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 25

Bobo
06-02-2008, 03:30 AM
on a side note, bobo I've just relocated to vancouver as of today - so I will make sure I pop over to winterpeg for a few beers, or in your case coffees :)see you at siggers mate

Hurry up before the summer has come and gone again ;)

PiXeL_MoNKeY
06-02-2008, 04:17 AM
I have worked with Allan Mackay, And I have recently found my own work on his showreel, work he never touched.. He is also claiming some practically shot explosion as being 3D which be claims to have created. This Guy is a Hack.. Ask around! anyone who's worked with him will tell you the same stuff.I wonder Scott who should I believe, Allan the Autodesk Master, Autodesk demo artist, Autodesk Masterclass instructor, and has released various free and commercial training material and scripts. Or you who has done...um...who are you?

-Eric

Glacierise
06-02-2008, 11:25 AM
I wonder why people reply to flame posts at all.

On a constructive note - I need some advice. What's the good route to go on building dynamics capabilities? I'm trying my best with reactor, but there are some seemingly insurmountable obstacles, and I have a small limit of rigid bodies I can use before it starts crashing - about 1500. I see the way to go is with fewer, big fragments with reactor, and lots, smaller debris/dust with PFlow, and I am doing my best with that too. Of course, sometimes it doesn't cut it, and especially when I am using the havok 1 solver (because 'inactive' doesn't work in havok 3), it's very inefficient. I am not looking for some ultimate solution - that would be silly :) But I would appreciate an advice on the other options out there. If some serious stuff could be done with Box#2, that would be awesome. I know that there are people out there with betas and info, and I don't need any specific info, just a 'Yeah, you'll be able to do useful dynamics, don't bother with other options, just wait' would be enough, if that's the case. TP3 is a thrilling option, but possibilities to work with it are limited, as I see it. I don't really know anything about rigids in Maya, but I've found on the Blastcode site that 'The next generation of stuff will be available for most platforms' by the guy in the Blastcode team. Then there is Houdini, the starving artist edition looks cool. Again, if I can use the Ageia solvers in a reasonable, functional way in Max in a few months that would be awesome. So please advice ;) Thanks!

holycause
06-02-2008, 12:39 PM
hi Allan,

did you get my email?

On a constructive note - I need some advice. What's the good route to go on building dynamics capabilities?

I'm learning TP, it's really amazing all the stuff you can do with it.

Mills
06-02-2008, 04:12 PM
Hi ya.
I am trying to do a combination of Birth script and Allen's tutorial on "objectAB"
I am having problems getting the objects back to the exact position. They are a fraction out.
What am i doing wrong here?
Pls if someone has a sec take a look here. I did a simple test. Max 9

www.500mills.com/woking_concept_BirthObjectAB.zip (http://www.500mills.com/woking_concept_BirthObjectAB.zip)

Thanks
E

JohnnyRandom
06-02-2008, 07:02 PM
On a constructive note - I need some advice. What's the good route to go on building dynamics capabilities?

Hey Hristo,

Also a note about Box#2, it is still pretty early in dev, but it has gotten x10 better since I first used it, it uses Aegia, and is looking very promising. I have had some simple test scenes using upwards of 5000 inter-colliding particles and it holds it's own quite well. There is no release date set.

I have been playing with the new TP3 demo as well, and I have to say, holy SH!T, what an amazing job cebas did! I agree though something you don't just buy on a whim, it has a hefty price tag, but it's capabilites are seemingly endless.

I must say though at this time simple to average size pflow scenes are much faster to setup. As with anything though it just takes time to learn and be more preficient.

amckay
06-03-2008, 12:22 AM
Hi holycause, can you send again just in case? I'm a bit behind on emails - I just moved to a new city, broke my arm and been on vacation - so its made it a beeyotch to keep track of things + type at all right now :P

Glacierise - yeah honestly havok/reactor arent the best of solvers.. if you look at bs2 or tp3 both will be a lot more stable, especially for lots of particles. Realflow also has very stable dynamics, as does maya. Also if its just rigids, oddly enough sim cloth has a rigid solver - so assuming its out for max 9/2008 grab it its free and very solid. Reactor makes me cry sometimes...

TP 3 is finally starting to become a very solid tool, price wise and well cebas support are pretty wishy washy - however with 3.0 its definitely worth looking into, prevously its been a good tool but not for everything, like I worked on a film a few years back with 2. and got everything approved I just needed to put the textures on it, and then I found out that I couldnt use the shader I wanted with TP and stupid things like that. But now more studios are starting to use it its starting to age into a very decent product

PsychoSilence
06-03-2008, 01:15 AM
how the hell did u break your arm?!?

greets to the frantic bwwwwoooooyyyyyys once u go out with em!

holycause
06-03-2008, 03:42 AM
oh...
not cool...

It's done I send it again to u.

Glacierise
06-03-2008, 05:17 PM
Thanks for the replies guys! Allan, what the hell happened to your arm dude?! You swam to Canada or what :D Hope you get well very soon. And thanks for the info on the simcloth solver, how could I not have known that. And it seems I will be tinkering with the TP demo, too bad box#2 won't come soon. By the way, Intel have made the Havok middleware free, have you seen that news?

PexElroy
06-03-2008, 11:55 PM
I need some advice. What's the good route to go on building dynamics capabilities?

like Allan and others have noted, TP3 is getting better and is now available to use, look into it and learning it will allow a lot of options. TP3 can do a lot of strong real-time particle dynamics, all with fast results, and no reactor preview mojo.

I wait also to see what we'll be able to do with Box #2. Most of these new consumer level dynamic rigid body tools are in their growing stages from what is being done in film, and in a few more years they will be even more robust than TP3/BC is today; and agree reactor is good for low-accuracy rigid simulations and fussy on complex events;

PsychoSilence
06-04-2008, 04:56 PM
http://www.vfxhack.com/2008/06/02/this-just-in-all-nighters-crap-visual-effects/

this is going out to all the weekend warriors and over-nighter-fighter!

JohnnyRandom
06-04-2008, 05:50 PM
Anse thanks for sharing, checking his blog brought up this (http://www.vimeo.com/993998) ... that sh*t is insanely awesome stop-mo!


Oh yeah, Allan how did you break your arm?:curious::eek: At least doing something fun, I hope, at the least :D

Glacierise
06-04-2008, 06:13 PM
Thanks for the advice guys, I'll be diving into TP3!

@Allan - dude, doesn't stunts get done in VFX instead of live, just to be safer? You kinda undermine the industry :D :D :D

@Anselm - great blog dude! Added to the RSS. And the stopmotion face stomping is awesome reference :D

Michael-McCarthy
06-04-2008, 07:48 PM
Thanks for the advice guys, I'll be diving into TP3!
For those of you guys that are diving in to TP3 and/or trying the demo, please take a look at some of these resorces:

3+ hours of free, video training/demos on ThinkingParticles stickied at the top of our forum here:
http://www.cebas.de/forums/smf/index.php?board=26.0

Register for the forum and there is a great and very professional group of guys there that are full of help.

Also take a look at CGfluids.com. This is a good resource for FumeFX, and TP info and support.

Lastly; on top of the video training we already have we will be releasing a ton more in the in the next few months. So stay tuned :D

Thanks a bunch
Michael McCarthy
cebas USA

holycause
06-04-2008, 09:19 PM
Thx for this link Michael,
I'm sure it will be helpfull for a lot of poeple :mrgreen:

amckay
06-04-2008, 10:35 PM
I dont know whether to tell the truth or make up a story about saving a kid from a burning building... actually on my flight over to vancouver all the flight staff had put it on themselves to come up with a good story before I land that made me out to be bit more courageous ;)
No story to tell really, my friends big dog was excited and jumped on me unexpectedly (he was about 20kg/44lbs) so he pushed me over when he leapt onto me and I managed to fall on my wrist wrong, thats about it to the story! Not the best time when I have to pack and move all my stuff to canada.. thankfully I have some very very cool friends who did most of the work!

Mills, I only have a mac laptop with me so I cant check the scene, but I believe the idea is to have dummy objects in the original position so at frame 0 you tell your renderable objects to reposition to them. try using transform if position is off, I'd need to look at the scene to know whats going on but I'm pc'less right now! doh!

psycho - yeah catching up with all the vancouver frantic gang for beers tonight

If anyone heres from vancouver lemme know, happy to grab some drinks once I settle in - itd be good to organize a night in vancouver to get all the local 3d'ers out for drinks to meet everybody, same with siggraph I'm sure :)


cheers Michael you guys are going really well with putting up a lot of resources fot tp 3!

JohnnyRandom
06-05-2008, 01:17 AM
Lastly; on top of the video training we already have we will be releasing a ton more in the in the next few months. So stay tuned :D

Thanks a bunch
Michael McCarthy
cebas USA

Swweeet :)

so he pushed me over when he leapt onto me and I managed to fall on my wrist wrong,

Not swweeet :argh: As for friends, cheers:beer:...The Beatles (http://www.imeem.com/nonongsky/music/djEFdxXJ/a_little_help_from_my_friendsbeatles/)sung it best and well Joe Cocker (http://www.imeem.com/michael2nd/music/haaTiNX_/joe_cocker_with_a_little_help_from_my_friends/) did pretty good too... How about the Kennedy Choir (http://www.imeem.com/rykim/music/595odkXb/the_kennedy_choir_with_a_little_help_from_my_friends/) :D Or Joe Anderson& Jim Stugress (http://www.imeem.com/people/AkyFGL/music/s5JiKQZF/joe_anderson_jim_stugress_with_a_little_help_from_my_friend/)

Digital Masta
06-05-2008, 03:16 AM
I dont know whether to tell the truth or make up a story about saving a kid from a burning building... actually on my flight over to vancouver all the flight staff had put it on themselves to come up with a good story before I land that made me out to be bit more courageous ;)
No story to tell really, my friends big dog was excited and jumped on me unexpectedly (he was about 20kg/44lbs) so he pushed me over when he leapt onto me and I managed to fall on my wrist wrong, thats about it to the story! Not the best time when I have to pack and move all my stuff to canada.. thankfully I have some very very cool friends who did most of the work!


44lbs is big? Good thing it wasn't my dog he's about 110lbs...anyway although I guess that has to really suck as far as getting work done goes.

Glacierise
06-05-2008, 12:08 PM
Wow, thanks for the TP videos, very helpful! Allan, the simcloth solver is such a gem! Only problem I have with it is the clumping of rigids together in some cases, but I guess that's due to interpenetrations and can be solved with more substeps/col tolerance. And beware in Canada, there's scarier beasts then just big dogs there :D

amckay
06-10-2008, 12:43 AM
Hi guys does anybody habe John O'Connells email address?

