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

PsychoSilence
05-28-2009, 05:41 PM
I'd IDKFA and blow your immortal ass to smithereens! Cheers guys, I know what will kick the wasps round here - box#2 release! Cmon Oleg! :D

i actually have the IDKFA shirt too :D

http://www.retrogt.com/retro-t-shirts/

Steve Green
05-28-2009, 05:43 PM
Hi,

it's easy enough to spawn particles based on the luminance of an image, but I wondered if it's possible to have an image attract particles, like the find target op, but based on image rather than an object?

Thanks,

Steve

PsychoSilence
05-28-2009, 05:56 PM
u can place them on the object by grayscale and store their location, then send em out and use a find target by script vector :)

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

Steve Green
05-28-2009, 06:05 PM
Cool, thanks - unfortunately Allan's download limit is exceeded, so I can't see the vid.

My idea was to have it working from an animated map, so it wouldn't be stored once it would constantly be hunting for the image sequence.

I Thought it might look quite interesting to have particles speeded towards an animated image, but I'm not sure if your suggestion would work with an animated sequence, or if there's another solution...

Cheers,

Steve

DustinBrown
05-28-2009, 06:17 PM
Hey guys. I need to fairly quickly generate wispy cigarette smoke for a piece I'm working on. I'm mostly a sculptor/modeler, not a particles/fx guy, so while I don't mind making it myself, I'm not particularly interested in learning all the ins and outs of PFlow...I just need the smoke to use as a design element.

Something like this:
http://forums.cgsociety.org/attachment.php?attachmentid=142776&stc=1

I tried to download the cigarette smoke video tutorial on Allen McKay's website, but apparently he's exceeded his download limit, so I can't get the video file from there. Also, there was quite a bit of discussion about creating cigarette smoke in the first pages of this thread, but those are from 2004, so all of the linked files (video files, max files, etc) are all broken/dead links. That's really unfortunate, I wish people would attach files here instead of externally.

I really just want a max file I can keep in my asset library, which has nothing but cigarette smoke in it, ready to render. Any time I need that sort of smoke for a cigar or cigarette or incense, I can just jump into that file, tweak some wind properties or something, and bring the smoke into my project.

Or if anyone has the Allen McKay video available, PM me and I can give you my email.

Thanks

Bercon
05-28-2009, 06:37 PM
Thanks guys.

Mental ray version might be a bit tricky. I'll have to look into it some time. Its on low priority however as I have BerconMaps update on the way and it'll probably eat my allocated coding time.

I think Brazil version might be easier to do, when I get the Brazil SDK.

The milk coral scene was in the same folder as the plugin files: http://www.ylilammi.com/BerconMetaball/MilkCoral.7z

NOTE! It contains some custom spawn operator, just load the scene ignore the warning and replace the missing spawn operator with default spawn operator and set the spawn by distance option in that operator, then it should work just fine.

SuperRune
05-28-2009, 07:33 PM
Bercon, that is beautiful!! I'm also crossing fingers for a Brazil version :)

DeKo-LT
05-28-2009, 10:26 PM
@Bercon
good job, thanks for sharing ;)

@Steve Green
By image you can select some faces with Vol. Select, then use delete modifier and you have pure animated geometry by image luminance.

@DustinBrown
Everyone dreaming about MakeArt button :)
Here is my very old project, maybe this helps:
arbat_FX_smoke06.zip (http://deko.lt/tmp/Forums/arbat_FX_smoke06.zip)

DustinBrown
05-28-2009, 11:30 PM
DeKo-LT, thanks man. I'm not looking for a Make Art button. PFlow is a bit esoteric IMO and I was just looking for a tutorial or something to point me in the right direction, like Allan's video. Unless you know what you're doing, you could easily spend an entire day playing Slider Jockey just to get the right effect, and that's what I'm looking to avoid. I'd rather be drawing or sculpting.

floopyb
05-30-2009, 05:35 AM
Awesome stuff Bercon, did a quick test with 500,000 particles.
Would be great to be able to get the particle scale/size from the pflow :)

PsychoSilence
05-30-2009, 07:09 PM
slight OT: i added some more pics to my gallery on my website

3DMadness
05-31-2009, 02:13 PM
Awesome stuff Bercon, did a quick test with 500,000 particles.
Would be great to be able to get the particle scale/size from the pflow :)
Hey, nice test! I agree with you, would be nice to controle the size from pflow to add some random to the size and change it based on the flow.

Bercon
05-31-2009, 06:37 PM
Perhaps indeed.
http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/Metaballs_26.jpg

JohnnyRandom
06-01-2009, 12:32 AM
A box#2 update, it is now in RC stage so it shouldn't be too much longer :)

How about some basic Newton physics:

Newton's Cradle QT 8mb (http://www.4rand.com/TEST/ParticleFlow/b2b/PhysXGlue/Pendulum/NewtonCradle.mov)

http://www.4rand.com/TEST/ParticleFlow/b2b/PhysXGlue/Pendulum/NewtonCradle.png


A knock off of Ronald Fedkiw's Two-way coupling of rigid and deformable solids (with Tamar Shinar and Craig Schroeder). I like the faking concave shapes ability ;)

Donuts QT 18mb (http://www.4rand.com/TEST/ParticleFlow/b2b/PSkinner/Donuts/donuts.mov)

http://www.4rand.com/TEST/ParticleFlow/b2b/PSkinner/Donuts/donuts.png


Bercon, fine work them metaballs are a hoot!

HeadSmell
06-01-2009, 05:21 AM
Dam cool stuff John...Cant wait for Box #2

this and zbrush 4...gona be a good year :)

JohnnyRandom
06-01-2009, 05:36 AM
Thanks man :)

Just finished a quick render.

Another Box#2 and Bercon Metaball - 1000 frames on 2 cores in under 2.5 hours, that's pretty quick and none of the other usual metaball overhead.

PhysX Force - Spherical Gravity QT 45mb (http://www.4rand.com/TEST/ParticleFlow/b2b/PhysxForces/Newton/nwtncbll_brcnmtbll.mov) (sorry for the large file its a 1000 frames)

http://www.4rand.com/TEST/ParticleFlow/b2b/PhysxForces/Newton/nwtncbll_brcnmtbll.png

3DMadness
06-01-2009, 02:33 PM
Perhaps indeed.
Great! I'm looking forward version 1.01. :D

I was going to say that it is the perfect match to box #2 but Johnny had already showed it. ;)

Nickolay411
06-02-2009, 07:22 PM
Bercon Thank You. This Is Really Impressive And Like Someone Else Said Very Generous!

The Max Community Is The Best. :)

galagast
06-03-2009, 03:21 AM
bercon! a milliom thanks goes to you! :buttrock:awesome stuff man.

loran
06-03-2009, 08:44 AM
That rocks Bercon! can't wait for a Mental Ray one

Nickolay411
06-03-2009, 03:27 PM
BOX2 is out.

http://www.orbaz.com/forum/viewtopic.php?t=2227

:)

loran
06-03-2009, 03:40 PM
the final animation
>video (http://www.youtube.com/watch?v=54OxE2OS_OE)

watch my blog for more
http://loran-cg.blogspot.com/

http://laurent.renaud.free.fr/divers/blog/demo-chateau2.jpg

http://loran-cg.blogspot.com/

JohnnyRandom
06-03-2009, 03:56 PM
Nice work Loran :thumbsup: that came out great ;)

PsychoSilence
06-03-2009, 04:08 PM
awesome job, loran! vanilla pflow still rocks butt :)

noouch
06-03-2009, 04:18 PM
Looks great, only thing that bothers me is the chunks pop a bit when they're crumbling.

holycause
06-03-2009, 05:22 PM
Nice Work ;)

JonathanFreisler
06-03-2009, 11:05 PM
:drool: box 2! *insert buttrock*

irwit
06-04-2009, 09:44 AM
WARNING

There must be a glitch in teh matrix or something because it looks like box 2 has been released ...
:bounce: :beer: :bounce: :beer: :bounce: :beer: :bounce: :beer:

loran
06-04-2009, 10:00 AM
pf box2 = 595$ !!
unfortunatly, 3ds is the most expensive 3d package over the galaxy and you need to spend thousands and thousands $ more after buying it!
every pf boxes and fumefx and afterburn and glu3d should be part of 3dsmax...

Do Autodesk guys knows Blender will integrate volumetric system comparable to fumefx soon and ever have a robust fluid simulation system... (and I don't speek about the amazing rig, skinning, UVW unwrap, nodale material and compositing, sculpting solutions).
unfortunatly the particle sytem is not comparable to PF... for now

TwiiK
06-04-2009, 10:11 AM
And yet people continue to buy 3ds max and its plugins...

I've tried Blender quite a bit and it's awesome, but still far from a package like 3ds max. Just like The Gimp is awesome yet far from a replacement for Photoshop.

loran
06-04-2009, 10:51 AM
And yet people continue to buy 3ds max and its plugins...

And yet people continue to crack 3ds max. Still the most cracked 3d package...

I've tried Blender quite a bit and it's awesome, but still far from a package like 3ds max. Just like The Gimp is awesome yet far from a replacement for Photoshop.

Not sure this is the right place to discuss this, but Blender is far foreward MAx in many points. Rig and bones are simply very very much more powerfull and easy to use. Automatic Skinning is awsome, there is no equivalent solution in max.Automatic Unwrap UVW in Blender is awsome too, the max solution is ridiculous if you compare (http://www.raylightgames.com/xrayunwrap/xrayunwrap.htm is a plugin for max based on the Blender system).There is a good sculpt solution in Blender (the next version will release micropolygones sculpt), for max you have to use another expensive software. There is a very good painting and projection painting in Blender, poor solution in max. Modeling tools are very close to the maya ones... So what is bad in Blender? Basic Material editor and Poor Rendering if you compare to MentalRay and no very good particle system. I am sure you know Blender is FREE and is 10Mo to download (Max2010 =7 Go).Blender works on PC (32&64bits) OSX and Linux too.

TwiiK
06-04-2009, 11:13 AM
This is not the right place to discuss this and you know it, I don't even know what made you bring it up in the first place. The price of box #2 is perfectly fair and whining about it in the particle flow section of a 3ds max forum will get you nowhere.

Think about how much it would cost you in terms of work hours to replicate the features of box #2. Also, if there were low cost or free alternatives that offered the same or better features of course they would be forced to lower their price, but there isn't. Thinking particles cost like $1700 and people buy that.

I can understand the price is high for the low budget hobbyist, and for those there are the cheaper or free alternatives, but for the serious hobbyist or commercial artist I think the price is very justified.

I know many people who spend $2000-$3000 a year on their hobbies. I probably spend that amount myself and my hobby is computer graphics.

Glacierise
06-04-2009, 11:13 AM
Make a new max-bash thread please, it seems that there isn't a fresh one to vent in right now ;)

loran
06-04-2009, 12:42 PM
Ok I appologiz but you don't cry so loud when others post their party pictures... ;]

What I said about plugins prices is that autodesk should integrate all of this in 3ds max. Other package ever have this function for lower price. What I said is 3dsmax will meet strong competion more and more. Of course plugins author have to be paid, so why autodesk doesn't buy their works.
ok give it up...

irwit
06-04-2009, 01:20 PM
Ok I appologiz but you don't cry so loud when others post their party pictures... ;]

What I said about plugins prices is that autodesk should integrate all of this in 3ds max. Other package ever have this function for lower price. What I said is 3dsmax will meet strong competion more and more. Of course plugins author have to be paid, so why autodesk doesn't buy their works.
ok give it up...

