

the thread whore :D :D

OlegB
04-12-2007, 08:03 PM
Anselm,

Check out an alternative option - Particle View > Options > Update Type > Forward.

Thanks,
Oleg B.

Glacierise
04-12-2007, 08:29 PM
Nice touch, anselm :) will check the forward update too, never used it before :)

JasperCG
04-15-2007, 04:01 AM
Is it possible, and if so how, to create explosions that realistic without the afterburn or similar plugin? Everywhere I look the examples are afterburn...which granted, are very cool, but I would like to try creating one without out.

I have tried a little, but they turn out very flat. I can't get the depth that is obvious with afterburn.

thanks for the info or help.

ArtiZta
04-15-2007, 02:28 PM
Hi Jasper, Have you tried using Atmospheric Apparatus under helpers with "fire effect" in environment? I think that fire effect is used to be called Combustion. It will still be far from afterburn but it can do some tricks if set properly.

cheers

loran
04-24-2007, 10:43 AM
I just play around with PF after viewing the last Digitaldomain Reel (http://www.digitaldomain.com/links/screening_room/character_reel.html)

Here is a similar effects I ve tested :)

http://laurent.renaud.free.fr/divers/cgtalk/DigitalDomain_Logo_FX.jpg (http://laurent.renaud.free.fr/divers/cgtalk/DigitalDomain_Logo_FX.mov)

PsychoSilence
04-24-2007, 11:10 AM
sweet bro!

maybe a little smalller but even more numbers and the disintegration of the logo could be a little better concerning the masking. over all kick ass as usual from you :)

cheers
anselm

TunnelLight
04-24-2007, 11:24 AM
looks a nice effect Loran!! Top Notch :thumbsup:

ArtiZta
04-25-2007, 03:36 AM
That should be... Lorandomain ....
cool man :)

rdg
04-26-2007, 08:11 AM
[1] http://forums.cgsociety.org/showthread.php?f=98&t=489438

Bobo: Rotate by Image
A pflow that rotates planeframgents (particles) based on image values.

As this is the comprehensive pflow thread this setup shoudn't be missing here.


Georg

PsychoSilence
04-26-2007, 09:49 AM
indeed! played around with it in a way like in the budweiser stadium add:
http://www.youtube.com/watch?v=p-F6jL21bzc

fun to gamble around with! a must-have preset ;)

kind regards

anselm

rdg
04-26-2007, 01:24 PM
Bobo's "Rotate by Image" and the odforce challenge 'Sunburn' inspired my to try this pflow setup:

RDG Shadowrun (http://proforma.preset.de/cgtalk/pFlow/shadowrun/)
A pflow that emitts particles only in the shadows of objects.

It is an quick'n'dirty - rather dirty than quick - implementation.

Georg

erilaz
04-26-2007, 01:33 PM
Okay guys, I'd like to start condensing the powerhouse of knowledge we have going here. I've set up a tutorial section on the wiki here:
http://wiki.cgsociety.org/index.php/3ds_Max/Particle_Flow

Here's a short tutorial on how to set up a wiki page if you don't know:





Click on the edit link next to the Tutorials heading
Type in the title of the particle flow tutorial. Now this is very important. Make sure you STRUCTURE YOUR TITLE. In the example I have added, the title Particle Flow/ precedes the actual title of the topic. This way if people are searching for things away from the main page they are easier to find. You type in the title using double brackets like so: Particle Flow/FAQ (http://wiki.cgsociety.org/index.php/Particle%20Flow/FAQ)
Once you save the page (you may have to save twice... there's some sort of weird error), a placeholder link will appear in red under the heading.
When you click on this link, you will be presented with a blank edit box. Here you can type your intro and start the tutorial. Try to keep some conventions, such as using Acknowledgment and External Links headings.
If you're unsure, feel free to ask any questions. Uploading images is handled via the http://wiki.cgsociety.org/index.php/Special:Upload paage, then you use that link in your document. Don't link images externally. :)

loran
04-26-2007, 01:45 PM
Very good Idea! I make some tutorials about different things over 3ds max to condense my knowledge without having to search to remmenber each time I need a particular effect.
I have done one simple about Pflow on my website and on Area-autodesk.com too.
If I have time I'll create some other

erilaz
04-26-2007, 01:48 PM
Please note, i've edited the page to go to: http://wiki.cgsociety.org/index.php/3ds_Max/Particle_Flow

The other was a more generic page about particles in general. This one is actually about particle flow!