JohnnyRandom
06-10-2008, 02:06 AM
Drop him a line through the boards email :)

check chaosgroups forum

PexElroy
06-15-2008, 10:25 PM
Yeah thanks for the TP3 videos, wish though they also included those exercise files ;)

Michael-McCarthy
06-16-2008, 02:08 AM
Yeah thanks for the TP3 videos, wish though they also included those exercise files ;)
Heya Rob,

Exercise files for a lot of the videos are installed with the demo and full version of TP3. There is both a "Tutorials" folder and a "Video Training Scenes" folder.

Of course not all of the Siggraph demos and film explanation files are available. However some of these Siggraph demos scenes are in the "User files" folder of TP3 scenes.

Thanks a bunch
Michael McCarthy
cebas USA

PeteDraper
06-16-2008, 05:14 PM
wotcha all! been a while since I've been on here as I've been knee-deep in more book-related stuff :) Finally time to announce some of the new DTE tutorials I've been developing over the last few months, most of which are very particle heavy including a pretty cool terrain weathering system, air bubble motion, tornado destruction and galaxy setup. This book has 100% new tutorial content, but I'm also giving away the 2nd edition totally free as an electonic download if you buy the 3rd edition.

Should be out soonish once I've finished the content ;)

few teasers below... as per usual, all without plugins :)

Cheers!

Pete

http://www.xenomorphic.co.uk/images/dte_3_largebubbles.jpg
http://www.xenomorphic.co.uk/images/dte_3_snowdrifts.jpg
http://www.xenomorphic.co.uk/images/dte_3_atoll02.jpg
http://www.xenomorphic.co.uk/images/dte_barntornado.jpg
http://www.xenomorphic.co.uk/images/dte_3_island.jpg
http://www.xenomorphic.co.uk/images/dte_3_galaxy.jpg
http://www.xenomorphic.co.uk/images/dte_3_traceflare.jpg
http://www.xenomorphic.co.uk/images/dte_3_weathering.jpg

also the arbitary "ball o' fire" is featured too :)

EDIT - removed the ball o fire image as I've rejigged that scene and the original version looked crap in comparison. I'll post another one at a later date //pete

PeteDraper
06-16-2008, 05:15 PM
few more... :)

pete

PeteDraper
06-16-2008, 05:19 PM
forgot to mention... if there's anything that you feel would be worth chucking in there, please let me know as I'm open to suggestions on additional tutorial content :)

ta

pete

PsychoSilence
06-16-2008, 05:35 PM
i wanna pre-order a signed version hereby :)

iīm doing tornado destruction lately. will it be out on your website too?

EDIT: pre-ordered via amazon.com a minute ago...

Glacierise
06-16-2008, 05:43 PM
Pete you're still kicking dude that's awesome! When's your book coming out, seems on the brutally cool side of things?!

PeteDraper
06-16-2008, 05:56 PM
yep still alive (barely tho!). The new one should be out in a few months; I've got until next month (ish) to finish it, then they're into layout and editing before the print job. I'd say it'll be on the shelves around october give or take a month, depending on where you live and where they print the thing.

It's cheaper this time around than last time (think it's about $10 cheaper if memory serves) as the publisher didn't want to go down the dvd rom route, so all of the assets will be on a secure website maintained by them for readers to grab. How that's gonna work I'm not sure as there's a HELL of a lot of assets to download; including the free 2nd edition stuff, you're talking about 8gb...!! ... but hey, I'm not paying for their web hosting charges hahahaha!!

I'll chuck some more samples up here once they're at a decent stage.

psychosilence - check yer pm's in a sec :)

pete

Glacierise
06-16-2008, 06:10 PM
I'm getting them through a local bookstore that deals with amazon, etc - it borders on ripoff, but there are things too cool to pass by. What's the approach this time - is it still the hardcore 'no plugins' deal or you've changed :D

PeteDraper
06-16-2008, 06:16 PM
hey Glacierise - yeah still plugin-free :) Which unfortunately means I have to deal with a lot of blobmesh crapola, but feel free to use PWrapper :D

Glacierise
06-16-2008, 06:41 PM
Heh, you're deep in the machine! Awaiting the new book enthusiastically, and you keep an eye out here - we have some cool stuff to show too! Btw, you showing any scripting? That tronado destruction based on mass sounds interesting!

PsychoSilence
06-16-2008, 06:50 PM
Please check PM, Pete :)

PeteDraper
06-16-2008, 06:52 PM
the massed based system is something I'm holding back for a bit as it's not finished, but the barn system utilises multiple baked setups with different influences (which we build), simulating masses for each section of the barn, then tieing it into the one main dust & debris setup. Similar approach, but not as "accurate" due to it not directly referencing the source geometry volumes; if I'd had incorporated that, it would have taken up about 1/3 of the book hahahaha.

There's a bit of pflow scripting, but due to space it's mainly a case of loading in pre-designed scripts. However are pretty basic and are just there to extend functionality, though their useage and structure are explained in detail.

pete

JohnnyRandom
06-16-2008, 09:42 PM
...pre-ordered via amazon.com a minute ago...

LOL me too:D

Pete, I was hoping you were going to say release it at Sig, then have a book signing;):D

Cheers Pete your books are a great help! Really, really looking forward to seeing how you did the tornado shot, we've been bouncing around for a couple weeks off and on trying to figure out a good way to accomplish it, Is it built up different than in the 1st edition? looks like it(i don't have the second ed.)

PeteDraper
06-16-2008, 10:10 PM
yeah it's totally different to the one in the first edition... that one was childsplay compared to this one! The main effect is all down to the geometry animation baking systems and the dust and debris plays off that which feeds back into the main system that drags up debris from the ground, churns it up, then dumps it back down again, constantly checking to see whether it needs to be picked up again incase it got dumped in the tornado's path :) Bit of a heavy one... when I've taught this it's normally taken an entire day to teach as a workshop :)

pete

PS yeah was hoping to get across to sig... not gonna be the case I don't think this year due to one reason or another...

Michael-McCarthy
06-16-2008, 10:14 PM
Very cool Peter! I cant wait for the new addition. I always recommend your books to students and its super that the older title will be a free ebook too.

Thanks
Michael McCarthy
cebas USA

PeteDraper
06-16-2008, 10:24 PM
oh yeah totally forgot, on something completely unrelated, there's a nice ice comet breakup system I wrote for a forthcoming 3dworld issue. The system's pretty straight-forward for you guys but there's also some heavy procedural modelling in the base scene (not mentioned in the tutorial due to space, but the entire stack is there for you to pull apart if ya need to :) )

http://www.xenomorphic.co.uk/images/comet_ice.jpg

the geometry doesn't actually fragment apart from the three main keyframed chunks, its actually multiple systems which interact with each other and proxy geometry creating the nice distribution of debris, in essence automating the system :)

pete

TwiiK
06-16-2008, 10:52 PM
Hang on...

I'm not done with the second book yet. :)

Tried for months to succesfully order the second book, but apparently the norwegian postal system hates me and it never arrived. When I finally recieved it I was knee deep in work and have barely had any time for 3d since.

Even so, I'm nearing completion of the fire based tutorials now, trying to personalize each effect as much as possible, and underway taking notes as I go. The idea was that when I finish all the tutorials, to compile all the notes into questions which I'll fit into a mail headed your way. :)

Looking forward to the next book and that 3d world issue as well.

PsychoSilence
06-17-2008, 03:47 PM
Effects guru Stan Winston dies of cancer at 62

http://www.msnbc.msn.com/id/25196077/

:cry::cry::cry:

Glacierise
06-17-2008, 04:18 PM
That's sad. He did make movie magic. Rest in peace.

PsychoSilence
06-17-2008, 05:16 PM
With the hype around Speed Racer these days I recently finished a tutorial about procedural carpaint in Compositing:
http://www.cgarena.com/freestuff/tutorials/max/procarpaint/index.html

itīs pretty basic and only shows off the idea...hope u still like it :)

PeteDraper
06-19-2008, 12:29 PM
nice one psychosilence :)

Very very sad news to hear about Stan Winston... a legend has passed.

Pete

PeteDraper
06-19-2008, 12:31 PM
finally managed to update my reel for the first time in two years!!

http://forums.cgsociety.org/showthread.php?f=154&t=644469

have a butchers and see what ye thinks - also some of the DTE v3 stuff in there too :)

cheers!

Pete

TunnelLight
06-19-2008, 02:24 PM
Howdy Pete, How goes it?

That show reel looks awesome, DTE 3 stuff looks good too. Still waiting for the ok on the stuff you did for us for your reel, but I let you know as soon as I do.

Catch you later.
Mark

PeteDraper
06-19-2008, 02:41 PM
allright mark!! Long time no heary :) Cheers for the comments mate! Yeah no probs on the stuff... whenever it gets released :)

hope all's well down saaaf... bloody chilly oop here...!

cheeeeeers!!

Pete

nitrocom
06-19-2008, 02:41 PM
hey pete,

very good reel, especially i like the intro (sun blows out) it's impressive...

PeteDraper
06-19-2008, 04:44 PM
arggh... double post (bloomin' firefox)

cheers nitrocom! :)

PexElroy
06-19-2008, 09:38 PM
great reel Pete, looking forward to that book's release -- it has a lot of slick material ;)

PeteDraper
06-21-2008, 11:38 AM
cheers pexelroy :) looking forward to finishing the thing :D

PeteDraper
06-24-2008, 01:32 PM
on a completely unrelated note, anyone in the UK going to Glastonbury this weekend to study computational fluid dynamics at the bottom of a pint?

gimme a yell if you are

pete

TunnelLight
06-24-2008, 02:28 PM
I'm sorry to say Pete but JayZe is going to be keeping me away, besides computational fluid dynamics is best observed multiple times in an indoor environment with GI also known as my local.

Catch you later dude

Have fun a Glastonbury

PeteDraper
06-25-2008, 10:22 AM
heheheh nice one mate :) have a good weekend everyone and I'll catchya next week :)

pete :beer::beer::beer:

PsychoSilence
06-26-2008, 11:10 AM
http://www.the123d.com/inside/staying_inspired/staying_inspired_0.shtml

Nice article that hitīs in the same chamfer as mine...

i feel like itīs time for a new one too :)

Also ALLWAYS worth visiting every now and then (like weekly):

http://freelanceswitch.com

JohnnyRandom
06-26-2008, 11:15 PM
Nice read Anse:)

BTW a little something for you (not anywhere near perfect but I think a good start) I haven't tested it any other scenarios to see how well it reacts (other than adding some force ops) but anyway...

birth, pos icon, box#3 op, shape

Watch me (http://www.4rand.com/TEST/ParticleFlow/Misc/TornadoTest_01.mov):wavey:2.5mb

PsychoSilence
06-27-2008, 12:46 AM
Nice read Anse:)

BTW a little something for you (not anywhere near perfect but I think a good start) I haven't tested it any other scenarios to see how well it reacts (other than adding some force ops) but anyway...

birth, pos icon, box#3 op, shape

Watch me (http://www.4rand.com/TEST/ParticleFlow/Misc/TornadoTest_01.mov):wavey:2.5mb

Itīs always john or charley that shoot from an unknown angle :) Adding some force and collision is all cenario it needs i suppose. For now it moves a bit toony and fast but i guess with lil tweaks a touch down cenario would be easy to achieve :) If that way is faster then controlling and locking particles to a loft i guess it is the future way using (in general) faster toolbox setups.

anselm

PsychoSilence
06-30-2008, 04:21 PM
http://www.vfxhack.com/2008/06/27/are-you-in-the-money/

http://animationguild.org/_Home/home_FRM1.html

http://www.vfxhack.com/wp-content/uploads/2008/06/008-wages.jpg

PeteDraper
07-02-2008, 02:03 AM
interesting read... yeah it kinda translates over here too... for anyone interested in uk rates, check out BECTU.