They have included polyboost, CAT, PF box 1 not least to mention mental ray :D and probably numerous others. There's are just too many to choose from with max ( which isn't a bad thing :) )

3DMadness
06-04-2009, 01:40 PM
Ok I appologiz but you don't cry so loud when others post their party pictures... ;]

What I said about plugins prices is that autodesk should integrate all of this in 3ds max. Other package ever have this function for lower price. What I said is 3dsmax will meet strong competion more and more. Of course plugins author have to be paid, so why autodesk doesn't buy their works.
ok give it up...
I didn't want to keep it off-topic, but wich app have this kind of things integrated, besides blender? I was thinking about maya unlimited, but for the price I can get max + box2 + glu3d + fume... oh, and box1 comes with max 2010. ;)

Nickolay411
06-04-2009, 02:28 PM
burnt sienna oils 200ml 15 dollars
36X48 Fredix Canvas 48 dollars
kolinsky 3 brush set 111 dollars
oak lyre easel 283 dollars

ORbaz technologies box#2 priceless.


All art is expensive. :)
use your mastercard.

noouch
06-04-2009, 04:31 PM
I didn't want to keep it off-topic, but wich app have this kind of things integrated, besides blender? I was thinking about maya unlimited, but for the price I can get max + box2 + glu3d + fume... oh, and box1 comes with max 2010. ;)
Houdini has that functionality and lots more, but at $8,000 for Houdini Master, it's quite expensive.

JonathanFreisler
06-04-2009, 11:04 PM
Wait, and autodesk "intergrading" plugins isnt up to autodesk! they need to buy out that product. Its probably in sitni sati and olegs best interist to maintain there own products.

Its a bit of a nieve attitude to say "autodesk shoudl just have allt eh stuff i want in it".

CapitanRed
06-04-2009, 11:14 PM
they coul'd try to keep it backwards compatible, so plugins have not to be rewritten for each version :)

plejboy
06-05-2009, 08:56 AM
Cant believe the whining...
THIS IS AWESOME NEWS!
and about the price! wtf? It's SO worth the money.

Great work Oleg and everyone that betatested.
This will boost our workflow like crazy here at mainframe.

Good times ahead!

evanschaible
06-05-2009, 09:46 AM
Oh dear...

"Autodesk should just integrate all those plug ins inside Max"

Well... Autodesk cant do that, becuase they dont own the plugins. They actually do what you are complaining about. I mean, if you have a subscription plan you can get CAT for free because Autodesk bought out SOFTimage. But I like the fact that they remain plug ins honeslty, that keep the price of Max reasonable and allows you to cutomize for exactly what you need. With Houdini some peope pay 8 grand for an app, and probably about 500 or 600 dollars of that is to pay for developement of PyroFX which they will never use. I am glad that isnt the case with Max.

"Oleg shouldnt charge so much for the Boxes"

Okay, you spend the thousand of hours that is required for the devlopment of these tools, and lets see what you charge. Where do you think Oleg gets these great plugins? He isnt just using someone elses stuff, as far as I know he has no development team that codes for him, HE MAKES IT HIMSELF! That takes time (countless hours), resources, money and even alot of the times health and sleep on his part. I think his prices are perfectly fair, and with his prices, I think in the long run he loses out because he could charge alot more.

"Max is too expenzive"

Okay then, buy out Autodesk and lower the price. We will all be grateful to you when you do.


Sorry for the sarcasm, but I couldnt help it.

evanschaible
06-05-2009, 09:55 AM
And yet people continue to crack 3ds max. Still the most cracked 3d package...



Not sure this is the right place to discuss this, but Blender is far foreward MAx in many points. Rig and bones are simply very very much more powerfull and easy to use. Automatic Skinning is awsome, there is no equivalent solution in max.Automatic Unwrap UVW in Blender is awsome too, the max solution is ridiculous if you compare (http://www.raylightgames.com/xrayunwrap/xrayunwrap.htm is a plugin for max based on the Blender system).There is a good sculpt solution in Blender (the next version will release micropolygones sculpt), for max you have to use another expensive software. There is a very good painting and projection painting in Blender, poor solution in max. Modeling tools are very close to the maya ones... So what is bad in Blender? Basic Material editor and Poor Rendering if you compare to MentalRay and no very good particle system. I am sure you know Blender is FREE and is 10Mo to download (Max2010 =7 Go).Blender works on PC (32&64bits) OSX and Linux too.

Yeah, and I have downloaded the last 3 or 4 releases of Blender and still cant get past the UI. It took my no time to learn Max, but after I think two years of trying to learn Blender so I could play with its fluids, I threw it away because the UI is rediculous.

I would pay the money for Max just because I can use the perspective view and I dont have to spend 5 years learning a GUI.


Anyhow... Box 2 is amazing.

HeadSmell
06-05-2009, 10:21 AM
dam...gona be 600 broke...there goes rent

noouch
06-05-2009, 12:57 PM
dam...gona be 600 broke...there goes rent
"Screw a roof over my head, I've got PhysX Particles bitches!"

jigu
06-05-2009, 02:13 PM
As box #2 uses physx technology, Is it going to have physx fluid simulation too? or soon?

SoLiTuDe
06-05-2009, 04:50 PM
Yeah, I was gonna go off about how the pricing is fine, and that 8 grand for Houdini is about what you'll end up spending on max + all the fx plugins (some of which don't work with each other, or even the built in max plugins very well), and that no matter what, you will pay for development of stuff you might not use (i'm not a modeler, but I'm sure going to be paying for modeling tools development...)... so instead of going on and on about that stuff which is starting to get way OT, I just want to congratulate Oleg for all that he's accomplished... Box2 is a great addition to max / pflow... and will definetly be coming in handy in the future! :D:D

--Ian

^^I don't see why the physx fluids wouldn't be on "the list" but i'm sure it would've been too much to get done for this release of box2. I don't think it'd be any sort of replacement for something like glu (or realflow, or houdini fluids) but it'd probably be pretty cool for simple stuff... though this example (only good one i've found) looks really promising: http://www.youtube.com/watch?v=XmqyyeNXA6A&feature=related

3DMadness
06-05-2009, 05:53 PM
Maybe someone can write a fluids solver with box#2 + box#3 like the one thiago costa made for XSI using ICE: http://www.vimeo.com/thiagocosta/videos

It's based on the same SPH paper, I think it will come to PF someday. :)

SoLiTuDe
06-05-2009, 06:17 PM
^^I'm pretty sure it's been possible to write your own fluids solver since day 1 with the sdk. There's been one or two people that have done stuff like this already... just not the extent of the xsi community -- from what i hear ICE is very open / and is probably a bit easier to develop for.

http://www.zhangy.com/main/index.php?module=documents&JAS_DocumentManager_op=viewDocument&JAS_Document_id=24

http://www.zhangy.com/main/index.php?module=documents&JAS_DocumentManager_op=viewDocument&JAS_Document_id=26

http://dimo3d.blogspot.com/2007/10/fun-with-fluids.html

JohnnyRandom
06-05-2009, 06:32 PM
That is bad ass thanks for posting that link, I certainly have some envy for Ice and Houdini's architecture...some amazing stuff!

3DMadness
06-05-2009, 08:30 PM
Hey Ian, I forgot about this one dimo showed... too bad he never released it.
Also the one from Zhang looks promising, I've tried the plugin and works fine but only for the demo scene. I guess now that the physx comes with box#2 it could be easier to get fluids now. Should we wait for box#4? :D

SoLiTuDe
06-05-2009, 08:47 PM
I guess now that the physx comes with box#2 it could be easier to get fluids now. Should we wait for box#4? :D

Yeah, the physx fluids stuff should be "easier" now ...but that's assuming an sdk is availalbe for box2. I'm sure Oleg has it on his list of things to do. You guys can download the physx plugin from nvidia: http://sourceforge.net/projects/physxplugin/ which includes particle fluids... my first impression is that it's not really all that great... maybe good enough for some basic splashes or something...? I haven't played with it much. I could be totally wrong, maybe it's bad ass... just all the stuff i've seen (minus that one examples) doesn't look that production ready/useful. :D

Nickolay411
06-10-2009, 07:40 PM
BERCON, will you compile a version where particle scale affects the radius of the metaball?

thanks,
nick

reForm
06-14-2009, 03:11 PM
BERCON, will you compile a version where particle scale affects the radius of the metaball?

thanks,
nick

I hear the next version, due soon, will have this functionality :)

Bercon
06-15-2009, 12:20 PM
Here is the new 1.10 version. Again works only with VRay 1.5 SP3a.

Download: http://www.ylilammi.com/BerconMetaball/BerconMetaball_2009_06_14.7z

Instal, help, what the? Read the readme.txt that comes with the archive.

Supports max 9, 2008, 2009, 2010, 32/64bit. Again tested only briefly on each system and no large scale testing done so its possible that there are still some bugs lurking around.

Small warning, the field functions were changed so old scenes don't look exactly the same when rendered with this new version. You have to change the Distance Function dropdown to correct. Squared = Power of 2, Cubic = Renderman = Power of 3, Wyvill = Wyvill.

It now has support for variable particle size (pflow), N number of motion blur samples and texture (fixed in world coordinates, doesn't move with the blobs).

http://www.niksula.cs.hut.fi/%7Ejylilamm/graphics/Metaballs_29.jpg

loran
06-15-2009, 02:28 PM
Bercon this is really great. but i am a Mental Ray user! bouuuuuhouhouuuu

3DMadness
06-15-2009, 03:57 PM
Great job Bercon, thanks again for sharing and for this great update! :)

Nickolay411
06-15-2009, 04:02 PM
Bercon,

SWEEEEEEEEEEEEEEEEEEEEEET! :)

reForm
06-15-2009, 04:13 PM
Hi Bercon, great update on the metaballs. One problem I have encountered, using the new variable blob size option seems to break DR rendering in vray. Switching to specified blob size allows the metaballs to render correctly using DR.

Bercon
06-15-2009, 04:49 PM
Breaks it how? Did you also update the newest version to DR nodes?

reForm
06-15-2009, 04:53 PM
The nodes were returning black buckets.

Curiously, the problem appears to have corrected itself. Maybe it was just a vray problem. I had updated all the nodes ofcourse! ;)

JohnnyRandom
06-16-2009, 08:43 PM
Nice Bercon, thanks, going to have another go at it, it just keeps getting better :)

jigu
06-19-2009, 05:17 AM
2012 new trailer up! http://www.traileraddict.com/trailer/2012/trailer

Congrates to all of you max guys whoever worked on this film! Best VFX shots I have ever seen so far.. I see Best VFX oscar for this one.

Congrates to all you CGTalk VFX max fellows whoever worked on it......:beer:

HeadSmell
06-19-2009, 07:01 AM
great googly moogly.....