Oh, and while you're there, feel free to start editing the FAQ I created! :D

PsychoSilence
04-26-2007, 06:24 PM
recently added some external links...i will add more when i find the time during the weekend.

cheers

anselm

avolution
04-27-2007, 07:52 PM
Hi! Where would one purchase the Krakatoa package; I do not see it for sale at
Blur.com

thanks

krak does nice sand twisters etc because of the high amount of particles it can handle after saving them into a file sequence. thatīs an awesome functional option!!! you build a huuuuge pflow handling like 10million particles. then you save it as krakatoa file sequence and delete the pflow! as you can imagine there is no 10minutes taking particle flow calculating every frame. instead krakatoa renders with 8x mblur in less then a minute per frame alowing me to render most of my stuff local... :) itīs faaaaaaaaaaast, thats why i love it...

cheers

anselm

EDIT: some left over gravel or tiny stones on the cargo area...dunno if it will be used in any kind of form later...just had a sec to throw it together...still lotta tweaking here and there:
http://www.core2core.de/spielwiese/leftover-cargo.avi

Steve Green
04-27-2007, 08:01 PM
Hi,

it's not written by Blur, it's written by Frantic Films.

It's not available to buy yet AFAIK.

- Steve

TTYO
04-30-2007, 03:10 PM
Hi I am a greenhorn at Particle Flow. Basically I can do a little bit of it but now I have a problem regarding flames.


I wish to make realistic flames like in the Ghost Rider.

charleyc
04-30-2007, 06:00 PM
Hi I am a greenhorn at Particle Flow. Basically I can do a little bit of it but now I have a problem regarding flames.


I wish to make realistic flames like in the Ghost Rider.


While you can make some pretty cool flames in PFlow...http://www.allanmckay.com/index.htm has a good tut on this, they are not really good as realistic flames. For those you are better off with a product such as FumeFX (http://www.afterworks.com)

Wicked
05-03-2007, 03:14 PM
Aight, here's a problem I've been crackin' my head about all day!

In the attachment you'll find the .max(9) file of a (some kind of) vine growing out of a wall. My task was to make some leaves grow on the vine at different moments.
In other words: they have to grow when the front of the vine reaces a certain point.
To me it seemed not that difficult to make, but this how far I got..

preview + max file: http://members.chello.nl/y.yigit/plant_test.rar
flow: http://members.chello.nl/y.yigit/plant_flow.jpg


(Also I don't know where that one leave comes from at the end. :shrug:
It's a particle, not a mesh.)

PsychoSilence
05-04-2007, 12:18 AM
Aight, here's a problem I've been crackin' my head about all day!

In the attachment you'll find the .max(9) file of a (some kind of) vine growing out of a wall. My task was to make some leaves grow on the vine at different moments.
In other words: they have to grow when the front of the vine reaces a certain point.
To me it seemed not that difficult to make, but this how far I got..

preview + max file: http://members.chello.nl/y.yigit/plant_test.rar
flow: http://members.chello.nl/y.yigit/plant_flow.jpg


(Also I don't know where that one leave comes from at the end. :shrug:
It's a particle, not a mesh.)

Deko made some nice vine tutorial: http://www.deko.lt/cg_education.html

scene files: http://www.cgtalk.lt/tutorial/video/Pflow_GrowPlant_scenes.zip
video tutorial: http://www.cgtalk.lt/tutorial/video/Pflow_GrowPlant.avi
codec if needed: http://www.cgtalk.lt/tutorial/video/TSCC.exe

kind regards

anselm

PsychoSilence
05-04-2007, 02:46 PM
guys and gals letīs feed the FAQ! (<< can i make that one blink?!?):

http://wiki.cgsociety.org/index.php/Particle_Flow/FAQ
(http://wiki.cgsociety.org/index.php/Particle_Flow/FAQ)
just made one entry for now. i will add some more later the day after work...

how about collecting the best tutorials in this thread (by using the search?) and on the net on certain tasks an artist might run into? then making a pol on it and adding them (if this is necessay, i donīt think so btw...)? i would volunteer for adding them then. who else?!? (green is the color of hope!)

jornas
05-05-2007, 02:28 PM
i have a problem with HaywoodTools..

i installed mx9 - boosterpack and now some scripts that i use very often stopped working, one very good and helpful script what i want to work is BlowChunks.ms .. any advice ? :sad:

Unknown property: "isValidMeshObject" in undefined

noouch
05-08-2007, 02:22 PM
Tested a couple methods for vegetation, one with Split Amount, Spawn by travel distance and sphere shapes, and the second with Spawn Rate, Spawn by travel distance and shape facing with a quick gradient bob map:

http://noouch.de/video/pfbaum1.mov

http://noouch.de/video/tree2_small.mov

PsychoSilence
05-08-2007, 02:39 PM
VEEERY nice growing time-lapse indeed! looks pretty much like DEKOīs growing approach: http://www.deko.lt/cg_education.html

any chance in sharing the file to the community?

cheers

anselm

noouch
05-08-2007, 02:58 PM
I got the basic idea from Deiko, so no surprise there.
What I'm missing is a way to generate a mesh skin from the particles, because spheres and bobs are a really bad way to do it. I'm even considering migrating to houdini, where it's a snap to do stuff like this with particles, plus it has a L-system, which is the better way of doing it anyway.