Pete

PeteDraper
07-02-2008, 02:07 AM
as I pulled the fire image from the samples I posted not so long ago, I thought I'd update with the new one. This is the final "Taken Further" result with motion blur and additional fresnel shading on the edge of the fire, though for those who saw the original sample will notice that the simulation and colour have also been tweaked to achieve a more gaseous fuel source effect.

Again, 100% vanilla 3ds max, no plugins (ie particle flow, blobmesh, scanline renderer)

Pete

SoLiTuDe
07-02-2008, 03:38 AM
Again, 100% vanilla 3ds max, no plugins (ie particle flow, blobmesh, scanline renderer)

Pete

BOOOOOO!!! Pwrapper > Blobmesh. :D FumeFX > Pwrapper :D

(awesome looking result Pete!)

frogspasm
07-02-2008, 05:42 AM
Whoa! That's pretty impressive for just using Max out of the box.

I'll have to start playing around with Pflow again one of these days! (It's been a while)
When's that book coming out? :)

(And how was Glastonbury? I always get so jealous of the UK around this time of year :P
Festivals up the wazoo...)

superhypersam
07-03-2008, 03:05 AM
ya this method is pretty cool.

though pwrapper is really worthwhile replacement for blobmesh

from a couple of yrs ago, but i still like.
www.giantapestudios.com/Anim/Glu_fire.mov (http://www.giantapestudios.com/Anim/Glu_fire.mov)

I dont think I would use this method anymore, not with fume around.

PeteDraper
07-03-2008, 03:24 AM
yep, Id definately agree on all points hands down (though it's difficult to do plugin-free tutorials with plugins heheh :) )

blobmesh is evil

evil evil evil

I think I've even mentioned that in the text...

pete

Steve Green
07-03-2008, 07:38 AM
I couldn't believe how much quicker PWrapper was than Blobmesh when I downloaded the demo of Glu3D. I remember there being a few speedier versions out there - Habware? but this was something else.

- Steve

PsychoSilence
07-03-2008, 03:27 PM
Yeah! The latest built is multithreadded finally :) I bought both Glu3D and PWrapper about 3-4 Month ago and really enjoy it despiter the fact i never did any crazy simulations tho. I had trouble "calming/settling" the water with the new version but that might be fixed soon...same goes for mixing two fluids.


Back in the days they sold glu and the wrapper togwther for like 300 bucks if i remember right...Now it's 500 for the glu and 300 for the wrapper...BUT PWrapper alone (especially because there is a TP connexion if u ask 3daliens kindly...) is woth it's money. I basicly bought glu and wrapper for one single 5 day remote freelance job i did from home on my notebook and it payed off the very first day :)

cheers,
anselm

Nazgul
07-03-2008, 08:57 PM
Superhypersam,

Ya you're absolutly right. For off the shelf software, nothing beats FumeFX for fire.
But you have to agree that the fire you achieved with PWrapper is pretty stylish!

It's kind of hard to style FumeFX stuff...it tends to always look the same. Perhaps if they could provide us with with different shaders...

PsychoSilence
07-17-2008, 10:02 PM
I am very disappointed...:wip:

...well, maybe you guys are just as busy as me!



In the meantime some fun with the network:
http://youtube.com/watch?v=nGtWYuJ5f64

ArtiZta
07-19-2008, 10:53 AM
Hi Guys,
It's been a while since I've posted a question in here...
Now I'm wondering has anyone found a workaround for UDeflector so we can have particle collide with and custom shaped mesh without taking all the system resources?. (static shape, not moving or deforming)

I came to think if there is a script to layout/align many standard Deflector's to meshes faces. Is that possible?

Cheers

Glacierise
07-19-2008, 11:20 AM
Just model a proxy and use it as a deflector ;)

SoLiTuDe
07-19-2008, 11:32 AM
he means use a very low poly version of your mesh instead of the actual mesh.

ArtiZta
07-19-2008, 11:32 AM
Just model a proxy and use it as a deflector ;)

Sorry Glacierise, you got me lost there..
can you please explain more?

Glacierise
07-19-2008, 12:15 PM
Say you have a scyscraper, modeled with all the details, windows, arcs etc. Say you have a city destruction scene. Now do you really wanna collide the particles with the individual windows and details? Probably no. So you make a box over the skyscraper, put some detail into it, tweak the verts until you have a good enough resemblance of the skyscraper, and use that as a collision object.

JohnnyRandom
07-19-2008, 07:22 PM
Psychodude made me feel bad for not posting in here forever.

A little r&d been working on, still trying to figure a few things out but anyway...

http://www.4rand.com/TEST/ParticleFlow/Box3/Rebuild/rebuild_thumb.png

Rebuild.mov 7.5mb QT (http://www.4rand.com/TEST/ParticleFlow/Box3/Rebuild/rebuild.mov)

Bobo
07-19-2008, 09:17 PM
In the meantime some fun with the network:
http://youtube.com/watch?v=nGtWYuJ5f64

So THAT is how you crash Deadline? ::applause:
I wonder how I missed the fun...

JohnnyRandom
07-19-2008, 10:20 PM
LOL

I have to say chucking pieces of hardware off a four story building looks fun... wonder if they can fix that hardware in post:scream:

ArtiZta
07-20-2008, 07:43 AM
Say you have a scyscraper, modeled with all the details, windows, arcs etc. Say you have a city destruction scene. Now do you really wanna collide the particles with the individual windows and details? Probably no. So you make a box over the skyscraper, put some detail into it, tweak the verts until you have a good enough resemblance of the skyscraper, and use that as a collision object.

Ok, my case is, I have a bended cylinder, like tunnel in which I will have particle move through but with gravity so they have to collide with the inner walls of the tunnel. Now to achieve that I'm using UDeflector, even if I really simplify the tunnel mesh it's still too heavy, especially coz I have to use thousands of particles, and still couldn't find a workaround for this.

CapitanRed
07-20-2008, 02:29 PM
I'd suggest to not having all the particles go trough all the tunnel. make them born behind the camera (I assume that your camera is going trough the tunnel like one of the particles), and delete them when they are far enough from camera...

If you have a view from the outside, born them outside the view, and delete on the other side. or use camera culling. and if there are to much particles, make only a part of them colliding, and mask the not colliding particles in post.

you can do the same for the inner view: particles that are far away don't have to collide with the cilinder.

ArtiZta
07-21-2008, 06:07 AM
Thanks Gorane
I did try it out... but I couldn't manage to lighten the system. Still working on it..

Btw, I saw Bobo's post in other thread about Tool Box #3 performing faster is some cases, I was wondering if he passes by this post, how does Box #3 perform with deflectors especially UDeflector?

Glacierise
07-21-2008, 11:33 AM
Bobo said it performs faster then the same thing done with scripting. As far as I know Box#3 doesn't optimize collisions. If I were you, I'd use small amount of particles to get the motion I look for, then increase the number of particles and cache them, so you can watch the cached animation, adjust, cache again. You should have done most of the work in the phase with few particles, so you won't have to tweak too much in the second phase, and it will be bearable :)

Bobo
07-21-2008, 03:23 PM
Bobo said it performs faster then the same thing done with scripting. As far as I know Box#3 doesn't optimize collisions. If I were you, I'd use small amount of particles to get the motion I look for, then increase the number of particles and cache them, so you can watch the cached animation, adjust, cache again. You should have done most of the work in the phase with few particles, so you won't have to tweak too much in the second phase, and it will be bearable :)

I did not say that. I said it is faster doing geometry lookups like the old "Position Object" operator does. It IS accelerated when dealing with heavy geometry. One could write collision tests using the Data Test operator and those would be faster, too.

ArtiZta
07-21-2008, 05:03 PM
If I were you, I'd use small amount of particles to get the motion I look for, then increase the number of particles and cache them, so you can watch the cached animation, adjust, cache again. You should have done most of the work in the phase with few particles, so you won't have to tweak too much in the second phase, and it will be bearable :)

Yes, that's it the only technique I use till now coz I couldn't find any other workaround. Still it's time consuming. I'll just have to face it.

Bobo, so there might be a chance that box #3 can perform some faster collusion events then??..

Glacierise
07-21-2008, 05:14 PM
Hey thanks Bobo, I didn't know that, I stand corrected.

Bobo
07-21-2008, 05:27 PM
Hey thanks Bobo, I didn't know that, I stand corrected.

Don't give up so quickly - I just wrote a Collision DataOp in Box #3 and it turned out not to be faster than a UDeflector+Collision Test. In fact, it was slower. Of course, it might be related to my relative inexperience with Box #3, as I am still learning all the fine details.

But we are working on our in-house collision operator and the early tests were encouraging.
I created a simple "paddle" test where lots of particles fall onto a highly tessellated box moving up and down, bouncing the particles away.

Here are some benchmarks - they list the time needed to play the flow without collisions, then the Krakatoa Collision Test time, the UDeflector+PFlow CollisionTest time and the speed up factor.


0-100 frames, particles emitted on frame 0:

--Step: 1 frame, 100K particles, 12 faces
--NO COLLISION: Time: 4.953 sec.
--KRAKATOA COLLISION: Time: 11.5 sec.
--PFLOW COLLISION: 13.406 sec.
--Factor: 1.16574

--Step: 1 frame, 100K particles, 1,200 faces
--NO COLLISION: 4.953 sec.
--KRAKATOA COLLISION: 14.734 sec.
--PFLOW COLLISION: 230.266 sec.
--Factor: 15.6282

--Step: 1 frame, 100K particles, 12,012 faces
--NO COLLISION: 4.953 sec.
--KRAKATOA COLLISION: 23.328 sec.
--PFLOW COLLISION: 2328.34 sec.
--Factor: 99.8088


As you can see, the PFlow Collision operator hates high polygon counts because it does brute force processing. So with 100,000 particles and only 12 faces, it is almost as fast as our operator, but with 12,000 faces, it is 100 times slower. It appears to scale linerarly to the number of faces, so from 1,200 to 12,000 faces the PFlow Collision goes from 230 to 2300 seconds.