TwiiK
06-19-2009, 08:11 AM
Hah, looks retarded, just like Roland's other films. :)

The VFX did indeed look very nice.

jussing
06-19-2009, 08:41 AM
I looks just plain awesome! :drool:

Yea, the story will be yadder yadder, like Day After Tomorrow... it's not like ONE PERSON holds the key to wave off the disaster (apparently), so it'll propably be father & family with their problems, dodging waves, meteors and tumbling aircraft carriers while re-discovering their true love for each other.

kogden
06-19-2009, 10:17 AM
Fantastic job guys! cant wait too see it :D.:bounce:

Aldarion
06-19-2009, 01:28 PM
Hristo has a shot in there, I presume Allan might have some too. Can you guys share any info about them ? Or the files :P

Glacierise
06-19-2009, 02:05 PM
I have a shot in the trailer, in the movie I have ten! And 3-4 I contributed elements too :P As soon as it's out guys, we'll tell as much as possible about it :)

JohnnyRandom
06-19-2009, 03:08 PM
F***n VFX extravaganza! Sweet, who needs story when your in a motor home getting bombed by meteoric debris :D

3DMadness
06-19-2009, 04:06 PM
All the story I'm looking forward is the one from the guys that have worked on it. :D

Congratulations for all the ones that worked!

PexElroy
06-21-2009, 05:25 AM
Great work on the VFX on this beast. :thumbsup:

galagast
06-22-2009, 04:30 AM
Awesome!
It's been a loooong time since I last felt goosebumps after watching a trailer! :D

TwiiK
06-22-2009, 08:49 AM
I have a shot in the trailer, in the movie I have ten! And 3-4 I contributed elements too :P As soon as it's out guys, we'll tell as much as possible about it :)

I read somewhere, your blog I think, that the car jumping out of the plane was the shot you did. What does that involve actually? :)

Did you do the whole thing? Plane, car, effects, post processing? Also, seeing as you specified "from the side", does that mean that someone else did the shot of the front of the car and the front and back of the plane?

Just very curious how these things work and how much work you need to have done to call a shot yours. If you can't discuss this until after the release of the film then I'll patiently wait. :)

evanschaible
06-22-2009, 09:48 AM
Oh my.

I hadnt been following that, it has Max written all over it. My mouth was open the whole time. I would have a field day with a project like that... and... well you know... a budget like that too. Wonderful work... next one... well... I need a job. :wavey:

- Evan

PsychoSilence
06-22-2009, 06:04 PM
Just very curious how these things work and how much work you need to have done to call a shot yours. If you can't discuss this until after the release of the film then I'll patiently wait. :)

Do it once, then re-do it, then re-do it, then re-do it, then re-do it, then re-do it, then re-do it, then re-do it, then re-do it :D

SoLiTuDe
06-22-2009, 06:15 PM
I read somewhere, your blog I think, that the car jumping out of the plane was the shot you did. What does that involve actually? :)

Did you do the whole thing? Plane, car, effects, post processing? Also, seeing as you specified "from the side", does that mean that someone else did the shot of the front of the car and the front and back of the plane?

Just very curious how these things work and how much work you need to have done to call a shot yours. If you can't discuss this until after the release of the film then I'll patiently wait. :)

When you say "that's one of my shots" it usually just means you have something in it... ie: if you're the fx guy, then you did fx for it, then you can say it was one of your shots, if you're an animator, then you did animation on the shot (maybe the anotov and the cars), you can say it was one of your shots, etc... Technically I have a shot in the trailer at 0:59 ...but all I did was the ground breaking... Still one of my shots though! :thumbsup: There are always a lot of people involved, and calling it your shot usually just means you worked on a part of it.

TwiiK
06-22-2009, 07:08 PM
I see, interesting. :)

Is this normal or is the cg in this film sort of a collaborative project between a ton of different VFX artists seeing as so many of you have worked on it?

Industrial Light & Magic are the ones doing the cg for transformers, right? Do they also outsource certain parts of the VFX in Transformers to other artists or studios or do they do everything themselves?

PsychoSilence
06-22-2009, 09:01 PM
I see, interesting. :)

Is this normal or is the cg in this film sort of a collaborative project between a ton of different VFX artists seeing as so many of you have worked on it?

Industrial Light & Magic are the ones doing the cg for transformers, right? Do they also outsource certain parts of the VFX in Transformers to other artists or studios or do they do everything themselves?

this is going to the edge of off topic but yeah, this is normal. the demands versus the production time frame of a shot/project dont allow much for lone rangers. IMO being a team player is as important as the actual skill set itself if you wanna "survive" at a company...

regarding the outsourcing:
again demands versus the production time frame only allow a few companies of a decent size to tackle a show alone. Usually you see multiple vfx houses rolling in the credits. Sometimes companies jump in last minute to help out. Sometimes a show is broken into logical chunks that are given to multiple companies intentionally. 2012 is/was handled by at least 5 companies e.g.

joconnell
06-22-2009, 10:40 PM
Folks does anyone know any meta solution that respects uv mapping coordinates? I want to do an effect of a painting starting to melt and run so I was doing to map the image to a surface and use pflow to gradually start the particles to slide downwards and "drag" the image with it - the built in max blobmesh doesn't pick up coordinates and since it's for a freebie favour on a short film there's no budget for something like the mesh skinning in box# 2. I thought that a meta solution might give me some nicer folds than trying to animation geometry - Since it's going to be camera mapped essentially I'd like to get a nice smooth leading edge to get some interesting highlights on it. This ring any bells with anyone?

And yeah - the 2012 stuff looks ****ing epic - more consistent than the day after tomorrow bits, great job all around - you're all making me jealous :D

PsychoSilence
06-23-2009, 12:39 AM
Folks does anyone know any meta solution that respects uv mapping coordinates? I want to do an effect of a painting starting to melt and run so I was doing to map the image to a surface and use pflow to gradually start the particles to slide downwards and "drag" the image with it - the built in max blobmesh doesn't pick up coordinates and since it's for a freebie favour on a short film there's no budget for something like the mesh skinning in box# 2. I thought that a meta solution might give me some nicer folds than trying to animation geometry - Since it's going to be camera mapped essentially I'd like to get a nice smooth leading edge to get some interesting highlights on it. This ring any bells with anyone?

And yeah - the 2012 stuff looks ****ing epic - more consistent than the day after tomorrow bits, great job all around - you're all making me jealous :D

Galagast once wrote a scripted solution to glue vertecies to particles:

http://www.orbaz.com/forum/viewtopic.php?t=342

although i would try doing that kinda task in post. render a blobmesh grayscale render and then use that as displacement in comp.

PsychoSilence
06-23-2009, 07:01 AM
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford.avi

yet another take on stanford...

Glacierise
06-23-2009, 07:08 AM
While I can't, of course, give you the internal workflow, I can give you some common sense :) Being an FX artist, I would hardly be able to do film quality modeling, rigging, animation, shading. Even if I would, it's not the point :) In production what you do is iterate, A LOT. Read what Anselm said, and then redo it :D So everybody that has elements in a shot should say that it's his shot - the modeler, rigger, shading artist, animator, fx artist, compositor etc. I feel that shot more 'mine', because we weren't ILM with hundreds of people workin on it you know :) That much about that subject. And, the movie's gonna be cool :D

Glacierise
06-23-2009, 08:01 AM
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford.avi

yet another take on stanford...


Argh pour gas into my veins :) I will buy it these days and then see what happens :buttrock:

OlegB
06-23-2009, 10:10 AM
Anselm,

http://www.3delicious.de/spielwiese...on-stanford.avi (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford.avi)

yet another take on stanford...
Is it with Box#2 or something else?

Thanks,
Oleg B.

joconnell
06-23-2009, 11:44 AM
Galagast once wrote a scripted solution to glue vertecies to particles:

http://www.orbaz.com/forum/viewtopic.php?t=342

although i would try doing that kinda task in post. render a blobmesh grayscale render and then use that as displacement in comp.

Sure thing - I thought about doing an all 2d solution but I'm not going to get the same quality of light / bump i'd get from nice uv bits - that script is great though, it offers some very interesting possibilities to get quick results and different looks!

Cheers Anselm!

JohnnyRandom
06-23-2009, 03:51 PM
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford.avi (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford.avi)

yet another take on stanford...

Glad you didn't squish the bunny :D

PsychoSilence
06-23-2009, 04:30 PM
Glad you didn't squish the bunny :D

i thought of doing so but the only OBJ i could find on my machine was the dragon and they dont want you to cause harm to it...so i took a good ol' teapot. poor bastard...

JohnnyRandom
06-23-2009, 05:32 PM
AFAIK the bunny is fair game :D

Choose another model for these sorts of experiments. (You can do anything you want to the Stanford bunny or the armadillo.)

Was that teapot a PSkinner? Your deformations held up really well ;)

PsychoSilence
06-23-2009, 05:44 PM
there we go, downloaded the budda as it is and created a proxy for the sim and a skinned one which is as super poly as it comes from stanford :)

www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_budda.avi

Glacierise
06-23-2009, 05:51 PM
So it does work well with highpoly objects?! Way awesome. Also, I started a new religion this morning, so BEHOLD THE ALMIGHTY TEAPOT FROM THE SKY! Images depicting the holy teapot are offensive and if you do that one more time, I'm gonna come drink all your beer!

JohnnyRandom
06-23-2009, 05:56 PM
LOL, DUDE you can't squish the Buddha! ONLY the bunny or the Armadillo!

Hurry quick, replace it with the Bunny :)

PsychoSilence
06-23-2009, 06:02 PM
LOL, DUDE you can't squish the Buddha! ONLY the bunny or the Armadillo!

Hurry quick, replace it with the Bunny :)

http://physbam.stanford.edu/~fedkiw/animations/buddha_gear.avi

fedkiw did ti himself with the budda :(

JohnnyRandom
06-23-2009, 07:19 PM
Ok now I'm jealous :P

Hasty setup, birth grid w/ apprx 1100 control particles, Gorilla head 75000 faces, 7 min render...

GorillaHead QT 4.5mb (http://www.4rand.com/TEST/ParticleFlow/b2b/PSkinner/GorillaHead/gorillahead.mov)

No gorilla's were harmed in the getting of the skull...I hope

PsychoSilence
06-23-2009, 08:11 PM
hah! i love it! little squishy, rubbery but proofs the concept. u can skin almost everything to particles now :D looove it!

grury
06-23-2009, 08:34 PM
U box#2'ers are havin too much fun...I wanna play too :(

jussing
06-24-2009, 08:21 AM
@PsychoSilence - I feel like I'm the odd one out here - I have no clue how you did that squeeze thing.... are you skinning a mesh with particles, or what? Can you please give me a quick run-down? :)

Thanks,
- Jonas

jigu
06-24-2009, 08:44 AM
he is using box #2 physx skinning I guess?

JohnnyRandom
06-24-2009, 02:38 PM
@PsychoSilence - I feel like I'm the odd one out here - I have no clue how you did that squeeze thing.... are you skinning a mesh with particles, or what? Can you please give me a quick run-down? :)

Thanks,
- Jonas

It is Box#2 chicanery ;) its the opposite, skinning control particles with a mesh.