Didn't save the first tree scene, but here's the continuously growing branchy thing:

oatz
05-08-2007, 03:56 PM
Really great, Noouch.

Khye

SoLiTuDe
05-09-2007, 12:46 AM
#1: I have a system set up where in the first event the particles are split into 5 different events with a shape instance each. The shape instance is a group where I've animated a few objects before grouping... so I click animated shape and even when i turn off the display it feels the need to calculate like mad crazy to the point where the scene is unworkable.

#2: If I use a shape instance with animated shape then send out all the particles to the next event, they lose their animation from the animated shape... I'd like to use a single event to define their global motion, rather than using instances of a bunch of operators.

#3: I know the shape instance can randomly pick between sub objects kind of things, but not if they're groups withing a group... any way to get it to randomly pick between my different sets or groups of animated objects without using a separate event for each one?

Any pointers appreciated!

Thanks guys!
--Ian

CapitanRed
05-09-2007, 08:43 AM
well, I don't know about the group in the group thing.
but about the slow calculation. If you have objects with animated bones, or perhaps only with additional controlles on your base controllers, the calculation takes extremly long. I had five simple bird meshes, with only wings animated up and down by bones for a birds crowd. And it was so slow that it was not possible to work with it.

what i did then is, I baked the animation with bobos meshOven script. It takes also alot of time, but then the calculation of the particles was much faster.

by using the meshOven, only bake one object at a time, and watch your memory usage ;)

PsychoSilence
05-09-2007, 09:58 AM
hy ian,

yeah complex geometry as shape instance is always a hassle :( have you considered only using flow as a position driver? i had that issue back them when i had to create a shot where a sensor builds up out of thousends of little sensors(each high poly!) for an add. one of those solutions below saved my ass! capitanīs way baking the animation should ease the pain in addition. of the polycount doesnīt change during the animation use a simple pointcache on each animation and delete all animation helpers.
about the group thing i was told never to use more then one group because grouping in max is sh*t on my first day at work :D ! especially for animated objects! please try to ungroup the group inside the main group and parent-link everything to a dummy and treat that dummy as a group. that might work. or use TP :D

allan has a tutorial about how to transfer particle motion to meshes via a little scripting:
http://www.intrinsia.net/amckay/tuts/_amckay_PRB.avi
tested and working fine for me at least...

bobo has a workaround as well:
http://www.scriptspot.com/bobo/mxs5/pflow/pflow__MovingObjects.htm

i use the motion/rotation transform method more and more because the flow only has to shift around ticks then and not thousends of polys which are simply attached to itīs particleTM anyways...

kind regards bro!

anselm

OlegB
05-09-2007, 10:26 AM
If you have Box#3 (non-ACAP version), it has a new option "Fast Shape Evaluation" for Shape Instance operator. The reference for option reads: "When on, this option speeds up execution of the Shape Instance operator by evaluating the particle shape at the final integration step only..."

Another valid point here is that Shape Instance operator needs a check-box option with name like "For Rendering only" that indicates that operator should be disabled in viewports and enabled only during rendering. I'll add this option for 1.07 update of Box#3.

Thanks,
Oleg B.

PsychoSilence
05-09-2007, 10:33 AM
If you have Box#3 (non-ACAP version), it has a new option "Fast Shape Evaluation" for Shape Instance operator. The reference for option reads: "When on, this option speeds up execution of the Shape Instance operator by evaluating the particle shape at the final integration step only..."

Another valid point here is that Shape Instance operator needs a check-box option with name like "For Rendering only" that indicates that operator should be disabled in viewports and enabled only during rendering. I'll add this option for 1.07 update of Box#3.

Thanks,
Oleg B.

thatīs great news for the upcoming update! :) iīll definitely try the "Fast Shape Evaluation".

cheers
anselm

OlegB
05-09-2007, 12:35 PM
Hmm, on the second thought, we need more general solution for all operators in PFlow - that's related to "History Dependent", "Viewport Only" and "Render Only" option. I need to think about it - the solution should be on Particle View level in order to see by the operator's thumbnail icon the current status of the operator.

Thanks,
Oleg B.

joconnell
05-09-2007, 01:06 PM
Definitely some kind of cache loader would be great too - literally a single operator in the head of the flow with nothing else that can pull open cache files with no kind of calculation overheads and as little memory as possible - like the krakatoa prt loader or realflows plugin loader.

SoLiTuDe
05-09-2007, 05:56 PM
Hey Guys thanks for all your input. I actually did end up doing the geo->particle position script last night actually. It runs better, but it's still slow because it has to go through the script operators I think. It's at least better than the shape instance in this case, but it's still hard to work with.