We hope to have this new operator in Krakatoa 1.2.

Glacierise
07-21-2008, 05:54 PM
That speed improvements are dramatic! At least! And the way it scales is very impressive! Quite a big feature for the next Krak! And what happens if the mesh is deforming (i.e. characters)?

JohnnyRandom
07-21-2008, 06:01 PM
Wow that is a huge difference.

I am immensely curious about this operator you are working on:) any more details?

Bobo
07-21-2008, 06:14 PM
That speed improvements are dramatic! At least! And the way it scales is very impressive! Quite a big feature for the next Krak! And what happens if the mesh is deforming (i.e. characters)?

There are two sides to this - on the one hand, it would have to reload the geometry into the acceleration structure on each frame the geometry is changing, which is typically the time-intensive part. Then it has to take into account the velocity of the geometry surface to deflect the particles correctly, which is still left to be done.

That being said, we are doing some heavy character work on a movie using Krakatoa and its PFlow operators, so there should be some more surprises in 1.2 related to characters and particles...

Bobo
07-21-2008, 06:18 PM
any more details?

Sure. Here is the rest of the data, minus the PFlow Collision Test times, as I had no time to wait to measure those ;)
Still the same setup, played over 100 frames:

--Step: 1 frame, 100K particles, 120,000 faces
--NO COLLISION: 5.797 sec.
--KRAKATOA COLLISION: 116.031 sec.

--Step: 1 frame, 100K particles, 480,000 faces
--NO COLLISION: 5.938 sec.
--KRAKATOA COLLISION: 463.718 sec.

--Step: 1 frame, 1M particles, 1200 faces
--NO COLLISION: 40.75 sec.
--KRAKATOA COLLISION: 132.422 sec.

JohnnyRandom
07-21-2008, 06:38 PM
minus the PFlow Collision Test times, as I had no time to wait to measure those ;)


That almost isn't funny:D

The 1200 faces 100K vs. 1m is interesting, time equates to actually less than tens times.

I have to say impressive work, it's nice to see some improvements in the collision arena.

ArtiZta
07-21-2008, 07:48 PM
That sounds great!
finally some hope to have collision lighter then using UDeflector?
can't believe how far it goes against pflow collision. So, that Krakatoa Collision, is it going to be usable with non Krakatoa renderers??

I think it's about time Max does something new to its particle system... it's been ages since there's been something new.

Bobo
07-21-2008, 07:51 PM
That almost isn't funny:D

The 1200 faces 100K vs. 1m is interesting, time equates to actually less than tens times.

I have to say impressive work, it's nice to see some improvements in the collision arena.

Actually it is almost 10 times slower, particle count changes result in linear scaling.

14.734-4.953 = 9.781 (100K + 1,200 faces)
132.422-40.75=91.672 (1M + 1,200 faces)
Factor: 9.37246

Here I am removing the non-collision time (spent redrawing the particles in the viewport and in the rest of the operators), so the pure Krakatoa Collision Test times for 100 frames are 9.781 and 91.672 respectively.

PsychoSilence
07-21-2008, 07:55 PM
I think it's about time Max does something new to its particle system... it's been ages since there's been something new.

Itīs been taken care of :) i guess Siggraph will reveal some new stuff thatīs going on right now...maybe Oleg can join in and please his followers with news :)

Bobo
07-21-2008, 08:54 PM
So, that Krakatoa Collision, is it going to be usable with non Krakatoa renderers??

All Krakatoa PFlow Operators come with the evaluation version of Krakatoa and can be used as general operators in any flow, with any renderer. The only limitation of the evaluation version: you cannot net-render with them unless you buy a license of Krakatoa. (Krakatoa itself does not support network rendering in evaluation mode and adds a watermark on the image, otherwise all other features are available). So if you work with PFlow on a single workstation, the Krakatoa PFlow operators are sort of bonus tools in evaluation mode. If you need to render on the network, you would have to buy a license.

For example, the Geometry Test we currently ship with Krakatoa could be used to find particles inside a volume. See this tutorial for ideas (http://www.franticfilms.com/software/support/krakatoa/using_krakatoa_geometry_test_operator.php). Both Box #1 and Box #3 have such a feature, but are slightly slower.

Benchmark - 1M particles in a 120x120x120 box icon volume tested against a default Geosphere with Radius 50 (4 segments Icosa, 320 faces) on a single frame:

Box #1 Group Selection > Inside Object > Split Group : 14 seconds
Box #3 DataOp (Select Object > Inside Objects > Output Test): 9 seconds
Krakatoa Geometry Test: 6.1 seconds

Same with 1M particles and 2K faces (10 segments):
Box #1 Group Selection > Inside Object > Split Group : 72 seconds
Box #3 DataOp (Select Object > Inside Objects > Output Test): 17 seconds
Krakatoa Geometry Test: 6.8 seconds

Same with 1M particles and 8K faces (20 segments):
Box #1 Group Selection > Inside Object > Split Group : 278 seconds
Box #3 DataOp (Select Object > Inside Objects > Output Test): 41 seconds
Krakatoa Geometry Test: 7.1 seconds

DISCLAIMER:
I would like to point out again that the PFlow Tools Boxes are highly flexible while our simple PFlow operators are highly specialized to deal with large particle counts. The above comparison is in no way an attempt to badmouth the Boxes as they work great in combination with Krakatoa and I couldn't live without Box #3 DataOps. So please don't go around saying "Bobo hates Orbaz" ;)
Hat off to Oleg - I greatly enjoy working with DataOps and PFlow in general. :bowdown:

JohnnyRandom
07-21-2008, 09:14 PM
Actually it is almost 10 times slower, particle count changes result in linear scaling.


It seems, I could be wrong I haven't clocked it but the more particles you add with a vanilla collision test the time seems to go up more exponentially than linear.

The link you posted, the final result is sweet.

Bobo
07-21-2008, 09:17 PM
It seems, I could be wrong I haven't clocked it but the more particles you add with a vanilla collision test the time seems to go up more exponentially than linear.

The link you posted, the final result is sweet.

I was talking about the Krakatoa Collision Test scaling linear with particle count (each particle has to shoot a ray, each ray intersection takes about the same time).
The PFlow Collisions ARE exponential, as each particle has to run through all faces to find the one, just like the MAXScript IntersectRay().

JohnnyRandom
07-22-2008, 12:31 AM
Yeah I know, I was comparatively speaking :)

Can't wait to try it.

Glacierise
07-23-2008, 05:17 PM
Itīs been taken care of :) i guess Siggraph will reveal some new stuff thatīs going on right now...maybe Oleg can join in and please his followers with news :)

Someone's being a naughty teaser - no presents for Christmas :D

PeteDraper
07-23-2008, 05:40 PM
i've been in conversation with the guys over at 3daliens for the last couple of days and they've pulled a late one and managed to get particle flow scale referencing sorted out in pwrapper after I kicked up a bit of a stink about it :blush:

The current version works best for particle scale values under 100% (some issues with surfacing with scale values over 100%), however I've done a quick test and if you scale the particle down and increase the size of the shape before applying any scaling it works fine. The next version will apparently have the +100% scale properly sorted.

Hopefully they can also implement static particle issues in the next one too... ie if a particle isn't moving, don't update the mesh within a threshold level (to prevent mesh flickering)

Fingers crossed :thumbsup:

Pete

PsychoSilence
07-23-2008, 05:53 PM
i've been in conversation with the guys over at 3daliens for the last couple of days and they've pulled a late one and managed to get particle flow scale referencing sorted out in pwrapper after I kicked up a bit of a stink about it :blush:

The current version works best for particle scale values under 100% (some issues with surfacing with scale values over 100%), however I've done a quick test and if you scale the particle down and increase the size of the shape before applying any scaling it works fine. The next version will apparently have the +100% scale properly sorted.

Hopefully they can also implement static particle issues in the next one too... ie if a particle isn't moving, don't update the mesh within a threshold level (to prevent mesh flickering)

Fingers crossed :thumbsup:

Pete

i want a glu velocity op next!!! :)

http://forums.cgsociety.org/showthread.php?f=206&t=655530

JohnnyRandom
07-30-2008, 09:28 PM
7 Days no post so here some great news for all of us...and in case you haven't heard elsewhere...

Basically optimized and integrated Box#1 into standard particle flow:beer:


Advanced Particle Effects

Incorporate sophisticated particle effects into your scenes. PFlowAdvanced includes 12 operators new to 3ds Max Design, giving you easier access to a whole new range of creative options. The addition of the PFlowAdvanced technology to 3ds Max not only greatly enhances the functionality of Particle Flow, but it also gives you a high-performance, streamlined 3ds Max workflow, through which to access its features.


- Numerous paint tools enable you to precisely place particles in your scene. Operators—including Particle Paint, Birth Paint, Placement Paint, Birth Texture, and Mapping Object—enable you to paint both particles and emitters onto specific parts of your objects. Other special effects include the ability to emit particles based on animated color data in emitter objects.
- The utility operators have been integrated into the 3ds Max core to extend Particle Flow's capabilities, while giving you a streamlined workflow for creating particle effects. Clean up particle flow, synchronize layers, repair the cache system, save a custom flow as a preset, and automatically delete operators that are not in use and lock or bond particles to an object (improves on the former position object operator).
- The ability to group particles enables you to perform specific operations on arbitrary subsets of particles. Quickly create multiple groups and work with as many of them as you like through such operators as Group Selection, Group, and Split Group
- The PFlowAdvanced Shape operators greatly expand the Particle Flow Shape toolset. Define a particle shape beginning with a pre-existing shape—a vast range of 3D and 2D shape presets are available to choose from. Additionally, the Fast Shape Evaluation operation evaluates a reference mesh object on the final frame quickly and accurately, significantly improving performance.
- Workflow enhancements include an Express Save operator that has been integrated with the 3ds Max auto-save functionality, backward compatibility with Orbaz Particle Flow Tools, Box # 1-based scenes, and general memory and performance improvements based on source code optimizations.

PsychoSilence
07-30-2008, 10:25 PM
maybe i should people let peak at my animated collision object...

http://www.3delicious.de/spielwiese/animated_collision_box3.avi

working on that one...still not implemented: speed by grayscale of collision object*puts down note*

amckay
08-05-2008, 06:55 PM
great work mate, I rarely will say this to a guy but nice box!(es)

So SIGGRAPH is around the corner, whos going this year? I'm flying out next Saturday from Vancouver I think there should be some good parties this year

Shoot me an email guys if you want to grab a beer amckay@allanmckay.com

Phibmobil
08-05-2008, 08:08 PM
i remeber how i cried the first time i realsed you cant deflect off an animated surface.. good job =) ... are those boxes rigid ?

loran
08-06-2008, 08:06 AM
I ll be at Siggraph. I hope to meet you Allan, maybe at Autosesk users party ?