It basically works likes this: (there are other ways to setup your control particles but this is the simplest most straight forward)

1. You have two copies of a mesh, mesh01 will be your particle birth and culling object, mesh02 is your Particle Skinner object.
2. You birth your particles in a birth grad around mesh01 and then cull (or delete) any particles outside of the mesh using the culling options in the birth grid.
3. apply your pflow PhysX operators to get gravity, collisions, binding, and any other external forces you want to apply.
4. mesh02 particle skinner is linked to mesh01's particle flow system, the pflow system is basically now a set of bones for mesh02
5. Squish, squash, smash, tear, shoot, explode, whatever you heart desires.



-John

3DMadness
06-24-2009, 02:42 PM
U box#2'ers are havin too much fun...I wanna play too :(
I couldn't agree more...

box#2 :bowdown:

jussing
06-24-2009, 03:10 PM
Thanks, Johnny, that sounds awesome!

- jonas

PsychoSilence
06-24-2009, 06:10 PM
and another one from last night...

original from stanford:
http://physbam.stanford.edu/~fedkiw/animations/filling_bag.avi (http://physbam.stanford.edu/%7Efedkiw/animations/filling_bag.avi)

box2 approach:
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_fillingbag.avi (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_fillingbag.avi)

i should start the water flow later so the bag is already fully rested but never the less the bag acts a sa softbody geometry getting filled up wobbling from the water pouring in...

proof of concept: can be done :) YES WE CAN

check the attachment for hints. the birht stream is caching glu3d guide caches. then i pointcached the bag so i can entirely turn off the pflow. as last step i emitted glu3d by particle cache and meshed it. more pflow birth stream particles more detail...but my computer is not the fastest under the sun...

PsychoSilence
06-24-2009, 11:17 PM
original from stanford:
http://physbam.stanford.edu/~fedkiw/animations/armadillo_tear.mov

box2 approach:
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_tearing.avi

JohnnyRandom
06-25-2009, 02:48 AM
interesting glu solution ;)

nice tear,. physbammer :)

PsychoSilence
06-25-2009, 05:06 AM
some update on the filling bag...this one gets the point more or less :)
and yes, the water needs a ton of more detail but this one calculates in only a few minutes :D

www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_fillingbag2.avi (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_fillingbag2.avi)

jigu
06-25-2009, 08:13 AM
some interesting tests.. Tearing apart one looks great!

about water, It doesn't have volume but hard gravity force which makes water surface bends a lot.. (I mean water has bag curvature instead of volume..) Is it with particle flow or glu3d?

PsychoSilence
06-25-2009, 03:59 PM
some interesting tests.. Tearing apart one looks great!

about water, It doesn't have volume but hard gravity force which makes water surface bends a lot.. (I mean water has bag curvature instead of volume..) Is it with particle flow or glu3d?

it is a farely low amount of particles that are then fed into glu as guide particles. with more particles the water looks more convincing of course. it's a proof of concept video of a "douche bag" :D

PsychoSilence
06-26-2009, 07:40 AM
a little inspired by pixomondo´s cloth burning tool they developed for Red Baron comes my little approach:

http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_BurnAfterReading.jpg

www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_BurnAfterReading.mov (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_BurnAfterReading.mov)

the fire´s resolution sucks (and the boyancy is too high) but again proof of concept that u can pipe the gray scale map that controls binding breaks into fume as emission map...in a more up scale manner this will be awesome for burning plastic or any kinds of synthetic compounds...

Glacierise
06-26-2009, 08:24 AM
The tearing ones are totally great, serious detail dude!

Edit - that burning one is also very promising! Cool tests!

PsychoSilence
06-26-2009, 04:32 PM
updated the QT!!! now the texture is warping fine with the distortion :)

JohnnyRandom
06-26-2009, 06:52 PM
Nice Ansi! Off the bat I can't tell, I assume you are emitting fume from a particle soucre using the pskinner control particles? Cool setup BTW

PsychoSilence
06-26-2009, 07:26 PM
check the screenshot :)

after the pflowAB at the bottom theres only the display...

JohnnyRandom
06-26-2009, 08:44 PM
LOL, F**k yeah, nice bind solvent!

JohnnyRandom
06-27-2009, 03:16 AM
PhysBAMM! :P

Chain QT 12.5 mb (http://www.4rand.com/TEST/ParticleFlow/b2b/PhysXGlue/Chain/chain.mov)

http://www.4rand.com/TEST/ParticleFlow/b2b/PhysXGlue/Chain/chain.png

Glacierise
06-27-2009, 07:05 AM
Hey Anselm, so you're basically solving the glue by texture? Way cool!

@Johnny - dude that looks nice, I thought PhysX doesn't do concaves?

PsychoSilence
06-27-2009, 09:54 AM
hah, htats nice, john! i assume same technique as on the donuts and lots of drag? how do u make them not stick to each others neigbour? i#d love to check out this scene! :D

@ christo: theres an older siggraph paper that describes how to create a concave structure out of convex meshes. thats all we are doing all the time. for the donuts for example we use one tiny cilinder per donut segment and in total when that s glued it makes a full circle e.g. a torus with 24 segments can be described by 24 cylinders aligned exactly as the torus' segments are BUT now its 24 convex meshes defining one concave object :) then u glu all that sh*t :D

JohnnyRandom
06-27-2009, 03:55 PM
Thanks, gotta love glue :)

Glacierise
06-28-2009, 03:05 PM
Thanks Anselm, I'm using that approach all the time in Rayfire, but I hoped something more hm, automated can be done :D

PsychoSilence
06-29-2009, 05:00 PM
some beauty renders from the weekend :)

http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_tearing-ww1.mov
http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_Buddha.mov (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_Budda.mov)
http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_2000Donus.mov

3DMadness
06-29-2009, 05:12 PM
Really nice anselm!

Do you guys use the physx with GPU or by software?

amckay
06-29-2009, 05:24 PM
Ansi, looking good mate!!

btw siggraph is looking more and more likely to happen
gotta fly to belgium, so I'll swing through USA on the way back
BEEEEEER

JohnnyRandom
06-29-2009, 07:27 PM
hah, htats nice, john! i assume same technique as on the donuts and lots of drag? how do u make them not stick to each others neigbour? i#d love to check out this scene!

Can't tell you all my secrets ;) :D, You'll see soon enough, I am building some training stuff for the Man :)

Really nice anselm!

Do you guys use the physx with GPU or by software?

I have tested all three, GPU, PhysX Card, and CPU (software)

The best performance funny enough came on one of my GTX8800's (I have an SLI config, one for graphics the other for PhysX), I figured I'd get better performance with the PhysX card, not so much. Software performs quite well really. I haven't done any benchmarks lately. If/when I get some extra time I'll make up a bench scene.

PsychoSilence
06-29-2009, 08:35 PM
Can't tell you all my secrets ;) :D, You'll see soon enough, I am building some training stuff for the Man :)

awesome! we are establishing something like pool of cool here at work. where u dump all ur test scenes when u figured out something cool.

kogden
06-30-2009, 03:37 PM
You guys look like you have been really enjoying the new pflow stuff! It does look awesome!

I have one other thing I'd like to share though :drool:

Sorry to break the flow BTW...

I've posted a topic on a FumeFX tool I've made, and would love all of you to check it out!

Reason I posted here is its seems to be the place the VFX guys hang out ;)

http://forums.cgsociety.org/showthread.php?p=5956897#post5956897

Sorry for breaking box #2 discussion

All your feedback would be fantastic!

Cheers Kieran.

PsychoSilence
06-30-2009, 07:00 PM
thanks kogden,
that is quite a nice collection of presets :) thansk for sharing

jigu
07-01-2009, 02:07 PM
As box #2 now supports interparticle collision, is it possible to build pile of sand now? (Like it happens in sand glass) Or Particle collision depends on particle size/shape or it can calculates collision between tiny dot shape particle?

JohnnyRandom
07-01-2009, 03:54 PM
As box #2 now supports interparticle collision, is it possible to build pile of sand now? (Like it happens in sand glass) Or Particle collision depends on particle size/shape or it can calculates collision between tiny dot shape particle?

You can pile up particles, yes it is dependent on the PhysX Shape operator, you can for instance have a shape operator set to sphere with a size of 10.0. The PhysX Shape operator will basically wrap that sphere particle shape with a collision primitive, it is adjustable, you can set the PhysX Shape to be larger or small than the particle shape if you want.

PsychoSilence
07-02-2009, 08:36 PM
updated the buddha...more gonna follow :)

www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_Buddha.mov

irwit
07-02-2009, 09:00 PM
Im so impressed with all this box 2 stuff. You just couldnt get these kinds of malleable objects before. I think this has really opened up some hugely creative oppertunities. Well done to all those beta testers, you took your time but a job well done and well worth the wait! Oh and I guess Oleg probably deserve a mention ;)

:applause:

PsychoSilence
07-08-2009, 04:17 AM
www.3delicious.de/spielwiese/BETA/public/PhysXGlue_CaptureNet.avi

some doodle on the side...

Glacierise
07-08-2009, 09:11 AM
That soft bodies are truly amazing, I've been missing this so much...

PsychoSilence
07-09-2009, 04:32 PM
www.3delicious.de/spielwiese/BETA/public/PhysXGlue_In-the-Net.mov

render and slight changes done to make it work better...

JonathanFreisler
07-10-2009, 12:10 AM
Just wondering if there is a 2010 version of the max pflow freebies? More of a question for oleg i guess :D

PsychoSilence
07-10-2009, 01:09 AM
they arent out yet which is saaaad but oleg is still working on box2 full steam still (fix release soon) and on the box3 1.5 beta :)

JonathanFreisler
07-10-2009, 01:18 AM
Shweet, I cant imagine to update them to 2010 would be too hard. Sucks becuase now i cant take my scene back to 2009 lol :S

OlegB
07-10-2009, 02:10 AM
to update them to 2010 would be too hard
Actually, there is a complication there. In 3ds Max 2010 I adjusted a few things related to image motion blur for moving/rotation camera - to make a good IMBlur for particles in scanline renderer. But I did not adjust the corresponding part in Camera IMBlur operator - that part is yet to be finished. So, it's not a straight recompile. And Box#3 1.5 work (as a commercial produce vs a freebie) takes priority.

Thanks,
Oleg B.

JonathanFreisler
07-10-2009, 02:13 AM
Yeah cool no problem. Back to max 2009 for a bit i guess.

Thanks Oleg.

PsychoSilence
07-13-2009, 04:51 PM
updated my showreel with some more beta testing stuff if anyone is interested :)

http://vimeo.com/5567057

irwit
07-13-2009, 05:14 PM
updated my showreel with some more beta testing stuff if anyone is interested :)

http://vimeo.com/5567057

it says I cant view it as it is a private video ?

PsychoSilence
07-13-2009, 05:26 PM
sorry, fixed now :)

Glacierise
07-13-2009, 06:09 PM
Dunno dude, some of the research stuff doesn't look like it belongs there, mayby you should only leave the more spectacular ones. But I LOVE the logo blowaway in the start!

irwit
07-13-2009, 06:47 PM
Looks good! Personally I would say keep the tests. Anyone who knows what they are looking at will appreciate the skills to create them.