Basically what I have is a bunch of big ships with little ships flying next to it. I need the big ship to be controlled by the particles.. (using a find target with a little bit of variation) there isn't a lot of motion here. ...The small ships need to have a little bit of variation (animation) and so what I did was keyframe some basic movement around the big ship, and then group them together. Then if I move the group the ships still move and whatnot. There is no vertex animation, it's just the little ships translating, and so I would think that using baked animation would actually be slower since it'd be like keyframes for every vetex, no?

I think it probably just comes down to the number of ships I have... I'll probably make a separate system for the ones near the camera (make only those animate) and just have the ones in the background be static.

Oleg: actually i was even going to suggest an "on at rendertime" feature for some of the slower operators. We don't have box 3 here yet, but soon enough I'll be looking into getting it.

Edit: Ok... so if I turn on my main system (A) it runs decently fast... as soon as I turn on another system (B) it lags beyond belief... and if I run another system (C) with system (B) it runs decently fast... as soon as I turn another sytem on, A freaks out... and I can't seem to find anything linking the two together... they're actually pretty basic setups.

OlegB
05-09-2007, 06:27 PM
so if I turn on my main system (A) it runs decently fast... as soon as I turn on another system (B) it lags beyond belief.
You may need to "clean up" your PFlow systems. If you have Box#1, it has utility "Clean up Particle Flows". With Box#3, it's done automatically for you.

Thanks,
Oleg B.

SoLiTuDe
05-09-2007, 06:35 PM
You may need to "clean up" your PFlow systems. If you have Box#1, it has utility "Clean up Particle Flows". With Box#3, it's done automatically for you.

Thanks,
Oleg B.


Thanks Oleg. I put in a PO to get it here like a week ago... god knows where it's at in the the paperwork though ... would love to have lock/bond right about now.


On another note...maybe I should change my profile so it says where I'm actually working (Gas Powered Games) instead of "looking for work" :)

PsychoSilence
05-09-2007, 06:46 PM
On another note...maybe I should change my profile so it says where I'm actually working (Gas Powered Games) instead of "looking for work" :)

leave it that way and change it to "blur studios" then ;) *crosses fingers*

CapitanRed
05-11-2007, 03:21 PM
simple thing...but why does it not work!!??!!

The File: http://www.megalomania.ch/capitanred/MaxFiles/ParticleSetup.max

The goal is to move a particle to the targets, but not get twice to the same target.

to do so, i have in the last event a scriptoperator before looping back to the start event. it says:

on ChannelsUsed pCont do

(

)

on Init pCont do

(



)

on Proceed pCont do

(

count = pCont.NumParticles()



for i in 1 to count do

(

pCont.particleIndex = i

deleteItem PositionArray CurrentNum

)

)

on Release pCont do

(



)



the currentNum is the index number of the current target in the array.

But i allways get a errormessage.

can somebody help me out?

nitrocom
05-11-2007, 03:31 PM
Hi all,

I have experienced a problem which related glu3d. I pasted the link of this .avi :
http://rapidshare.com/files/30723411/water_glu_test02.avi.html

Main problem here is, the water geometry and ice cubes can be seen in one shot render, however, if you want to render whole shot the water geometry cannot be seen...

As you will see in this preview, geometry can be seen both pers.view and preview. However when I try to take a movie, it doesn't work...

Any opinion will be welcome... :)

Regards...

nitrocom
05-11-2007, 03:36 PM
Hi all,

I have experienced a problem which related glu3d. I pasted the link of this .avi :
http://rapidshare.com/files/30723411/water_glu_test02.avi.html

Main problem here is, the water geometry and ice cubes can be seen in one shot render, however, if you want to render whole shot the water geometry cannot be seen...

As you will see in this preview, geometry can be seen both pers.view and preview. However when I try to take a movie, it doesn't work...

Any opinion will be welcome... :)

Regards...

avolution
05-11-2007, 05:24 PM
I am no master scripter; but it seems nowhere in your script do you ever
declare what these items

PositionArray CurrentNum are in the script such as

On init pCont do
(
GLOBAL CurrentNum=blahblah
GLOBAL positionarray=nahnah

etc



simple thing...but why does it not work!!??!!

The File: http://www.megalomania.ch/capitanred/MaxFiles/ParticleSetup.max

The goal is to move a particle to the targets, but not get twice to the same target.

to do so, i have in the last event a scriptoperator before looping back to the start event. it says:

on ChannelsUsed pCont do

(

)

on Init pCont do

(



)

on Proceed pCont do

(

count = pCont.NumParticles()



for i in 1 to count do

(

pCont.particleIndex = i

deleteItem PositionArray CurrentNum

)

)

on Release pCont do

(



)



the currentNum is the index number of the current target in the array.