PsychoSilence
08-06-2008, 08:35 AM
I wouldnīt be at Siggraph this year (again) :cry::cry::cry:

Loran, i got your mail but havenīt had the time to reply yet :( since i have guests in my house :)

Glacierise visited me here in Hannover and we along with former Soulpix members went to a party around our city lake:

http://3delicious.de/spielwiese/array-of-nerds.jpg

Left to right: Geri,Anselm,Hristo,Christian,Kai



@ Phibmobil: it is all box3, no RBD here...

Glacierise
08-07-2008, 12:34 AM
Affirmative, not all Anselm's photos are fake! :D I'm back from my little trip to fabulous Germany. Wacken - I'm still bruised, Dortmund, Hamburg and Hanover were awesome, and Anselm is a great host (don't overwhelm his place now though :D ) Don't die before seeing that, would be my advice. Back to CG thrills now, and awaiting the news-y fireworks at Siggraph, cheers guys!

JohnnyRandom
08-07-2008, 05:01 AM
Dunno, looks like you guys had some fun:D

PsychoSilence
08-07-2008, 09:11 AM
yeah my place sucks, iīm never there for a reason :D

so brando is holding a masterclass about water and pee, and allan is holding one about steaming poo he promised me ;) *wink*. best of both worlds :D
.
.
.
naaah, i bet this year liek any other rocks. i wish i could get my hands on brandoīs battle masterclass really!

PsychoSilence
08-08-2008, 08:23 AM
Hey guys since yesterday im at facebook too :D

Ansi @ facebook (http://www.new.facebook.com/profile.php?id=754050535&v=photos&viewas=1416564103#/profile.php?id=1416564103)

so if youīre there too letīs make friends :)

another classic from glacieriseīs visit:

http://www.3delicious.de/spielwiese/array-of-nerds2.jpg

feldy
08-08-2008, 07:44 PM
hey why does your facebook take me to a drew gunning lol

PsychoSilence
08-09-2008, 12:15 PM
oh foRRRk! thats one of my friends there...the product manager of turbosquid actually.

try that one:
http://www.new.facebook.com/home.php#/profile.php?id=1416564103

PsychoSilence
08-11-2008, 08:55 AM
[Shameless Plug]

http://forums.cgsociety.org/showthread.php?p=5316980#post5316980

[/Shameless Plug]

JohnnyRandom
08-18-2008, 08:28 PM
so brando is holding a masterclass about water and pee, and allan is holding one about steaming poo he promised me ;) *wink*. best of both worlds :D


No pee or poo

Allans - some nice workflow tricks, head was pounding from not drinking enough the night before, some interesting stuff though.

Brandons class was cool, great water tricks without a fluid plug-in, parray splash was interesting could really look believable with way low-overhead.

Bobo had a good class on something I knew very little about (hehe among other things), scripted manipulators, heh, so that's what that does:D

Had a good time, thanks, met some great folks:)

loran
08-27-2008, 02:26 PM
hey,
I am testing the Creativity Extension with the PF tool box #1 included.
I have big problems with the LOCK/BOND. many crashes when I use it. A lot of "out of memory" message. I try to put an LOCKBOND after a collision surface to tell particles to stay on the sea surface after collision but lockbond crash every time I do that :/

http://laurent.renaud.free.fr/divers/cgtalk/boatWake-01.jpg
(http://laurent.renaud.free.fr/divers/cgtalk/boatwake-01.mov)VIDEO (http://laurent.renaud.free.fr/divers/cgtalk/boatWake-01.mov)

Daniel-B
08-27-2008, 10:01 PM
Loran, for some reason the video isn't working. Also, if you wouldn't mind posting an example scene file, I'd like to try to create some of the water elements. You see, I've been playing with ways to create spray with Krakatoa, and this might be a good example to try. Here is an example of one the CG splashes I made...

http://img.photobucket.com/albums/v58/PixelMagic/splash.jpg

SoLiTuDe
08-28-2008, 04:53 AM
Loran, for some reason the video isn't working. Also, if you wouldn't mind posting an example scene file, I'd like to try to create some of the water elements. You see, I've been playing with ways to create spray with Krakatoa, and this might be a good example to try. Here is an example of one the CG splashes I made...



http://laurent.renaud.free.fr/divers/cgtalk/boatwake-01.mov <-- correct link
http://laurent.renaud.free.fr/divers/cgtalk/boatWake-01.max <-- max file

SoLiTuDe
08-28-2008, 05:00 AM
hey,
I am testing the Creativity Extension with the PF tool box #1 included.
I have big problems with the LOCK/BOND. many crashes when I use it. A lot of "out of memory" message. I try to put an LOCKBOND after a collision surface to tell particles to stay on the sea surface after collision but lockbond crash every time I do that :/


Works fine here... well it doesn't crash, but i think the settings are what they're supposed to. :D What machine specs? 32bit max or 64bit max? ...how's it hold up under 100 particles as opposed to the 100,000?

loran
08-28-2008, 08:21 AM
I work on Max32bit. It crash even with 100 particles. Every lock/bond tests crash rapidly for me. can you send me a scene wich not crash for you?

http://laurent.renaud.free.fr/divers/cgtalk/lockbond-crash.jpg

charleyc
08-28-2008, 04:32 PM
Loran - The scene in Solitude's post works fine without a crash for me. Perhaps there is an error with your creativity extension install. Have you tried re-installing it? Lock/Bond has in the past been a very stable operator and so far I have found no difference in this extension version. However, I do have Box3 installed and therefore I am not using all the same installed dll's from the extension. Perhaps that is why it is working fine here and not for you?

SoLiTuDe
08-28-2008, 05:47 PM
Loran, the file I used, was yours. :D I agree with Charley... your install must be jacked. I don't have box3 installed either, so it most definetly doesn't have anything to do with box3... however I haven't tested in 32-bit yet (I try not to open it). I can maybe test tonight in 32, but till then try reinstalling.

JohnnyRandom
08-28-2008, 06:25 PM
Woohoo, Particle Flow Tools Freebies recompiled for max 2009!

Thanks Oleg:thumbsup:

http://www.orbaz.com/forum/viewtopic.php?t=1801

Daniel-B
08-28-2008, 10:21 PM
Loran, are you using Max 2009? I tried opening your scene file, and it crashed on me before the file even opened.

SoLiTuDe
08-28-2008, 10:26 PM
^ Afaik, the creativity extension (which he mentioned) is only for max2009...

JohnnyRandom
08-29-2008, 12:39 AM
Loran, it works here too (max 2009 32-bit) albeit slow as f##k, it works. :D

Although at frame 71 max is already using 1.2 gigs of memory.

Having such a big high poly collision object is going to screw you big time, you need to do something about that. Either make an adaptable collision object the is only mid/high poly near and around the boat or think of something else. Max is trying to figure out collisions for that huge 80,000 face collision object. :)

Should have went to Brandon's water masterclass :D he built a bitchn' adaptable sea surface that is only high poly in the camera view. Could do the same thing with your deflector object I would imagine. Basically works like this...sea surface + subdivide + volume select + delete mesh, rinse & repeat. A extruded shape object controls the volume select and deletes the rest, giving you whats left in the camera view to calculate. I image you could do the same with the deflector around the boat. So a two copies of the sea surface one for looks the other cut down as a deflector. Should save you some memory space.

Loran, something like this, it could use some more love but this is basically what I am talking about. The whole timeline is now using under a gig of memory. Hope this helps:)

PsychoSilence
09-02-2008, 10:32 PM
Brandon Riza updated his website :)

http://brandonriza.com/3DVisualEffects/HTML/3DVisualEffects.htm

(maybe he updated it longer ago but i havent been there since a while. just in case u havent seen his new stuff go ahead :) )

JonathanFreisler
09-03-2008, 01:11 AM
wow. And crap, now i have to watch them all, and im meant to be working.

thanks for the heads up

SoLiTuDe
09-03-2008, 01:15 AM
Brandon Riza updated his website :)

http://brandonriza.com/3DVisualEffects/HTML/3DVisualEffects.htm

(maybe he updated it longer ago but i havent been there since a while. just in case u havent seen his new stuff go ahead :) )

Nah, that's a recent update... he put it up right after the war hammer trailer was released. :D

loran
09-03-2008, 10:33 AM
Thanks JohnnyRandom (http://forums.cgsociety.org/member.php?u=56504) ! You are right, collision was the problem. Thax for the scene and your good advices! You were at sigg to watch the Brandon classes? Do you know some classes will release in video??

pixel9
09-03-2008, 12:02 PM
Brandon Riza updated his website :)

http://brandonriza.com/3DVisualEffects/HTML/3DVisualEffects.htm

(maybe he updated it longer ago but i havent been there since a while. just in case u havent seen his new stuff go ahead :) )

Whow. I've not been on his site quite a few months. Great stuff!

Glacierise
09-03-2008, 12:15 PM
Hi folks, what's up! I've been pushed into some archviz jobs these days (the last ones, hopefully) but here's the opener for my reworked reel:

http://cg.glacierise.com/reel_opener.avi

7-8 pflows, 3 Fume grids, and of course Rayfire ;) Enjoy!

P.S.: Brandon Riza's new stuf is awesome! I loved that warhammer film, longer and sweeter then the other recent Blur stuff!

grury
09-03-2008, 12:56 PM
Nice one Hristo! crazy sims.
Pity the smoke didnt have a bit more detail, or mayb is just the video compression killing it.

loran
09-03-2008, 01:15 PM
wow, great opening Glacierise! I definatly have to push into fumefx!
any cool links on tutorials for PF/fumeFX?

PS: glad to see this thread rebirth

Glacierise
09-03-2008, 02:16 PM
As soon as I finish my current work, I'll make some FFX tuts. Meanwhile - go to Allan McKay's site ;)

@grury: which one are you talking about? The wave or some of the other smokes?

grury
09-03-2008, 03:06 PM
The huge wave coming from the right.

JohnnyRandom
09-03-2008, 06:03 PM
Thanks JohnnyRandom (http://forums.cgsociety.org/member.php?u=56504) ! You are right, collision was the problem. Thax for the scene and your good advices! You were at sigg to watch the Brandon classes? Do you know some classes will release in video??

No worries Loran, thank Brandon for the idea:)
Yeah, was cool to watch some particle masters at work:D. I would assume they will put them to DVD or E-learning, they did record them all.

Hristo, nice dude! :thumbsup:

I think it is the video compression killing some of the effect, divx has a hard time defining a lot of greys to blacks at higher compression ratios, probably the biggest downfall of mpeg based codecs.

Brandon Riza's an insane FX machine, pretty damn good photographer too:D

Glacierise
09-12-2008, 07:49 AM
No words for box#2 yet, ha? I wonder if after the last successful colaboration with box #1 AME is going to pull Oleg in and put box#2 in max 2010 in April :)

JonathanFreisler
09-12-2008, 02:57 PM
if anything we will see box tools 1 (creativity expansion) standard in the next max and perhaps box tools 3 implemented in and not as an expansion.