3DMadness
07-13-2009, 09:22 PM
I like the tests part, contratulations for the reel! ;)

JonathanFreisler
07-14-2009, 01:44 AM
Oh, on the topic of reel news, I finished mine about a month ago.

Youtube HD (http://www.youtube.com/watch?v=FO0mTxMVxSk&fmt=22)

Or on my slow Australian based website

www.jonathanfreisler.com (http://www.jonathanfreisler.com/reel.html)

Anselm, you make too many reels to keep up with man! Lol, good job. I slightly agree with Hristo, but then again I'm not a fan of people putting 'rnd tests' on there reels. But at least yours are rendered and pretty impressive :P

PsychoSilence
07-14-2009, 02:14 AM
Thanks, Jonathan :)

I like your reel alot! especially the FXWars entry. solid quaility thru and thru :) Im still waiting on all the movie stuff i did in the last 1,5 years...accept the avalance the reel is all commercial works.

PsychoSilence
08-01-2009, 10:10 PM
Hy everybody,
if any of you is going to wonderful new orleans to that thingy called siggraph let me know via PM and we can all meet and greet or just get tasted on bourbon street :D

john and me will demo new box2 and box3 magic at the TurboSquid and Frantic booth. The demoes will be streamed on THE AREA and the TurboSquid website with a little delay (they have to *beep* out all my coursing, i´m a grumpy german who spends way too much time with sassy southeners...).

I dreamed once that i would roam the Siggraph show on olegs shoulders cuz he is sooo big, a titan! No sh*t!...wonder if that´s gonna happen...

kind regards,
Anselm

JonathanFreisler
08-02-2009, 03:40 AM
Man I wish I was going to siggy. And by frantic don't you mean 'prime focus' :P

Have fun who ever is going!

Bobo
08-02-2009, 05:07 AM
And by frantic don't you mean 'prime focus' :P

He surely did. But he is in New Orleans already, I assume the company name changing waves travel slower than the speed of light and have not reached him yet ;)

JonathanFreisler
08-02-2009, 06:23 AM
He surely did. But he is in New Orleans already, I assume the company name changing waves travel slower than the speed of light and have not reached him yet ;)

Then how did it reach Australia? I hear were a bit behind in the scheme of things :P

when will you guys get a new website Bobo and the re branding complete. I dig the new name also.

Bobo
08-02-2009, 10:48 AM
when will you guys get a new website Bobo and the re branding complete. I dig the new name also.

We have a new site for the software side, but it has not been updated with the latest stuff yet. I assume it will be on Monday. Not sure about the VFX site.
We got new emails (old ones still being redirected, of course) and expect the worldwide change of CI (including logo) around September.

OlegB
08-02-2009, 01:29 PM
Particle Flow Tools Box#2 (already available) and Box#3 1.5 release (coming very soon) will be demonstrated by Anselm and John Rand at the Turbo Squid booth #3600 and Prime Focus VFX Technology Inc. booth #3500.

Tuesday, August 4th

10 – 11am by Anselm at Turbo Squid
2 - 2:30pm by John Rand at Prime Focus
2:30 - 3pm by Anselm at Prime Focus
Wednesday, August 5th

10 – 11am by Anselm at Turbo Squid
1:30 - 2pm by John Rand at Prime Focus
2 - 2:30pm by Anselm at Prime Focus
Thursday, August 6th

10 – 11am by Anselm at Turbo Squid
2:30 - 3pm by John Rand at Prime Focus
I'll be around too - to answer the most difficult questions :)
See you at the show!

Thanks,
Oleg B.

PsychoSilence
08-02-2009, 05:30 PM
And by frantic don't you mean 'prime focus' :P
Old habits die hard :D

Awesome to see John demoing too. I knew it when he told me he´s gonna be in town. :D

For all Saturday and Sunday arrivals: There is the Setchmo Festival around the mint in the far east corner of the french quarter. Live music, food, nola heritage and all open air :)

We gonna be there during the day!

kind regards,
Anselm

Glacierise
08-02-2009, 06:36 PM
Jazz and CG... I'm crying I miss that :( Anyway, I'll stay up all-nighters to watch the PFlow presentations dudes, cheers!

Nickolay411
08-03-2009, 12:53 PM
ohmannn... i'll be watching online too :) so freakin excited to see everything unfold.

amckay
08-03-2009, 02:30 PM
Thumbs up guys! I was hoping to make it over there but my flight plans got a little messed up last minute and now I'm helping some ex blur guys on a few cool cinematics. I'll definitely be over there next year though, back to LA!
Hope everything goes well, good luck with the hangovers, and great to see Ansi and Johnny rockin the TS stand!!

Nickolay411
08-03-2009, 06:59 PM
new krak 1.5 stuff... looks really nice

http://www.youtube.com/watch?v=7c4WYzr30B0&hd=1

PsychoSilence
08-08-2009, 10:44 PM
http://www.3delicious.de/spielwiese/siggraph2009_1.jpg

Michael McCarthy, Sir Oleg, Ansi, John Rand, Bobo.

\m/ means "Box#2" not Heavy Metal :D

More to come when I´m back in LA...

Bobo
08-08-2009, 10:52 PM
Michael McCarthy, Sir Oleg, Ansi, John Rand, Bobo.

\m/ means "Box#2" not Heavy Metal :D

More to come when I´m back in LA...

The button I am wearing says "Ask me about Krakatoa". Needless to say I had no voice yesterday, but I dare you - ask me! :buttrock:

galagast
08-10-2009, 09:21 AM
looks like siggraph was a blast for you guys! \m/ :buttrock:

Bobo
08-10-2009, 02:03 PM
looks like siggraph was a blast for you guys! \m/ :buttrock:

Actually, this was my least favorite Siggraph since 1997. It was the smallest, most boring and worst organized one. Even the badges were printed on regular inkjet paper because the organizers had no money for the glossy material with holographic logo they used in previous years. And there was barely anyone on the floors. Recession indeed.

I loved the Autodesk booth which was made from cardboard tubes though. It looked very "green" and at the same time stylish. The Maya and Max logos cut into the tubes made some portions look like a pipe organ. Possibly the best design in years.

As some probably know, the quality of a Siggraph can be measured by the quantity of free T-shirt ("The Strongest Force In The Universe" according to The Brain). While a typical good Siggraph yields between 10 and 12 T-shirts, this one produced only 5, three of them from my own company. I actually had to buy a 6th one, mostly because it looked like made for me:
http://www.rabop.com/pictures/neworleans_090802/thumbnails/832.jpg

Being nearly 42, I did not miss the parties though (if there were any). I know there was a Chapters one and an NVIDIA one (I saw the latter but did not feel like joining) that I knew of, if there were more, I never heard of them. I am sort of thankful Autodesk had no huge Users Group meeting with Max, Maya and Softimage users in the same room. Those people do not mix. Good decision IMHO.

Glacierise
08-10-2009, 04:59 PM
That would just mean you guys take more space, relatively :D

JohnnyRandom
08-10-2009, 05:11 PM
Jeff, Hristo, I was hoping you guys were going to make the trip across the pond :)

I agree with Bobo, I was really surprised at the lack of attendees, although, I had gone to NAB earlier this year and it was the same thing, in years past you needed a cattle prod to walk the show floor this year there was only a 1/3 of the normal attendance which is a HUGE decrease. Hopefully it is just a lull.

Nice to finally meet a bunch of folks out from behind a keyboard!

BY far the Krakatoa Shot Glass was the best swag I came across :D

Ohh yeah and watch out for United Airlines Stewards, since they seem to act more like baggage handlers these days. I caught one slamming a 30+ pound carry on bag, in the overhead bin, into my backpack, which contained my laptop... the LCD is now DOA Argh:banghead:

http://www.youtube.com/watch?v=5YGc4zOqozo

Now I know what this guys was talking about!

charleyc
08-10-2009, 05:34 PM
Dang John. That sucks. Ansi has a nice fly home story too. I will let him tell it. I am thinking of staying pretty clear of United these days :P

This was my first missed siggy in a while, but it sounds like a good one to have missed :(

PsychoSilence
08-10-2009, 05:49 PM
So apparently my luggage never left New Orleans yet!!! I called TSA and we identified my bag 100% as LOST AND FOUND.......and i thought when i drop it off at teh United counter it stays in there posession until i grab it from the conveyer belt at my destination...

they say it didnt have a TAG on it and so after XRay etc they were like "this bag is clear but where the hack does it fly to?!?"

end of story: Im on the phone all day with United to just frakin f*ckin throw my bag on the next birdie to LA (united flies direct twice a day so cant be too hard!) and they wanna go thru the whole "i lost my luggage"-prcedure with me...so after 15 min u finally talk to a real person and he wants to know where I lost the bag....THEY lost it!!! Since i called TSA earler they just need to go over there (it is the counter RIGHT NEXT to united...nola is a small arse airport!), grab the back and ship it...

okay, now i "lost" it :D

msmith81
08-10-2009, 06:26 PM
so... now that Box 2 has been released, what is the word as far as a fully multi-threaded PFlow? I'm always waiting for things to update and I see my new i7 machine with only 1 out of 8 processor cores at work.

JohnnyRandom
08-10-2009, 06:31 PM
Charley maybe we get to meet next year :)

I will say New Orleans is beautiful, first time I have been there, it was great to Ansi and Christine to guide us around! and Bobo to save me from Big O'

Ansi, double argh, I remember when I was little flying used to be fun :)

JohnnyRandom
08-10-2009, 06:32 PM
so... now that Box 2 has been released, what is the word as far as a fully multi-threaded PFlow? I'm always waiting for things to update and I see my new i7 machine with only 1 out of 8 processor cores at work.

Parts of it are multi-threaded, but as with any physics (or standard particle motion) it is cause and effect, something needs to happen before it can figure out what to do next.

PsychoSilence
08-10-2009, 07:27 PM
oh, you still need to send some pictures my way :)

i havent heared back from united yet btw :D

EDIT: i totally forgot to ask the vray guys about mapping object and all those mapping wooees in general we are having in pflow to the very day :((((

msmith81
08-10-2009, 07:43 PM
Parts of it are multi-threaded, but as with any physics (or standard particle motion) it is cause and effect, something needs to happen before it can figure out what to do next.

I'm just using the programs not writing them, I don't have a clue HOW the multi-threading would work in theory. I just know that I need it and the guys at my work who are using XSI are running circles around me with ICE (in terms of speed and viewport playback). So I'll take your answer as no news?

JohnnyRandom
08-10-2009, 10:14 PM
Anse, they are on my laptop :(, it is supposed to get fixed tomorrow (fingers crossed)


msmith81, the answer is not no but it is not fully multi-threaded. I just use them too. :) What I wrote was to help you understand but anyway. What are you trying to do?

You are now comparing apples to oranges, you can always use ICE:shrug:

PsychoSilence
08-10-2009, 10:37 PM
all good :) exitement!

http://www.imdb.com/media/rm3623586048/tt1046173

imdb has more screenshots of GI Joe in the movie's gallery now. finest in krakatoa/fumefx/box3-goodness.

charleyc
08-10-2009, 11:00 PM
Charley maybe we get to meet next year :)

I will say New Orleans is beautiful, first time I have been there, it was great to Ansi and Christine to guide us around! and Bobo to save me from Big O'

Yeah, I think it is back in LA next year. I will definately be attending when it is here.