But i allways get a errormessage.

can somebody help me out?

CapitanRed
05-11-2007, 09:44 PM
I updatet the file.

now it is looking for the targets it didn't go , but when it went all trough, again an error message apperas.
I tried to solve this with the if condition in the last script operator, but it does not work.

avolution, the variable and the array are declared in the previous script operators.

CurrentNum = random 1 PositionArray.count

BoxArray = $box* as array
PositionArray = #()

for s in Boxarray.count do
(
append Positionarray Boxarray[s].position
)

jigu
05-12-2007, 04:33 AM
i have one question :

say " we have one particle spawned by birth operator in first event and it's using "sphere" shape as "shape instance". now if we use "spawn by travel" test to generate trail from that one particle.. it generates trail from that particle but not from the surface of the "sphere" which we use as "shape instance"?

now question is : how to tell pflow to spawn particles from the surface of the sphere?
does it involve scripting?

i hope my question is clear.


excuse my bad english.

charleyc
05-12-2007, 06:55 AM
Jigu

There are a number of ways to do this. The simplest is to use either Mesher or a script to position scene objects to the particle transforms. You can then use that to create the spawned particles as you like. In older versions of max I have run into issues in the flow that required the clean up utility of Box1, but when it works, it is by far the simplest way to do this. I believe you will run into issues if your particles birth over time for the Mesher source. A solution is to birth them all at once and use another method (such as a collision test) to send to another event with the appropriate particle positioning. Other than that, you have to use a script to position particles with the appropriate transform offset. It can be done. The final way is to use Box3. It still requires the idea of the script, but it would be easier and probably run a lot faster.

PsychoSilence
05-12-2007, 11:27 AM
i posted a demo scene in another thread covering the same issue:

http://forums.cgsociety.org/showthread.php?f=206&t=493553

cheers

anselm

jigu
05-12-2007, 03:49 PM
thanks charley,PsychoSilence : i don't know much in maxscript. I found that mesher approach more easier.:)

PsychoSilence
05-12-2007, 08:00 PM
hy guys,

i have issues with allanīs "drip_system III" rain system on moving objects. everything works like a charme in his demo scene on a sphere. so does it on my custom object (a character). but on moving and animated objects it wouldnīt work properly due to following the char-meshes movement issues. lock/bond was my first thought but i still donīt get it working :cry: any input?

kind regards

anselm

charleyc
05-12-2007, 08:25 PM
Lock/Bond is the solution for things like this. What specific problems are you having with it?

PsychoSilence
05-13-2007, 08:05 PM
the prob i get mainly is that the lock/bond fixes the particles in position where they hit the deflector. but i still want the particles to run down the surface to build trials as they do without lock/bond but on animated shapes...

charleyc
05-14-2007, 01:55 AM
If you decrease the Position Lock/Bond Force, they will slide across the object. Also you can set the Break Off When Exceeds: parameters so that the droplets fall off realistically. It is really made for just such effects.

PsychoSilence
05-14-2007, 12:59 PM
yeah i know...used that on my venus image e.g. but iīm getting trouble with my rainsetup...

i did some reverse engeneering and started from scratch with allanīs setup. i attached a max file below (max8, box#1 needed) evetually one of you ccan fix it so i can finally dance in the rain :)

kindest regards

anselm

PS: i know itīs a simple setup. iīm just seriously stuck somewhere really need to get my head free of sh*t. damn GF :D...embarrassing embarrassing embarrassing :cry:

charleyc
05-14-2007, 05:18 PM
I posted an example file here

http://www.orbaz.com/forum/viewtopic.php?p=5761#5761

PsychoSilence
05-14-2007, 05:47 PM
thanks a bunch, bro! iīll have a look at it immediately.

kindest regards,
anselm

PsychoSilence
05-15-2007, 07:08 PM
[shameless plug]

An article written by me was just released at www.CGArena.com (http://www.CGArena.com) :
Freelancer's Manifesto - A Collection of advices by Anselm v. Seherr - Thoß, Germany

http://www.cgarena.com/freestuff/tutorials/misc/freelancers/header.jpg (http://www.cgarena.com/freestuff/tutorials/misc/freelancers/index.html)

This text is neither really an article nor a tutorial. Itīs more a kind of collection of advices and guidelines but may be informative for young artists planning on starting a freelancer career and need to get going.

online article version:
http://www.cgarena.com/freestuff/tutorials/misc/freelancers/index.html

kindest regards,
Anselm

Glacierise
05-15-2007, 09:27 PM
Read and bookmarked for daily reference :D

Very usefull article man, I wish there were more of those! I've only read a couple of insights into the lives of CG wizards, and I'm very excited for more.