I think the futures looking good for Oleg, well i hope so!

Glacierise
09-12-2008, 03:17 PM
Box#3, being quite a niche specialized tool would be a strange thing to put straight into max, but who knows? I hope the future's good for Oleg too, so he can bring us particle goodness for some time to come :)

charleyc
09-12-2008, 05:32 PM
Box3 is actually a huge compliment to PFlow and really needs to be adopted by more users. It seems a lot more complicated than it really is, but its usefulness is very broad. Many of the questions that get asked in this forum have very simple box3 solutions. I don't know whether or not it is advantageous for Oleg to have these tools integrated (as opposed to selling them as plugins), but if it is good for him, Box3 would be a very good tool to get in the hands of more people.

PsychoSilence
09-19-2008, 12:50 PM
Many of the questions that get asked in this forum have very simple box3 solutions.

ppl dont know because u havenīt done a training DVD yet :D i try to built more and more operators from sretch these days...some are waaaay faster then the shipped operators.# (simple collision tests e.g.)

cheers,
anselm

(one week no posting here...i should post more :D )

Glacierise
09-19-2008, 01:10 PM
We could do some kind of a fundraiser for him to do it :) With lots of steep tools it's like that - one just needs a good kickstart, and it's fun and games from then on. And I think we need the box3 kickstart hehe.

charleyc
09-19-2008, 05:22 PM
LOL, thanks guys. I am actually ready to do something like this, but I need to find a good time in my schedule where I can get started. I agree with you Anselm, I think more people would realize the need for the tool if they realized what all they can do with it.

depleteD
09-19-2008, 08:44 PM
Box3 would be a very good tool to get in the hands of more people.

Yea man, box 3 is really dope, good for passing data around and getting data from the scene to drive your particles. Box 3 I think should be staple in anyone's arsenal but I feel that it requires pretty strong math skills.

-Andrew

PexElroy
09-19-2008, 08:55 PM
Brandon's Warhammer Online work is great to see, good use of particle flow there,

charleyc
09-19-2008, 09:32 PM
Yea man, box 3 is really dope, good for passing data around and getting data from the scene to drive your particles. Box 3 I think should be staple in anyone's arsenal but I feel that it requires pretty strong math skills.

-Andrew


Yeah, it does require more thinking. But it really depends on the type of effect you are going for. It exposes some really useful data that requires almost no math knowledge to put together some really cool stuff. Particle to particle data (distance, number of neighbors...) and particle to mesh data, great stuff when creating a procedural system. But much of the vector or matrix stuff does require a lot more math knowledge. Really you just need to know what equations you need (Google :D ) and it is a fairly simple task of wiring up that equation in Box3. But even then it is not for the mathematically illiterate. And with the new subOps coming, the ability to adjust object parameters per particle really opens up the door for procedural setups.

depleteD
09-19-2008, 11:28 PM
Yea I agree, once you get a understanding of some basic equations possibilities really start opening up. I'm really excited about the per particle shape control

amckay
09-22-2008, 06:47 PM
Once you've gotten a good understanding of it you can begin building custom operators that do almost anything, and far quicker than pflow scripting can handle. So after a while you begin to build a huge arsenal of operators that speed up your workflow considerably. Its pretty amazing how complex your systems can get but be executed quite simply and intuitively by recycling custom made operators.
Box set 3 is p1mp!

PsychoSilence
09-22-2008, 09:48 PM
tried to built a custom find target today with box3 only and failed :D some days i suck :D

depleteD
09-22-2008, 09:54 PM
Turn the speed vector to the diirection between the particle and the goal point.
Drive the magnitude by the distance from the particle to the goal point.

PsychoSilence
09-22-2008, 10:03 PM
y, in theory i figured it all out, just the OP wouldnīt work :D prolyl shoudlnīt start that kinda stuff at 11PM...

charleyc
09-22-2008, 11:35 PM
Post what you have.

instinct-vfx
09-30-2008, 05:34 PM
Hey everyone,

i wondered if there's a quick and easy way to create scripted custom operators ?
Basically i'd just like to save a preset script operator.

Didnt find any info on that so i assume it doesnt exist. Just thought i'd ask.

Regards,
Thorsten

amckay
09-30-2008, 05:55 PM
Easiest way if you have pflow tools is to delete everything but that operator and go into pflow utils in the utility panel and save it as a preset, then it will show up in pflow and you can click it

if you wanted it do it outside of pflow tools, then kill everything but that, and save it out to a designated repository directory you make like C:\MyPresets\ and then make a simple script that you can save off that says something like

mergeMaxfile "C:/MyPresets/MyCustomScriptOp.max" ""

This way you can drag it off to the toolbar and click on it any time you want and it will merge it in easily enough. You can get more complicated than this, but thats the most straight forward way I can think of.

instinct-vfx
09-30-2008, 07:10 PM
Thanks Allan!

was thinking there's some workarounds. Thinking about the best way to do a script now *think*. Was thinking to make some scripted operators Easily available to Artists. If i come up with something reasonable i'll let you guys know.

Regards,
Thorsten

relief7
09-30-2008, 07:51 PM
I think another way to do it would be to directly create events / operators with a script. A similar example is also in the MAXScript reference.

ev = Event()
particleFlow.beginEdit()
op = script_operator()
op.Proceed_Script = "< YOUR SCRIPT GOES HERE >"
particleFlow.endEdit()
ev.appendAction op

You could wrap that within a macroscript section and build complete toolbars with operators that you frequently use. Or change it so that operators are inserted into the currently selected event in the PView window etc. You get the idea...

Another thing I am wondering in the context of "making scripted operators easily available to artists" is if you would like to expose parameters/variables which are used in the script to artists. Because that would be really cool. But I think that is what box 3 is meant for. Custom behaviours with changeable parameters... Plus its faster, too. So you might need to go that way if your operators should be interactive in some way.

JohnnyRandom
09-30-2008, 07:53 PM
You could/can just save the script out as a .ms and upon adding the script operator load the pflow script from the plfow mxs editor window.

Since max10 will have pflow box1 tools available (giving you the option of using presets) that may just be a better to do it (save it as a max file).

instinct-vfx
09-30-2008, 08:20 PM
Thanks again guys.

I have been thinking about adding the operators with a script to the current event. Exposing
the internal parameters is something i havent really thought about. Local variables can cause really bad slowdowns tho in script operators so i dont know if that's a too good idea.

I was thinking in general terms tho the actual use currently would be scripted ops to create krakatoa channels (velocity and alike).

Regards,
Thorsten

instinct-vfx
09-30-2008, 08:31 PM
Browsing through the MXS docs (wich i always find a tad confusing *sigh*) to find out if there's a method to get the currently selected event/action in order to be able to append an operator behind a selected one if there is a selected one.

Regards,
Thorsten

instinct-vfx
09-30-2008, 08:41 PM
Macrorecorder to the rescue ! Still havent found it in the docs...but $'Particle View Name'.selected returns the selected node

Regards,
Thorsten

Bercon
10-02-2008, 08:24 PM
Here is something I made with pflow. Had to rely on maxscript and sdk to get the exact effect I wanted. One mxs operator to bend the trails into spirals. Modifed Spawn operator to get spawned particles information of their parent so I get each individual trail colored. The actual color is applied with mxs operator that randomizes material id for each trail.

Animation:
[/url][url]http://www.niksula.cs.hut.fi/~jylilamm/Gallery/Spiral.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/pflowGrowth_06.mov)

Flow:
http://www.niksula.cs.hut.fi/~jylilamm/graphics/pflowGrowth_chart01.jpg (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/pflowGrowth_chart01.jpg)

rdg
10-02-2008, 09:11 PM
Here is something I made with pflow.
That's beautiful!
Very inspiring.

Glacierise
10-02-2008, 09:34 PM
Great work dude, veeeeery pretty :D

JohnnyRandom
10-02-2008, 10:13 PM
Excellent, the trail coloring is waaay cool:arteest:

"Modifed Spawn operator" as in modified the actual spawn operator through some techno-wizbang programming type stuff? (I ask since you were building/compiling materials not all that long ago)

PsychoSilence
10-03-2008, 12:07 AM
Very nice, Bercon!
Keep the thread alive, guys :)

iīm still dead busy relocating....iīll be back ;)

amckay
10-03-2008, 01:27 AM
hey bud how are you liking LA? I'll be in LA next thurs but for like an hour so dont think I'll have time to visit anyone!
I'll definitely over at xmas and say g'day!

bercon great work btw!

jigu
10-03-2008, 09:20 AM
wow bercon, Great work dude!

Bercon
10-03-2008, 11:14 AM
Thanks guys!

JohnnyRandom, most of the pflow operators source code comes with 3dsmax sdk so I took the sources of the existing Spawn operator and simply added few lines so it sets the Integer channel of each spawned particle the ID of their parent. Its fairly simple once you understand the basic structure of pflow.

JohnnyRandom
10-03-2008, 04:19 PM
Cool, that should be something that is included in the existing operator, great idea :)

depleteD
10-03-2008, 08:22 PM
Bercon- that is such a excellent display of aesthetic and technicall craftmanship. Top notch man.

PexElroy
10-05-2008, 12:25 PM
Bercon - very cool use of PFlow and script, and could use this flow for other cool effects and tweak it for more creative uses ;)

spaw
10-07-2008, 09:39 PM
Any chance you can post the file?


Thanks

-Michael

Here is something I made with pflow. Had to rely on maxscript and sdk to get the exact effect I wanted. One mxs operator to bend the trails into spirals. Modifed Spawn operator to get spawned particles information of their parent so I get each individual trail colored. The actual color is applied with mxs operator that randomizes material id for each trail.

Animation:
http://www.niksula.cs.hut.fi/~jylilamm/graphics/pflowGrowth_06.mov

Flow:
http://www.niksula.cs.hut.fi/~jylilamm/graphics/pflowGrowth_chart01.jpg

JohnnyRandom
10-07-2008, 10:18 PM
I concur, Mspaw, would be fun just to play with that spawn on steroids op. :)

Bercon
10-08-2008, 01:19 PM
Sure, put the BerconPFlow.dlo to your plugins folder. Requires max 2009 64bit.

http://www.niksula.cs.hut.fi/~jylilamm/misc/pflowGrowth_02.zip (http://www.niksula.cs.hut.fi/%7Ejylilamm/misc/pflowGrowth_02.zip)

ps. Why does pflow always reset scale when applying transformation matrix to particleTM? Whats the point of that?

If I write: pCont.particleTM = matrix3 [1,0,0] [0,1,0] [0,0,2] [0,0,0]
the actual value of particleTM is now: matrix3 [1,0,0] [0,1,0] [0,0,1] [0,0,0]

Is that whoever wrote the pflow script interface didn't understand what transformation matrix is all about?