JonathanFreisler
08-11-2009, 12:11 AM
all good :) exitement!

http://www.imdb.com/media/rm3623586048/tt1046173

imdb has more screenshots of GI Joe in the movie's gallery now. finest in krakatoa/fumefx/box3-goodness.

I shamelessly sore it, more for the FX. I could tell it was fume/krak. Very impressive.

msmith81
08-11-2009, 03:05 AM
You are now comparing apples to oranges, you can always use ICE:shrug:

yeah, I guess I'll just throw my solid 6+ years experience with Max out the window and start over. All that Maxscripting will easily translate to XS... oh wait no it wouldn't I'd have to learn Python. And what about the rigging, skinning, modeling, unwrapping, texturing, rendering, !modifier stack!, expressions, etc am I just supposed to magically know how to do that in XSI just because PFlow hasn't been multithreaded yet? It's a perfectly legitimate question to ask, sorry if I barged in on your little Siggraph party (interesting to read sure, but I was asking a serious particle question and you jump on me and tell me to switch programs?)

From what I can gather, Box 2 and the new version of Box 3 and updating Box 1 for 2010 have taken precedent, but I think it has been hinted that next would be multithreading for the core of PFlow. Thus, my original question.

I use PFlow all the time for medical animation, and particle counts are almost always limited by how much time I am willing to wait for the particles to process. Often shots are very long (>1000 frames) and maybe there is thousands of particles with complex instanced geometry that has to flow by and interact with forces/geometry the whole time. Update/cache times are often VERY long, and everytime I make a change and have to re-cache... I just sit there and watch those other 7 cores goes to waste. 3dsMax is a great tool and Pflow (+Boxes) is amazing, I just want the icing on my cake :cool:

maybe Oleg will stop by and fill us in on any future plans?

PsychoSilence
08-11-2009, 03:32 AM
I use PFlow all the time for medical animation, and particle counts are almost always limited by how much time I am willing to wait for the particles to process. Often shots are very long (>1000 frames) and maybe there is thousands of particles with complex instanced geometry that has to flow by and interact with forces/geometry the whole time. Update/cache times are often VERY long, and everytime I make a change and have to re-cache... I just sit there and watch those other 7 cores goes to waste. 3dsMax is a great tool and Pflow (+Boxes) is amazing, I just want the icing on my cake :cool:

that´s why movies are so expensive :D sorry, couldn´t help it :beer:

you are totally right about the shape instancing and large particle counts. some core essentials in pflow are from max 5.5 days where pflow was an extesion to max5...e.g. a collision object has a 5mio. poly count as maximum and such... that needs to be refurbished asap. But...one thing at a time. Oleg is just one individuum not a company with a team behind him. That´s why some implementations take longer then others. for a 1man show having these tools on the line up is not bad at all. So much for orbaz technologies defense. And now i want proxy instance :D shape instance but working with MR/VRay proxies etc. for exactly the shape instance issue.

kind regards,
Anselm

JohnnyRandom
08-11-2009, 05:29 AM
wow, where did that come from? you have asked the question that everyone particle flow person has been asking for ages.

Is ICE max related? no, then why are you bringing it into the conversation? if you do your are free to absolutely forget everything you have learned in max and start over or ask something that max people can help with :D

No one takes offense to you busting into the party to ask a legitimate question, you are certainly more than welcome, join the party. this thread is a party, about particles, things to do with particles, people that use particles, random bits of jibber jabber that sometimes is not about particles but it all flows(which is indeed better than starting a whole new thread to jibber jabber). Have a look through this thread:)


I use PFlow all the time for medical animation, and particle counts are almost always limited by how much time I am willing to wait for the particles to process. Often shots are very long (>1000 frames) and maybe there is thousands of particles with complex instanced geometry that has to flow by and interact with forces/geometry the whole time. Update/cache times are often VERY long, and everytime I make a change and have to re-cache... I just sit there and watch those other 7 cores goes to waste. 3dsMax is a great tool and Pflow (+Boxes) is amazing, I just want the icing on my cake :cool:


Now your talking or at least explaining something so we understand! I totally agree shape evaluation could definitely use some love. That is IMO the weakest part of pflow (other than multithreading), proxies would be a godsend for sure. Unfortunately the only thing I know of is reducing your viewport count until it is usable, which you probably do already.

Anyone tried point cached animated objects as particles? Does that work?

We all want that icing on the cake:)

SoLiTuDe
08-11-2009, 05:33 AM
^^Just want to throw in my two cents here... are you using box3 insead of the built in operators like position object / speed by surface... i didn't believe it until i got back into it, but box3 is like lightning compared to the standard operators, especially when dealing with animated geo.

and you should learn python anyway... it's awesome :) If you know maxscript well enough the transition won't be that hard, and will even make transitioning to another software even easier... of course learning all that other stuff is the pain in the ass


^Johnny -- haven't tried a pointcached object, (we don't have the latest box3 installed yet for the shape control operator) but there's no reason it shouldn't work... even with offset timing per particle. Using a script operator and having it track object motion to the particles will probably still be faster... I'll have to do some benchmarking when we get the new box3

PsychoSilence
08-11-2009, 06:43 AM
I tried a pointcached fish (as swarm) with random-subOps controlling start and end of the point cache cycle: works like a charme! not a single fish (all having the exact same point cache loaded) moves the same speed... will post a simplified file in the beta forum :drool: if only shape control would read proxies!

OlegB
08-11-2009, 11:25 AM
Often shots are very long (>1000 frames) and maybe there is thousands of particles with complex instanced geometry that has to flow by and interact with forces/geometry the whole time. Update/cache times are often VERY long, and everytime I make a change and have to re-cache...
Did you try to apply the modifications as post-cache operators? Cache Disk (from Box#3) has to ability to apply operators on post-cache basis - therefore you don't need to update the animation. For example, you can change particle shape, mapping, scale as post-cache properties.

Also, please keep in mind the ability to make history-independent operators with Box#3 - this is a great optimization technique. With the release 1.5 of Box#3 the reference documentation will be online too (as with Box#2), so it will be easier to explain.

As for the original question - now, with Box#2 out of the door (and Box#3 almost out of the door), there is some work planned to re-tune the core PFlow. However, I cannot disclose the extent of the changes (yet).

Thanks,
Oleg B.

OlegB
08-11-2009, 11:29 AM
with random-subOps controlling start and end of the point
Ansi,
For optimization sake - did you use Discretizator subop and turning On "Discrete Optimization" in Shape Control subop?

Thanks,
Oleg B.

Glacierise
08-11-2009, 01:00 PM
As for the original question - now, with Box#2 out of the door (and Box#3 almost out of the door), there is some work planned to re-tune the core PFlow. However, I cannot disclose the extent of the changes (yet).
Oleg B.


Yes! :buttrock:

JonathanFreisler
08-11-2009, 01:13 PM
Ansi,
For optimization sake - did you use Discretizator subop and turning On "Discrete Optimization" in Shape Control subop?

Thanks,
Oleg B.

Woot. Shot gun beta-ing.

Its really awesome to have almost instant feedback from some of the great minds developing the tools we love - what more could we ask for really.

msmith81
08-11-2009, 01:46 PM
Thanks everyone for some good tips and answers.

Did you try to apply the modifications as post-cache operators? Cache Disk (from Box#3) has to ability to apply operators on post-cache basis - therefore you don't need to update the animation. For example, you can change particle shape, mapping, scale as post-cache properties.

Yes, I love the post-cache operators, they definitely save me time. I often use Shape, Material, and occasionally Script operators (used to place geometry at the particle position as SoLiTuDe was discussing). But, when it comes to having to change the motion of the particles, it's back to waiting. Often the motion is what needs fine-tuning.

Also, please keep in mind the ability to make history-independent operators with Box#3 - this is a great optimization technique. With the release 1.5 of Box#3 the reference documentation will be online too (as with Box#2), so it will be easier to explain.

I haven't been able to wrap my head around this one yet, maybe I can make more sense of it with the new documentation. Or maybe it only applies to certain situations?

As for the original question - now, with Box#2 out of the door (and Box#3 almost out of the door), there is some work planned to re-tune the core PFlow. However, I cannot disclose the extent of the changes (yet).

Fantastic! That's just what I wanted to hear. :)

PsychoSilence
08-11-2009, 03:54 PM
Ansi,
For optimization sake - did you use Discretizator subop and turning On "Discrete Optimization" in Shape Control subop?

Thanks,
Oleg B.

I did used discrete optimization and discretizer. That was the front end study file i demoed at the booth :) ill clean up and share it in a bit. (still catching up at work)

JohnnyRandom
08-11-2009, 04:17 PM
Thanks guys for the point cached object info, will give it a test tomorrow, swamped today, something exciting to look forward too :)

Great news Oleg! Thanks for the update;)

tryhard
08-11-2009, 04:42 PM
Sorry to interrupt the discussion, but i have a simple question:

Max 2010 comes with a lot of PFlow samples. But how do i get these presets into my Particle Flow Preset Manager ???

JohnnyRandom
08-11-2009, 04:55 PM
AFAIK you have to open each file and save it as a preset. Each Pflow preset has an accompanying text file that the preset dialog draws its data from. So in theory you could copy all of the max files into your:
(this path is truncated) Local Settings\Application Data\Autodesk\3dsmax\2010 - 32bit\enu\plugcfg\Particle Flow Presets --folder

Add a text file with the same name as the Max Scene and within that file add a small description.

So it looks like this:
Bubble Pin Burst.max
Bubble Pin Burst.txt --add a description in this file if you like


EDIT: Just tried it works like a charm :)

EDIT EDIT: This only issue I can see is the scene files that have dependencies like mocap, bitmaps, sounds ect. You will need to add their paths in the User Config dialog.

tryhard
08-11-2009, 05:34 PM
Thanks John, i´ll give that a try ... thought that it would end like that ... :p

OlegB
08-11-2009, 11:59 PM
More online tutorials for Box#2 - http://www.orbaz.com/forum/viewtopic.php?t=2351

Thanks,
Oleg B.

JonathanFreisler
08-12-2009, 12:25 AM
Oh dude, that is one killer help file!

MOUSE SCROLL OVERS!

grury
08-12-2009, 07:37 AM
Thanks John and Oleg, very helpful indeed.

Cheers

Glacierise
08-12-2009, 08:15 AM
Wow seriously nice production quality on these! Getting the bundle this week!

JohnnyRandom
08-12-2009, 04:19 PM
Thanks, I hope and it was our intention that these would be a nice mix between written and video tutorials, since with written it is easy to get tired of reading an overabundance of text and with videos finding that little bit of information halfway through the video that pertians to what you are doing can be a drag (literally) :)

There are 4 more in production, although on hold until they Dell fixes my laptop (which was supposed to be yesterday),

As always any other ideas and feedback are more than welcome :)

amckay
08-12-2009, 05:16 PM
great work johnny!

btw all the 3dsmax pflow elements that ship with max 2010 - do you guys know where the hell they are stored when max installs them? or do you download them separeately? I was curious as I never have been able to find them

PsychoSilence
08-12-2009, 05:22 PM
John did an awesome job maintaining the same style and quality of the help documentation :)

the Box#2 help is one of the best i've seen in the Max world so far.