And, your writing is sincere and passionate. I feel you as a friend now!

Thanks,

Hristo

PsychoSilence
05-15-2007, 09:39 PM
Hey Hristo,

thank you veeeery much for the kind words! i feel honoured *blushes* :blush::blush::blush:

kindest regards,

anselm

Glacierise
05-15-2007, 10:28 PM
The pleasure is entirely mine, I assure you. I am kind of a noob too, started less than a year ago. I am slowly building my staircase away from lameness (in the arch-viz limbo right now), and guys like you are a great help and inspiration. Keep livin' passionately and contributing!

jigu
05-16-2007, 04:01 AM
hey 4 pages .....gr8 article. must read.:thumbsup:
thanks for it.

rdg
05-16-2007, 04:18 PM
robotik experiments

Hi am investigating the possibilities of simulating an industrial robot using pflow.
The task is to collect an unknown number of moving 'objects' and place them on a specific point.

This is the first setup/try:
http://proforma.preset.de/cgtalk/pFlow/robotik_3.wmv

The red point is a particle that collects the four objects in a loop.
(the process is inspired by this thread: Every Target 1 time SCRIPT (http://forums.cgsociety.org/showthread.php?f=206&t=495915))

the main problem here is:
I have'nt managed to dynamically add objects to the findTarget list.
But an robot will unlikely try to 'find' its way to the object - it will propably know when the object is at a specific place ...

Anselm allready mentioned that box#3 will hold some answers.
And as you might have noticed - i allready used the box#1 freebees ...

PS: Anselm your article rocks!

Georg

charleyc
05-16-2007, 04:27 PM
If you treat this like reality, you are right, the robot will be programed to 'know' exactly where the part is. Based on that, logicaly, the parts will always be in the same spot. You can easily use collision planes to transfer the particles to different events as required.

-The particles move down a conveyor belt until they reach a specific point where the robot will grasp them.
-Either there is a collision object at that specific point, or it is linked to the robots grasp, but at that point, they hit the collision.
-That transfers them to an event where they inherit the motion of the robot.
-At the drop off point, there would be another collision object that would change the particles motion again.


If you need something a little more intelligent, you would have to get a bit more creative with the Find Target Script you are using. Although collision tests would probably find their way into that method as well.

rdg
05-16-2007, 04:46 PM
charley
thanks for your input.

The communication between particles and scene objects is one of the great white spots on my max map.

I am not sure if the 'grasping particle' should controll the robot or the robot should interact with the particles.

I had an setup with a conveyor belt and two deflectors.
particles passing the first deflector are entering the "ready to be picked up zone".
particles passing the second deflector would start falling down and are deleted.


In a working setup particles should never reach deflector 2, though :D


The event 'ready to picked up' has a script that should check for new particles.
This particles would be added to a global list/variable holding all particles in that zone.
The robot should pick one by one and place them on another belt.

I had so many problems that I started to rip the scene apart and test small parts.
The above posted movie was the part:
A particle tries to find multiple targets.

One of the problems was:
I tried to build a list of the ready particles using:
if pCont.isParticleNew i then (
append readyParticleList i
)
which should append new ready particles to a global array.

But each particles was added 3 to 5 times in a row.

Georg

charleyc
05-16-2007, 05:28 PM
I may be misunderstanding what you want, but it seems like you might be overcomplicating this. Here is a scene showing what I understand you to be needing. Let me know if this is not what you had in mind.

It is an R9 scene

rdg
05-17-2007, 11:24 AM
:blush:

complexifikation is my middle name.

Yeah this is an interesting and smart(!) approach.
An it is even scalable to support multiple parts with offsets on the belt.

Thank you!

Georg

PsychoSilence
05-21-2007, 12:36 PM
Oleg just released the long expected surface spread at orbaz!

http://www.orbaz.com/forumData/animations/20070402/SpreadWave.avi

http://www.orbaz.com/forum/viewtopic.php?t=1200&highlight=

cheers

anselm

andyburm
05-22-2007, 01:21 AM
really nothing to do with Pflow but I thought I'd post the making of sandman from spiderman3
http://www.imageworks.com/video/spiderman3_1.html

:)

PsychoSilence
05-22-2007, 05:35 AM
hy andy,
thanks for sharing this. downloaded it right away :)

DeKo-LT
05-23-2007, 08:31 AM
Great article Anselm, congrats :thumbsup:
Every young artist must be very thankful for you! ;)


Btw, guys, Autodesk Masters Nominations:
http://usa.autodesk.com/adsk/servlet/index?siteID=123112&id=7276914

PsychoSilence
05-23-2007, 12:45 PM
Hey Deko! Long time no see. We miss you much and i think i can speak for the entire flow community here :)

Congrats for your nomination to Autodesk Masters!! so you will be at sigg this year? once i will be in one of those lists too maybe *sigh* *boots max and does something cool* :D nah joking ;)

kind regards,
anselm

PeteDraper
05-23-2007, 12:50 PM
Yeah congrats to all nominated!! Helluva list on there...