EDIT: the video http://www.niksula.cs.hut.fi/~jylilamm/Gallery/Spiral.mov

3DMadness
10-08-2008, 01:48 PM
Hey Bercon, can you upload again the .mov file? I got a file not found here.

Thanks!

Flávio

JonathanFreisler
10-08-2008, 01:56 PM
wow, now thats what i'm talking about! thanks bercon, this is why cgtalk works so well, sharing and caring.

nice work

oh and oleg does occasionally linger around here, so ask him, why the TM was written like that :p

JohnnyRandom
10-08-2008, 05:56 PM
Thanks Bercon,


for i in 1 to count do (
pCont.particleIndex = i
local pID = pCont.particleID + 1
if pflow_RotAxis[pID] == undefined then (
local v = normalize (copy pCont.particleSpeed)
pflow_RotAxis[pID] = normalize ( [-v[3]/v[1], 0, 1] * (quat (random 0.0 360.0) v))
)
local s = length pCont.particleSpeed
local a = 30.0 * (pow ((pCont.particleAge as Float) / 40.0 / TicksPerFrame) 1.5)
pCont.particleSpeed = (normalize pCont.particleSpeed)
pCont.particleSpeed = (pCont.particleSpeed * (quat a pflow_RotAxis[pID])) * s

That is beautiful :) Nice motion math! (starts to try an build box#3 operator)

avolution
10-08-2008, 06:49 PM
Hi! Is this part of a script; I figure it goes in the ON PROCEEDS part of a script op.

But I have yet to get it to work; and I do not know what it does?


Thanks Bercon,


for i in 1 to count do (
pCont.particleIndex = i
local pID = pCont.particleID + 1
if pflow_RotAxis[pID] == undefined then (
local v = normalize (copy pCont.particleSpeed)
pflow_RotAxis[pID] = normalize ( [-v[3]/v[1], 0, 1] * (quat (random 0.0 360.0) v))
)
local s = length pCont.particleSpeed
local a = 30.0 * (pow ((pCont.particleAge as Float) / 40.0 / TicksPerFrame) 1.5)
pCont.particleSpeed = (normalize pCont.particleSpeed)
pCont.particleSpeed = (pCont.particleSpeed * (quat a pflow_RotAxis[pID])) * s

That is beautiful :) Nice motion math! (starts to try an build box#3 operator)

JohnnyRandom
10-08-2008, 07:33 PM
It is a pflow script op Bercon wrote.

Yes it goes in the proceed. It also appears to create the array pflow_RotAxis of the particle ID's in the initiator.

I am not great at the math but as I see it, it basically gets the particle id, age, and speed.

It alters the particles current rotation axis and moves the particle by time around the specific particles new rotation axis which get incrementaly smaller. (i'm probably stating the obvious :blush: )

I was impressed that it didn't use a sin/cos function combo but did it somehow by normalizing the speed vector and then multipling quaternions. Going to have to dig out the linear algebra book.

Jerry is going to have to give you his insight, it would most certainly better explain it than I can.

Bercon
10-08-2008, 08:26 PM
It create random vector for each particle, parallel to their speed vector (if clause), this is done once for each particle. The vector is stored into array. This is actually a bad way to do it and you should use particleVector channel instead. Anyway, then for each frame it rotates the speed vector of the particle around the random vector stored in array.

This rotates vectorA around vectorB by the given amount: vectorA * (quat angleInDegress vectorB)

Actually normalizing the speed vector might not be necessary either, it was just leftovers from slightly more complex stuff.

The angle (local a) is calculated with the formulae: 30 * (age in frames / 40) ^ 1.5

avolution
10-08-2008, 08:58 PM
Could you drop the file into an *.ms file so I can play!
If you give a max file, I can not read it as my max is accidently very old!

Thanks for the explanation


It create random vector for each particle, parallel to their speed vector (if clause), this is done once for each particle. The vector is stored into array. This is actually a bad way to do it and you should use particleVector channel instead. Anyway, then for each frame it rotates the speed vector of the particle around the random vector stored in array.

This rotates vectorA around vectorB by the given amount: vectorA * (quat angleInDegress vectorB)

Actually normalizing the speed vector might not be necessary either, it was just leftovers from slightly more complex stuff.

The angle (local a) is calculated with the formulae: 30 * (age in frames / 40) ^ 1.5

Bercon
10-09-2008, 12:10 PM
I did the fixes I mentioned.

on ChannelsUsed pCont do (
pCont.useVector = true
pCont.useSpeed = true
pCont.useAge = true
)
on Init pCont do (
seed 48257 -- To get consistent results each time the script is ran
)
on Proceed pCont do (
count = pCont.NumParticles()
for i in 1 to count do (
pCont.particleIndex = i
-- Calculate random vector only once for each particle
if (pCont.particleNew) then (
-- Random vector perpendicular to speed vector
local v = normalize (copy pCont.particleSpeed)
pCont.particleVector = normalize ( [-v[3]/v[1], 0, 1] * (quat (random 0.0 360.0) v))
)
-- Calculate rotation amount
local a = 30.0 * (pow ((pCont.particleAge as Float) / 40.0 / TicksPerFrame) 1.5)
-- Perform rotation on speed vector
pCont.particleSpeed = pCont.particleSpeed * (quat a (pCont.particleVector))
)
)
on Release pCont do ()

EDIT:

Something totally unrelated to above, some videos of my next project.

The vids show the progress of the effect. If you don't care about how its progressed just watch the final two (06 and 07).

http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_01.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_01.mov) (~5 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_02.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_02.mov) (~5 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_03.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_03.mov) (~10 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_04.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_04.mov) (~5 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_05.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_05.mov) (~5 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_06.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/boxGrowth_06.mov) (~10 MB)
http://www.niksula.cs.hut.fi/~jylilamm/graphics/boxGrowth_07.mov (~10 MB)

avolution
10-10-2008, 03:32 PM
This was done with particles but not particle flow.
I used simcloth ala old version plus render script called
"pollen" by the great Alex McCleod

http://agilemedia.biz/mark/spiral_test.zip

PsychoSilence
10-14-2008, 03:12 AM
thanks for sharing the script, Bercon :) seriously gotta try this one day.

JohnnyRandom
10-14-2008, 05:19 PM
^looks who's back :D

mustan9
10-14-2008, 05:31 PM
That is a really cool effect.

Thanks for sharing that Bercon.

It would be good to control the size of the boxes by the amount of curve on the surface of the target object.

PsychoSilence
10-14-2008, 05:38 PM
l try to stay in touch more :) iīll move this saturday and still need to order a bed and request the electricity to be turned on. candle light is nice and stuff but... :D

JohnnyRandom
10-14-2008, 05:51 PM
Grab a fire extinguisher too, I saw those fires burning on the news :sad:

Good thing, that is some great reference material :)

spaw
10-15-2008, 06:49 PM
Can anyone please put together a max 9 32bit scene file using Bercon's script. Im not getting a similar motion.

Thanks

-Michael

Bercon
10-15-2008, 07:45 PM
spaw, hmm.. The script should work just fine in all platforms. What kind of motion are you getting then?

Here is something I've been working on for few days.

http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/BerconPFlowSpline04.jpg

And movie: http://www.niksula.cs.hut.fi/~jylilamm/graphics/FistOfTheBercon_01.mov

I edited the Spawn on Collision operator a little so it stores parentID and surface normal to newly spawned particles. This allows me to send these particles in circles along the surface. Then I wrote new spline object (or modifier) which creates splines from particles for specific PFlow events. The animation is just an example. I'm after some kind of lightning effect along the objects surface. Writing the spline stuff was a nightmare... Fortunately I managed to steal big part of the UI/PFlow connection code from BlobMesh.

spaw
10-15-2008, 08:38 PM
Wow! thats great, hope you have a chance to release some of these when you get the chance. As for the script I am not seeing a thing, all the particles are disapering post creation. Im sure its a simple user error but any sample scene that works would help.

Thanks

-Michael

grury
10-15-2008, 09:45 PM
Here is something I've been working on for

I edited the Spawn on Collision operator a little so it stores parentID and surface normal to newly spawned particles. This allows me to send these particles in circles along the surface. Then I wrote new spline object (or modifier) which creates splines from particles for specific PFlow events. The animation is just an example. I'm after some kind of lightning effect along the objects surface. Writing the spline stuff was a nightmare... Fortunately I managed to steal big part of the UI/PFlow connection code from BlobMesh.

Genious dude! Endeless hours of fun.

Heipa!

charleyc
10-15-2008, 11:50 PM
Bercon

That is great stuff. What are your plans for the particle spline tool?

JohnnyRandom
10-16-2008, 01:35 AM
Bercons, going off, going to have a whole set of modded operators:D

Wow! thats great, hope you have a chance to release some of these when you get the chance. As for the script I am not seeing a thing, all the particles are disapering post creation. Im sure its a simple user error but any sample scene that works would help.

Thanks

-Michael

Includes the original script and the updated script op. :)

amckay
10-16-2008, 01:59 AM
wow thats awesome

ice-boy
10-16-2008, 10:21 AM
can we use random opacity on the particles? are there any settings?

JonathanFreisler
10-16-2008, 12:03 PM
you could use a material frequency with a multi sub that has a few materials with different opacity.

Bercon
10-16-2008, 12:06 PM
Writing shader that takes particle id and randomizes color based on it shouldn't be too hard either but that does requires some knowledge of sdk.

EDIT: Actually that would a stupid way of doing. The best way is to randomize UVW coordinates for particles and then apply gradient map on them.

ice-boy
10-16-2008, 12:23 PM
The best way is to randomize UVW coordinates for particles and then apply gradient map on them.i am not very good at 3ds max. so what should i search in google to make this? or is there a tutorial?

spaw
10-16-2008, 08:04 PM
Thanks for the file!


-Michael

Bercons, going off, going to have a whole set of modded operators:D



Includes the original script and the updated script op. :)

Bercon
10-18-2008, 09:05 AM
I'm getting closer to the effect I was after, however I still want more electric-like effect.

http://www.niksula.cs.hut.fi/~jylilamm/graphics/FistOfTheBercon_02.mov

Max really sucks with effects like these... In order to do this properly I'd probably have to rewrite the spline plugin again...

PsychoSilence
10-19-2008, 12:13 AM
really nice, man! hull shock impacts like in start trek!

wish i could post more stuff :(

OlegB
10-19-2008, 07:12 PM
I'm getting closer to the effect I was after, however I still want more electric-like effect.
Interesting idea - btw, quite doable with off-the-shelf Box#3. I did a quick test - here's animation - ElectricWaves.zip (http://www.orbaz.com/forumData/animations/20081019/ElectricWaves.zip) - and a snapshot of the data flow wiring

http://www.orbaz.com/forumData/images/20081019/ElectricWaves.jpg

Thanks,
Oleg B.