PsychoSilence
08-12-2009, 05:26 PM
great work johnny!

btw all the 3dsmax pflow elements that ship with max 2010 - do you guys know where the hell they are stored when max installs them? or do you download them separeately? I was curious as I never have been able to find them

i think u need to install them separate! tehn they should be in the MyDocuments/3dsmax/directory

JohnnyRandom
08-12-2009, 07:29 PM
Thanks Anse & Allan, credits to Oleg and Galagast (Jeff Lim) for some of the great example files to build upon.


btw all the 3dsmax pflow elements that ship with max 2010 - do you guys know where the hell they are stored when max installs them? or do you download them separeately? I was curious as I never have been able to find them

They may get installed with Max but I have yet to find them. I grabbed them off the DVD.
They are buried in the samples DVD (I downloaded the ISO from sub so it may be a little different path)

Autodesk_3ds_Max_2010_Autodesk_&_3ds_Max_Design_2010_English_WIN_Sample_Files\Samples\Particle Flow Elements Library

Nice work BTW Allan, some real cool/useful sample files ;)

PsychoSilence
08-12-2009, 10:45 PM
just went thru this years siggraph papers and those two cought my eyes:

http://www.cs.mcgill.ca/~cg/projects/composite/

http://www.cs.caltech.edu/~keenan/project_epifa.html

especially the first link. thats almost like we can do it with box2 :)

JohnnyRandom
08-13-2009, 12:14 AM
Cool links, interesting how they have materials of higher densities reacting within lighter density volumes.

amckay
08-13-2009, 12:22 AM
Thanks Johnny, honestly there was/is a lot of potential with the idea of building elements for max, but basically I was told no plugins, and also my scenes would have to play in real time. I was given a prebuilt list so for the most part it was just making the examples they asked for and a few others.
So I think for a lot of people new to particles there's a lot of things to get them started, and then some more complex examples buried in there too. But nothing too overwhelming.
If I was given the OK to even use afterburn, well that would have a whole different story, but try making cool explosions with sprites and no stock footage or even cool models to work from ;)

So its not too bad, but I was just curious because I hadn't seen them yet
been stuck on the other side of the line for the past year or so so havent had much time to do 3d too busy playing client. Doing some work for a few ex blur guys right now on a cinematic, so its fun to blow some stuff up again ;)

where are you working these days johnny?

amckay
08-13-2009, 12:23 AM
ansi the second one looks like a piece on effects of marijuana on bunnies

JohnnyRandom
08-13-2009, 04:36 PM
For being limit to Peter Draper style of vanilla max (this is totally a compliment :)) they are a great additional resource for people to use/learn from.

Prime Focus Salt Lake City Office :D just wishing, working bread and butter is with a local tv station, plus some side jobs. Too well rooted here, not much for globe trotting unless it is for vacation :)

Yourself, you were in BC not all that long ago? Beautiful place!

PeteDraper
08-13-2009, 04:39 PM
cheeky...!

using fume & afterburn at the moment if you REALLY must know :P

Pete

ps anyone know how to get AB to deal with IOR values beneath 1?? Seems to be clamped off...

JohnnyRandom
08-13-2009, 05:22 PM
Haha, good for you joining in the Volumetric fun, maybe throw some Krakatoa into that fume goodness for some real surprises!

Interesting, I take it you are using Hypersolids? I guess they figured no one would ever need an IoR of less than 1.0. Sounds like bug report time!

PeteDraper
08-13-2009, 05:42 PM
aye - I'll give kresimir a shout

p

EDIT: always have used fume et al... just the books have been plugin free for tutorial reasons

irwit
08-13-2009, 07:52 PM
Jonny, I'm over in Salt Lake in a few weeks, doing a tour of America and I have a friend living over there. Anything I shouldn't miss while I'm there, I hear there is some sort of lake ??? ;) Also I hear there are only 3 bars in whole City, hoping this inst true ... !

JohnnyRandom
08-13-2009, 09:03 PM
Yep there is a big briny body of water full if brine shrimp and mosquitoes, it is nice if you want a reminiscent whiff of the ocean :D

Ahhh, good timing they just dropped the Private Membership requirement for all of the clubs/bars :)

PeteDraper
08-14-2009, 12:28 PM
re - afterburn meta solids IOR - don't fret, sorted it. Left it for a bit and worked on the system, then it just came to me: of course, the Afterburn UI simply locks the IOR value, so it's just a simple case of dialling in the value into its controller. Boink - works first time :)

Pete

PeteDraper
08-14-2009, 12:47 PM
tAnd now i want proxy instance :D shape instance but working with MR/VRay proxies etc. for exactly the shape instance issue.


Definately, but you can obviously get away with scripting mr proxy offsets. I've just finished working on a feature where I built a crowd army system of 12,000 characters with about 30 to 40 different iterations (horse variations, flags, no flags, footsoldiers, camels, clothing variations, secondary dynamics etc etc) all baked out to mental ray proxy libraries. These could be called on by the crowd system which avoided slower characters (footsoldiers, camels), followed leaders (riders with flags). Once the director approved distribution and motion with proxy (static) geometry it could all be baked out to keyframes; this also gives you the opportunity to emit particles or fume systems from the positions of the characters (driven by the original flocking system).

the current system is limited in this sense, ie we can't call directly on proxies, but there are ways around it with a little bit of work. Personally I'd love to see support for mrproxies / vray proxies, but with the inclusion of animation timing offsets, animation speed variables, scaling (either via the internal scale of the proxy etc).

pete

JohnnyRandom
08-14-2009, 06:06 PM
the Afterburn UI simply locks the IOR value, so it's just a simple case of dialling in the value into its controller. Boink - works first time :)

Pete

Great find :D gotta love a clean workaround ;)

PsychoSilence
08-14-2009, 06:13 PM
what exactly are u doing, pete? water bubble-ish type of solid meta shapes with AB?

PsychoSilence
08-17-2009, 04:18 AM
i just test-recorded my first Tutorial for the DVD :)

www.3delicious.de/spielwiese/TUTORIALS/Box3_Intro_MappingObject/Box3_Intro_MappingObject_controller.swf

it highly aims for starters and belongs in the beginners and little tricks section so no news here for you Pros :) we save that for the DVD.

I´m not sure if i like my recorded voice, i think i get a voice over :D maybe Allan would dub me?

EDIT: some stumbling in the video but go easy on me, it was the first time :)

kind regards,
Anselm

SoLiTuDe
08-17-2009, 04:24 AM
'Grats Ansi! People really need to realize how awesome box3 is... and with some tutorials I think people will start to see even how easy it is.

grury
08-17-2009, 07:31 AM
Nice one, mayb finally I'll get into Box#3, altho my priority at the moment is to get Box#2.

Hey Anselm, when is the DVD scheduled to come out?

StevieMac
08-17-2009, 10:24 AM
Hey nice one Ansi! I've kinda avoided Box3 for a while so a DVD on it would be great.

Thanks for thinking of something we need.


Steve

instinct-vfx
08-17-2009, 11:15 AM
Hey everyone,

i am currently using the result of a realflow sim (wich is basically a puddle of fluid on a plane) and pipe it into PFlow to transform the liquid into a specific shape.

To do this i am using a Box 3 DataOp to store the closest point into the MXS Vector channel as the FindTarget "closest surface mode gives weird results and strange distribution.

While it seems to work and the shape does resemble the reference mesh object i created it has an offset in both rotation and position to the actual mesh object.

Any ideas what i might be doing wrong ? Tried playing with various settings and such but i cant seem to get it right.

Kind Regards,
Thorsten

instinct-vfx
08-17-2009, 01:29 PM
Okay am seriously getting stuck here. I tried to workaround now by creating a second PFlow System just to check. And use "closest particle" with "use O1" and selected the second particleFlow.

The second flow is a standard flow with speed turned off.

And my particles are getting the same offset even in that setup :shock:

Any pointers ??

Regards,
Thorsten

instinct-vfx
08-17-2009, 01:33 PM
okay. got it. For reference :

You have to turn off mesh object and switch to "Icon" mode for Maxscript Vector to work in Find Target :x

Regards,
Thorsten

PsychoSilence
08-17-2009, 04:21 PM
Nice one, mayb finally I'll get into Box#3, altho my priority at the moment is to get Box#2.

Hey Anselm, when is the DVD scheduled to come out?

The DVD will cover both boxes and rayfire and a little fume so it might be valuable for u anyways :) The release is highly depending on how frequent i can record which depends on workload. So far im recording every day btw. :)

JohnnyRandom
08-17-2009, 04:33 PM
Nice Anse, can't wait to see the DVD, nice teaser :)

instinct-vfx you can try using the transform matrix channel instead of the mxs vector too, you will get scale and rotation in this channel.

DeKo-LT
08-17-2009, 05:09 PM
Congrats Anselm, finally we will have box#3 materials :thumbsup:
And now, who will take krakatoa? ...especially those mega powerfull 1.5 features ;)

PsychoSilence
08-17-2009, 05:25 PM
Congrats Anselm, finally we will have box#3 materials :thumbsup:
And now, who will take krakatoa? ...especially those mega powerfull 1.5 features ;)

Oh, Krakatoa will be featured too :) just didnt have the time to lay out a road map for that yet so i dont know what i gonna do. Gotta get more familiar with the KCM stuff. For GI Joe we used box3 for everything...So yeah i gotta see what actually ends up being recorded on the Krakatoa side of things. But NOT showing any Krakatoa being emplozed at prime focus would be a shame :D

JonathanFreisler
08-18-2009, 02:08 AM
\just didnt have the time to lay out a road map for that yet \

Yous starting to sound like AD lol. Do you know how many times they say 'road map' for upcoming versions of max? ha ha

PsychoSilence
08-27-2009, 06:18 AM
Speaking of "road maps" I´m open for suggestions! If you want to see something explained i will most likely see if i can make that happen if it is not among the topics already :)

here´s another teaser:
www.3delicious.de/spielwiese/TUTORIALS/Box3_Nanomites.mov (http://www.3delicious.de/spielwiese/TUTORIALS/Box3_Nanomites.mov)

This is nanomites how did NOT do it on "GI Joe". I can´t tell due to NDA but how ever i have my own approach to show similar techniques, yet nothing close to the movie.



flocking driven by animated helpers
smooth change of behavior when approaching a target close enough
eating and eating-specific-flocking when hitting a target
shading based on their distance to the target
rendering including reflections(there are supposed to be metal) using the new Krakatoa 1.5.X
kind regards,
Anselm

jigu
08-27-2009, 06:55 AM
I would like to hear/know why did you use box #3 for it? (ofcourse you used it that means it wasn't possible with normal pflow)

PsychoSilence
08-27-2009, 08:04 AM
a similar effect would be possible with common particle flow operators. the speed decrease in calculation and render time would be hurting though since you would need more operators, forces, etc. this flow has 5 operators (not counting the birth-op, render-op and displays). rendering time on my 4 year old notebook(2 cores, 2gig ram) is under 5 seconds a frame with motionblur starting off with 50,000 particles.
It´s about stripping down what you have in your bag to the minimum of the math needed. I will provide a "pushed further" version that does the effect in 2 operators.

kind regards,
Anselm

Aldarion
08-27-2009, 10:18 AM
Perhaps an insight into how you'd approach doing the whole chunk isolation detection they did for the eiffel tower but in max ? So that pieces that are no longer connected to the main object would break off and fall (because the nanomites have eaten stuff around the chunk).