... already voted (and not for myself either hahahaha!)

TunnelLight
05-23-2007, 03:44 PM
(and not for myself either hahahaha!)

I'm sure you can get one of your young lads to Pete :buttrock:

SoLiTuDe
05-23-2007, 03:54 PM
Oh man! Too many flipping good choices there. I don't even know where to start. :banghead:

loran
05-23-2007, 04:07 PM
I'll be on Sigg07 this year. Any advices for the visit?

PsychoSilence
05-23-2007, 05:14 PM
meeI'll be on Sigg07 this year. Any advices for the visit?

meet me and get horribly drunk with me :D

@ Ian: will you be online the next days? because i plan to stay in LA for a lil while...and you will be a hell of a LA guy by then i suppose. :D

amckay
05-23-2007, 05:56 PM
Hey Pete haha yeah it's a pretty cool list! Whos going to sig? I'm going to be holding a masterclass, so good excuse to get drunk this year!!!

Love to meet everyone!

Glacierise
05-23-2007, 06:06 PM
Voted for Allan, couldn't help myself :D Too bad I couldn't post for Pete Draper, Jeff Patton, Deko, etc... But that's democracy for ya :D

SoLiTuDe
05-23-2007, 06:07 PM
Anselm: Talking to via icq... so yeah dude we're partying... prob with allan at that. :)

Guys: I'll be moving to LA in a few weeks... so if you're gonna be in that area let me know so we can hang out. ...San Diego is about 2 hours from LA, so I'm 99% sure i'll be going to siggraph this year.

SoLiTuDe
05-23-2007, 06:08 PM
Voted for Allan, couldn't help myself :D Too bad I couldn't post for Pete Draper, Jeff Patton, Deko, etc... But that's democracy for ya :D

Just get on more than one computer. :thumbsup: Kinda defeats the purpose though

cnaeger
05-23-2007, 06:11 PM
Has Autodesk posted the Masterclass schedule for this year yet?

Alan,

What is your masterclass session going to be on? PFlow I imagine but any details?

Thanks,
Chad

amckay
05-23-2007, 06:23 PM
Hah thanks Sol
you use msn btw? shoot me an email at amckay@allanmckay.com

Flying back to Australia today, just wrapped a bunch of jobs here in New York

cnaeger - it's not officially announced yet, I just agreed to do it today. But I'll just say key words are fume/digital pyrotechnics (Fluids explosions etc), case studies on big films I've worked on, and other cool fx stuff :)

SoLiTuDe
05-23-2007, 06:24 PM
http://usa.autodesk.com/adsk/servlet/item?id=9476889&siteID=123112&linkID=7161306

Where it will be when they announce the classes. :)

PsychoSilence
05-23-2007, 06:27 PM
Hah thanks Sol
you use msn btw? shoot me an email at amckay@allanmckay.com (http://forums.cgsociety.org/amckay@allanmckay.com)

Flying back to Australia today, just wrapped a bunch of jobs here in New York

cnaeger - it's not officially announced yet, I just agreed to do it today. But I'll just say key words are fume/digital pyrotechnics (Fluids explosions etc), case studies on big films I've worked on, and other cool fx stuff :)

iīll be there! germany VS aussie! sounds like fun. big time.
how much do they charge for the masterclass and who do i have to sleep with to join in for free?

Wicked
05-24-2007, 10:30 AM
Me and my big mouth..

For the short movie we're making, I told my teammembers I would take the VFX-tasks on me. But now it seems I've underestimated how much work comes with it. I guess I underestimated it because I saw tutorials/examples on how to make (for example) smoke, fire etc. and how simple it seemed. I made simple examples and showed them to my teammembers, and they were convinced I would be up for the job.

And now the pressure is building up: One of the problems I now run into is how to create realistic dust coming of debris.
Offcourse I used the search (as well here and almighty Google ;) ) but I mostly come out on using Afterburner. (http://laurent.renaud.free.fr/tuto02-uk.html)
Also I found a good tutorial on making a realistic smoke-material (http://tutorialscentral.com/index.php?option=com_content&task=view&id=40). It looks good,... as long you don't use it in a (particle)animation. Behind this link (.rar, 11 MB) (http://members.chello.nl/y.yigit/smoke_test4.rar) you can see what I mean. It's an animation where you can see the smoke not moving as it should. The (not so) funny thing is that the particlemotion is slightly different then what you see in the animaton.. Weird.
Anyway, some tips/advice are welcome. I'll be keeping you guys updated when I got something worth showing :)