PsychoSilence
10-19-2008, 07:48 PM
crazy flow! gonna have to try this during sim or render time!!! wish i could see more of the first data op :(

OlegB
10-19-2008, 08:56 PM
I'll drop the scene file at Orbaz forum later...

Thanks,
Oleg B.

Glacierise
10-19-2008, 09:29 PM
Wow does that look nice. The scene file will definitely be interesting to check out!

JohnnyRandom
10-20-2008, 03:21 AM
I did a quick test - here's animation...

LOL, Gees Oleg, wish I could come up with a "quick test" like that:D

Wicked
10-20-2008, 09:57 AM
LOL, Gees Oleg, wish I could come up with a "quick test" like that:D
Indeed, same here. I had this kind of effect in my mind for a long time, but never got to make a serious attempt.. :hmm:

Bercon
10-20-2008, 12:41 PM
OlegB, nice!

I don't have Box #3 here as doing particle effects is just a hobby of mine for now, so I'm working with vanilla max. The main problems I had with the effect in question were:

- Turning particles into spline
Made new spline modifier with maxsdk for this. I'm not sure this was actually a good idea. I can't really modify the spline and I don't get any information about the particles transfered besides position. I wonder how fast it would be to have script operator recreate existing spline along the particles?

- Displacing particles off the surface
Made the sliding particles spawn new particles that live only 1 frame, these shortlived particles are displaced along the surface normal and then used to create the spline.

EDIT:
Okey here is a little update on the effect in question. I added some turbulence, increased number of particles per spline to 400 and switched small random noise to Perlin noise gradient vector. Its now pretty close to what I was after. The next thing to do is add some sparks to the collision points.

http://www.niksula.cs.hut.fi/~jylilamm/graphics/FistOfTheBercon_03.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/FistOfTheBercon_03.mov)

EDIT2: Sparks, I don't like them yet.
http://www.niksula.cs.hut.fi/~jylilamm/graphics/FistOfTheBercon_04.mov (http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/FistOfTheBercon_04.mov)

PeteDraper
10-28-2008, 05:14 PM
right... I'm gonna go and pass out now.

Finally feckin' finished the 3rd edition with a shedload of particle flow goodies in it; I finally decided to do the barn tornado as a video tutorial else it would have taken up about 1/3 of the book - it's about 3.5 hours long.

Here's a quick youtube link of the animation samples in the book / online video tutorials. Obviously there's more than that but haven't been included in the video as they're not animated - check out www.xenomorphic.co.uk/gallery_dte.htm for the full skinny

http://uk.youtube.com/watch?v=IgWQEcyQcsc

cheers, and ...uhhh... what the hell did I do before this thing started???!

Pete

Glacierise
10-28-2008, 05:40 PM
Dude that's just awesome! Just waitin' to get my paws on it now!

PsychoSilence
10-28-2008, 06:21 PM
Great News! Canīt wait to get it!!! All looking great!

amckay
10-28-2008, 06:48 PM
pete so you're all done? lets go get beers now, you have no excuse

looks great dude, is this still without plugins? the fire is metaballtastic! It looks great man nice efforts

seriously though.. beer

PeteDraper
10-28-2008, 06:59 PM
yep still without em :)

deffo - though after this weekend I don't think I could touch anothe... oh yeah, mine's a Leffe... :)

TwiiK
10-28-2008, 07:42 PM
That looks phenomenal.

Can't wait to get my dirty paws on it.

JohnnyRandom
10-28-2008, 07:49 PM
Good news:) Can't wait! Let's hope the four month pre-order bumps me up in the queue.

depleteD
10-28-2008, 07:54 PM
Very awesomness

TunnelLight
10-28-2008, 08:05 PM
Nice one Pete, How much do you knock off the price if they're autographed? :p

Aldarion
10-28-2008, 08:20 PM
That looks really tasty ! When does it hit the stores here in UK ?

Steve Green
10-28-2008, 08:24 PM
Can't wait - particularly like the scuba bubbles and the wobbliness of the detergent bubbles.

And the very nice looking shallow DOF dust motes.


Cheers,

Steve

PeteDraper
10-28-2008, 08:30 PM
cheers guys :) much appreciated!

It should hit the UK in about a few weeks - est about a month as it's gotta be dragged across from the states by an asthmatic ant in a canoe or something

hey mark - it's actually 50% off is it's signed as some git will have scrawled over it hahahahaha :)

grury
10-29-2008, 09:33 AM
Can't wait - particularly like the scuba bubbles and the wobbliness of the detergent bubbles.

And the very nice looking shallow DOF dust motes.


Cheers,

Steve

Agree, my faves too. Very keen to see how those flames were made too, awesome.

Congrats Pete.

avolution
10-31-2008, 07:03 PM
What releases would your book be appropriate to?
Some of us "forgot" to upgrade after, err R6!


yep still without em :)

deffo - though after this weekend I don't think I could touch anothe... oh yeah, mine's a Leffe... :)

TwiiK
10-31-2008, 07:55 PM
I think Pete said a while back that the included scene files are all made with 3ds max 2009.

Glacierise
11-07-2008, 09:57 AM
Hey what happened with box#2? Things have been too quiet for too long. Or has AME decided to be cool and include box#2 in the next max :)

JohnnyRandom
11-07-2008, 02:52 PM
^Still in the works. Oleg is building some real sweet stuff, that's all I can say, but be assured he's working hard on it :)

JonathanFreisler
11-07-2008, 03:01 PM
Quick, johnny knows something GET HIM...

PsychoSilence
11-07-2008, 08:22 PM
conspiracy! conspiracy!

should i whisper "shape *******" or "animated ********* objects"? I better don't!!!

i'm just whispering we all might have to re-learn bits and pieces of PFlow...or maybe not?

Glacierise
11-07-2008, 09:05 PM
We guessed about shape collision :D And the relearning - bring it on :D Strange thing Johny said is that Oleg is building - I thought that when the beta started the features were locked down. Whatever, it's going to be cool. Maybe some appetizer demos? :)

PsychoSilence
11-07-2008, 11:23 PM
only form the magician himself :) i dont wanna violate NDAs here :D

JohnnyRandom
11-08-2008, 04:25 PM
whisper.....Building = Some truly unique surprises if all goes well:)

JonathanFreisler
11-08-2008, 04:46 PM
:O sounds exiting.

I like whispers.

Were the box tools supposed to come out in order? If so, what happened?

Steve Green
11-08-2008, 05:00 PM
Originally, yes. But then it was decided that it would be better that Box 3 came out first.

PsychoSilence
11-08-2008, 05:21 PM
the happy pflow family with a single mom (aka processor core) :D

there will be a majour update to box3 along with box2 thats what is no secret anymore (since siggy i guess). maybe oleg can validate any release rumours as i wouldnt say anything.

charleyc
11-08-2008, 06:00 PM
Yeah I am not certain what Oleg wants mentioned at this point either. There were only very minor limitations placed on me for the Siggraph demo, and I can't imagine that he would be worried about keeping what was shown there a secret.

As for the order, Box2 was to be made a lot better by Box3, so Box3 was pushed forward in development ahead of Box2.

Bercon
11-09-2008, 10:37 AM
I compiled max 9, 2008 and 2009 versions of the modified spawn and collision spawn operators.

You can download them from: http://www.niksula.cs.hut.fi/~jylilamm/tools.shtml

Wicked
11-09-2008, 12:46 PM
You have some very usefull stuff on your site! Thanks for sharing.

PsychoSilence
11-13-2008, 12:03 AM
sidenote to box3 owners:
atm the quickest but blessed with the most less features collision test is a siple data test as shown in the attachment. I make massive use of it when i literally ONLY need to detect if a particle penetrated a surface becasue it is sooo fast compared to the other collisions. thats all i need for 75% of my purposes :bounce::buttrock:
If u need multiple bounces or features such as "Will Collide" ur better off using what ships with PFlow.

my 5 minutes of thinking about collisions until the new boxes drop from heaven.


kind regards,
Anselm

JohnnyRandom
11-13-2008, 02:20 AM
Thanks Bercon:)

Nice one PsychDude:thumbsup:

PsychoSilence
11-13-2008, 03:32 AM
thanks john :)

just realized how bad my typo in the posting really is :D anyways :beer:

oh and berconīs stuff totally rocks! just thought i mention some box-erism along to it :)

itīs about time we get the thread going again :) we should publish some stuff for the peeps. free stuff always draw atention :D

JohnnyRandom
11-13-2008, 05:21 AM
Hehe, "mmmnnn Beer..yes.. oh yes...woohoo!" Famous quote by none other than......

PsychoSilence
11-13-2008, 04:47 PM
Allan McKay?

JohnnyRandom
11-13-2008, 05:02 PM
LOL, no, I guess it could have been any of us :P... Homer Simpson :D

Glacierise
11-13-2008, 05:30 PM
Nice stuff guys! Btw, I notice Anselm has 'input physx', hmmm.

It would be really great if Oleg makes a free version of box#3 that enables users to use data operators, but not to edit/create new ones. It will make it much more famous, and people will publish all kinds of cool operators, I think.

PsychoSilence
11-13-2008, 09:20 PM
LOL, no, I guess it could have been any of us :P... Homer Simpson :D

being german i guess i claim to drink too much of this good stuff :D

@ christo: i should have set the depot to hidden :D u mean like the sim only license that comes with fume? may only in connection with a box3 purchase. but even then if u install box3 and dont license it it is actually a read only/render license.

Glacierise
11-13-2008, 09:52 PM
Wow so Oleg did make it this way! Very cool, you guys that use box#3 heavily should publish some stuff out.

PsychoSilence
11-14-2008, 12:49 AM
Wow so Oleg did make it this way! Very cool, you guys that use box#3 heavily should publish some stuff out.

basicly yes, if u own a full license of box3 u can install it infinite without register to function as render only :) zombie license if u will. follows without own will......

JonathanFreisler
11-14-2008, 02:45 AM
@ Glacierise, heres my stab at doing the running cracks like you did. Done wiht 2 pflow systems.

The cracks are custom and i hand modeled (sigh) for a job that needed to follow the specific cracks in a specific texture.

tunneling running cracks (http://au.youtube.com/watch?v=LuFqWPNSwOI&fmt=6)

@ PsychoSilence, the first thing i noticed in tha screen shot was the 'input physx' lol, before anyhting else.

JohnnyRandom
11-14-2008, 02:53 AM
Nice Jono ;)

Anse ... http://au.youtube.com/watch?v=3d4gTAj8FSE :D

PsychoSilence
11-14-2008, 04:31 AM
hah! thats what i should have right now! was first at work, will be the last to leave :(

Glacierise
11-14-2008, 10:30 AM
Hey nice Johny :) You should have 2-3 levels though, looks cooler. I did mine with just 2 levels, it will be better with 3. I'll redo it these days, I mean - one level is the central deep crack, the next one is the more shallow, farter away, and the third - very shallow, farthest from the center.