I know digital domain used houdini but it would be cool if we could do this in max.

PsychoSilence
08-27-2009, 04:06 PM
Perhaps an insight into how you'd approach doing the whole chunk isolation detection they did for the eiffel tower but in max ? So that pieces that are no longer connected to the main object would break off and fall (because the nanomites have eaten stuff around the chunk).

I know digital domain used houdini but it would be cool if we could do this in max.

Good that you bring that up :) To avoid confusion we at prime focus vfx have NOTHING to do with the eiffel tower. that was done with houndini and a good chunk of scripting by the fabulous artists over at digital domain! :)

making of eiffel tower:
http://www.sidefx.com/index.php?option=com_content&task=view&id=1572&Itemid=68

here's what we had business with in the movie:
http://www.evermotion.org/news/show/31447/prime-focus-vfx-helps-elevate-g-i-joe (http://www.*******.com/archives/news/gi-joe.html)

back to topic though for the unlinking and binding you should look no further then breaking binding in box2 using box3 and grayscale maps:
www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_BurnAfterReading.mov (http://www.3delicious.de/spielwiese/BETA/yet-another-take-on-stanford_BurnAfterReading.mov)
(will be covered on the DVD!)
little abstract but hopefully the idea passes along.

kind regards,
Anselm

PsychoSilence
08-30-2009, 10:03 PM
I´m almost finished recording basic and medium level box#3 tutorials :) i leave the master level advanced stuff to oleg, bobo or chalrey... they are the true champions of "boxing".

another sneak peak snippet (kinda hard to read but the guys duck down as the arrow attack):
http://www.3delicious.de/spielwiese/TUTORIALS/Box3_ShapeControl_FINISH.mov

u gonna find in this one:



controlling morph targets and blending them smoothly with the new shape control sub-operator
baking particles
a few notes on speeding up your workflow
...and i updated my cg portfolio if anyone is interested in dropping by :) (if u are not annoyed by me by now... ;-p )

kind regards and have a great sunday ya´ll,
Anselm

jlelievre
08-30-2009, 11:20 PM
These are looking great Anselm! Put me down for a copy! Any idea when you'll be releasing them?

SoLiTuDe
08-30-2009, 11:36 PM
I would like to hear/know why did you use box #3 for it? (ofcourse you used it that means it wasn't possible with normal pflow)

That doesn't mean it wasn't possible with normal pflow at all. I'm finding the speed increase by not using normal pflow ops is insane... something as simple as position object with lock on emitter + animated shape is much much faster with box3.. if you throw a grayscale map in there for speed, the standard pflow ops become useless by comparison. :D I'll see if I can get a video of this out eventually... maybe Ansi will be kind enough to do it instead. I don't have box3 at home, only at work, so it's a bit harder to take the time to do.

PsychoSilence
08-30-2009, 11:49 PM
These are looking great Anselm! Put me down for a copy! Any idea when you'll be releasing them?

i hope to have it in stores for christmas :) most companies buy random stuff at the end of the year for a better tax write off. i wanna catch that wave :) im all about marketing :D so i have to be done with recording in about 8 weeks (thats gonna be tough, im only 1/3 thru yet) and then i will send pre-release versions to magazines and such for (hopefully a good) review so its in the last issue of the year in cgworld magazine for example. TurboSquid will give me 20 copies for give away so what does not end up at a magaines or websites desk i can roll out to friends :)

That doesn't mean it wasn't possible with normal pflow at all. I'm finding the speed increase by not using normal pflow ops is insane... something as simple as position object with lock on emitter + animated shape is much much faster with box3.. if you throw a grayscale map in there for speed, the standard pflow ops become useless by comparison. :D I'll see if I can get a video of this out eventually... maybe Ansi will be kind enough to do it instead. I don't have box3 at home, only at work, so it's a bit harder to take the time to do

Ian has a good point here! I´m all about gettig it DONE, not about WITH WHAT you do it. I wanna show A way, not THE way. i highly encourage people to keep it as simple as possible, use legacy tools where ever u can. i say the parray is still awesome for what it does! how ever speed and workflow incease wise box#3 is a beast! it just takes a little while to tame it :)

and yes, position object with lock on emitter using the box is already recorded and in the release line up :)

kind regards,
Anselm

StevieMac
09-13-2009, 11:01 PM
Hey peeps just made an update to my site & blog & thought you guys might wanna check it out.

The base background is from a tut in Pete Draper's present book, so thanks for that mate.

The rest is just PFlow with FumeFX & Krakatoa.

The breakdowns is on my website
http://www.auxf18.dsl.pipex.com/Particles.htm

& the final animation is on my blog.
http://smactheblog.blogspot.com/


attachment.php?attachmentid=146318&stc=1




Cheers guys.

Steve

Glacierise
09-14-2009, 07:15 AM
Wrote you a comment ;)

StevieMac
09-14-2009, 09:23 AM
Hey Hristo thanks for the comment. Yup your probably right I felt the Krak pass was a little too obvious but it's snow so I thought if I added LOADS of particles it would look too thick.

Nice Rayfire tutorials by the way fella, I'll be checking those out this week! ;0)

Thanks again

Steve

jigu
09-14-2009, 03:21 PM
Thanks guys for explanation.

So having box #3 is more advantage over old pflow operators which can be remade for faster workflow in box #3 and ofcourse it can create new operators as per need without using scripting in pflow.

So in near future, are we going to see faster particle flow operator made by box#3 and completely replacing older pflow operators?

Also Is it possible to make operators using box#3 that can make scene faster for emitting particles from deformable high poly objects? (or still we are limited to use low poly objects for such case..)

Look forward to your DVD, Anselm. :)

SoLiTuDe
09-14-2009, 06:14 PM
Thanks guys for explanation.

So having box #3 is more advantage over old pflow operators which can be remade for faster workflow in box #3 and ofcourse it can create new operators as per need without using scripting in pflow.
So in near future, are we going to see faster particle flow operator made by box#3 and completely replacing older pflow operators?


Well I don't think they'll replace any of the older pflow operators (backwards compatability?) but with box3, we can all share our operators with each other, and use those instead of the old ops when needbe.



Also Is it possible to make operators using box#3 that can make scene faster for emitting particles from deformable high poly objects? (or still we are limited to use low poly objects for such case..)


That's exactly what I was talking about earlier -- and one of the main reasons I'm using box3 for emission right now. I have a scene with a high poly emitter -- that is also deforming. The scene is really fast in box3, and completely unusable with standard pflow operators. Of course you want to cut down polys where you can still, but compared to the standard pflow operators it's LIGHTNING FAST -- you should try it out some time. :D

PsychoSilence
09-14-2009, 10:32 PM
chris bond and chris harvey talk about the work done on GI Joe if anyone is interested:

http://media.fxguide.com/fxguidetv/fxguidetv-ep067.mov

kind regards,
Anselm

JohnnyRandom
09-14-2009, 11:09 PM
Cool, that asset management system is insane! Realtime scene assembly before a render, that's a trip.

Glacierise
09-15-2009, 07:50 AM
Thanks for that link Anselm. FXGuide is overall quite cool, the podcasts, TV and articles.

PsychoSilence
09-18-2009, 12:09 AM
re-recorded and extended the first tutorial on the DVD as a teaser :)

http://vimeo.com/6633107

finished the box3 part with 21 video tutorials (including guiding 35 max files) and about 6hrs footage so far. gonna be a double DVD :)

Glacierise
09-18-2009, 08:06 AM
This is totally gonna shred, looking forward!!

noouch
09-18-2009, 12:59 PM
It's a slow day here, so I decided to dig out an old scene of mine to see if I can give it a touch up now that my scripting/math skills are more up to speed (all vanilla max, cuz plugins are 'spensive ;) ):

http://noouch.de/video/progplant_new_01.mov

The 1 year old version:

http://noouch.de/video/progplant1_test1.mov

kogden
09-19-2009, 01:54 AM
Dude that's cool! out of interest, how did you approach that technique? Because i have no idea were i would start doing something like that :)

Kieran

amckay
09-19-2009, 02:41 AM
Great work Niko, that turned out great!

JohnnyRandom
09-19-2009, 02:42 AM
It's a slow day here, so I decided to dig out an old scene of mine to see if I can give it a touch up now that my scripting/math skills are more up to speed (all vanilla max, cuz plugins are 'spensive ;) ):

http://noouch.de/video/progplant_new_01.mov

The 1 year old version:

http://noouch.de/video/progplant1_test1.mov

I remember that test, very nice progression :)

@kogden, I believe, if I remember correct, he had done it with a super secret script operator :)

my guess is an split amount, spawn, with some sine/cosine something or other, delete by age :D which could all be done via a single script test. Still looks XY though ;) (just messing with ya) :D

PsychoSilence
09-19-2009, 04:49 AM
It's a slow day here, so I decided to dig out an old scene of mine to see if I can give it a touch up now that my scripting/math skills are more up to speed (all vanilla max, cuz plugins are 'spensive ;) ):

http://noouch.de/video/progplant_new_01.mov

The 1 year old version:

http://noouch.de/video/progplant1_test1.mov

thats really kick a$$! Would you be so kind and say Markus "Hase" Hund that i miss him badly ;) he should be at BM right now. btw seriously i know 2 people by now who wanted to recreated your first mograph growing and didnt figure it out. I cant pull it off since i dont know how to script (cuz i have charley sitting next to me ;D) im so glad Long-haired freaky people are in our industry!!! You´re right, plugins are for people who dont know any better...i have no clue :D

noouch
09-19-2009, 01:00 PM
thats really kick a$$! Would you be so kind and say Markus "Hase" Hund that i miss him badly ;) he should be at BM right now. btw seriously i know 2 people by now who wanted to recreated your first mograph growing and didnt figure it out. I cant pull it off since i dont know how to script (cuz i have charley sitting next to me ;D) im so glad Long-haired freaky people are in our industry!!! You´re right, plugins are for people who dont know any better...i have no clue :D
Thanks. No Markus here at the moment, maybe in the coming weeks though.

The setup consists of an emitter animated along a spline spawning a single particle at frame 1. Then there are two split amount operators to create trails and branch particles that inherit the velocity of the parent.

All the script operator does is rotate and scale down the particle speed by a certain amount to give nifty spirals. Clockwise and counter-clockwise is determined by the particle ID.
Bercon did a 3D version that looks pretty cool as well. I'm guessing he used spherical coordinates, but somehow I'm having trouble with those at the moment.

BTW I haven't actually had long hair for quite a while ;)

PexElroy
09-21-2009, 01:48 AM
noouch - that was cool and smooth, dig how it dissipates as it ages & grows; has a lot of digital grace. :cool:

Lawless
09-21-2009, 12:43 PM
Hi Anselm, i did not read all the thread but i'm watching your videos and i read that you're producing a DVD about the tool box, are you going to sell this dvd? there is a release date?


tanks...