P.S. Voted on Allan because the first (video)tutorials I found on PFlow were his vids. Ever since I'm pretty much hooked on VFX. :)

PsychoSilence
05-24-2007, 01:17 PM
ORBAZ TECHNOLOGIES TO ANNOUNCE PARTICLE FLOW TOOLS: CACHE AVAILABILITY

http://www.orbaz.com/forumData/images/20070523/ParticleFlowToolsCache.jpg

http://www.orbaz.com/forum/viewtopic.php?t=1210&highlight=

PsychoSilence
05-24-2007, 05:40 PM
Hy community iīm bound to my bed because i got a cold :(
so iīm just gambling around on my notebook...

http://www.core2core.de/spielwiese/bitstream.wmv

i attached a still and the max8 file in case anyone is interested.

my personal task was to visualize that data transfer is fast so i took mocap data of a jogger (which i had to modify slightly thatīs why itīs not realytime but a lil slower actually :( ).

the two things really making me scratch my head over and over are the following:

the random lock/bond docking. i already tried allanīs script that stores the position before particles get send out to the original start event but since i spawn particles with a rate and the target is animated that wouldnīt work as expected. it only works when all particles are spawned at once and the target isnīt animated actually...

second thing pissing me off is that the particles donīt get in line properly after they rip off of the jogger dummy. itīs a very small second find target to allign them again...

maybe iīll play around with the speed by icon more...it should do a wired trip thru a tube of virtual reality with the camera following it (boboīs ride the flow tut...) :D and to pust the speed visualisation any further iīd lke to have a quadruped from a mustang running :)

anyways, nothing big. small stuff from the sicko.

kind regards,
anselm

andyburm
05-24-2007, 08:25 PM
pretty cool Anselm, thanks for sharing :)

rdg
05-30-2007, 01:02 PM
Anselm's got the cool tricks.

He suggested I try 'shapemarks' to solve one of my homebrewn problems.
I didn't try it yet, but started to use pflow to get my rolling cube stamping colors on the floor:

http://www.preset.de/2007/0530/rollingcube/

Oddly the rig starts freaking as soon as a pflow is active that relies on deflectors attached to the rig. After baking the movement the pflow solved just fine.

Georg

PsychoSilence
06-02-2007, 05:38 AM
http://core2core.de/spielwiese/SHBS.jpg (http://core2core.de/spielwiese/SHBS.WMV)

http://core2core.de/spielwiese/SHBS.WMV

http://core2core.de/spielwiese/SHBS_setup.gif

Self Healing Biomechanic Suit:

Closes wholes in the chassis. Delay and Speed fully scalable. Maybe iīll make a hero shot where the hero gets hit by a grenade in the chest and his body armor suit self repairs...

A Mystique Setup like in X-Men would be possible too without the meta shapes (in my case PWrapper)...

kindest regards,

anselm

Glacierise
06-02-2007, 06:06 AM
Great! Now just put RayFire in the mix... :D

charleyc
06-02-2007, 06:54 AM
Thats cool Anselm. There is an Operator in PFTools #3 that does this
www.charleycarlat.com/Tests/PFSsurfaceSpread.mov (http://www.charleycarlat.com/Tests/PFSsurfaceSpread.mov)

PsychoSilence
06-02-2007, 11:18 AM
i actually used box#3 surface spread for the repear. Very nice movie youīve got there! What were your settings for the spread? Mine were pretty chaotic to get that organic kinda look of growing flesh...

@ Glacierise: heh! yeah worth a try! next step would be to make it any more procedural. right now thereīs a predefined area where a deflector hits. maybe iīll make a nice grenande hit or rapid fire shot.

kind regards,
anselm

charleyc
06-02-2007, 06:13 PM
Ah..I see it now. I should look closer before posting :)

The settings are nothing fancy. Here is an R9 scene with a semi-procedural approach. The impacts are procedural, but the Surface Spread is not really procedural in that it is activated at a specific frame.

www.charleycarlat.com/Tests/PF_ProceduralHeal_01.zip (http://www.charleycarlat.com/Tests/PF_ProceduralHeal_01.zip)

Here is a preview
www.charleycarlat.com/Tests/PF_ProcHeal01.mov (http://www.charleycarlat.com/Tests/PF_ProcHeal01.mov)

JasperCG
06-06-2007, 07:25 PM
I am trying to use Pflow to create an effect similiar to snow falling on a skinned character or splattered jelly or similiar.

I have a Collision test, speed stop, but then how do I get the particles to follow the character?

Thanks

PsychoSilence
06-06-2007, 08:59 PM
hy Jasper,

you should use a position object operator after the collision event with locked on amitter checkend and set to Animated Shape. or use a lock/bond operator from box #1: http://orbaz.com/products/particleflow/box1

kindest regards,

anselm