PDA

View Full Version : Houdini=77mill particles, Lightwave=0.7 mill



erikals
08-26-2011, 04:29 PM
NT, can this be fixed,...?

Houdini 12> 77 million particles,
Lightwave 10> 700.000 particles,

(or 1 or 1,5 mill if you are lucky..)

P.S. lets make it clear that i'm not asking for 77mill particles here...!
 

Cageman
08-26-2011, 04:40 PM
As you may know, it can be dealt with:

http://www.newtek.com/forums/showthread.php?t=121449

(HC members only).

:)

erikals
08-26-2011, 04:58 PM
hm, weell... not really(?)
doesn't show a much bigger number of particles... (or...?)

caesar
08-26-2011, 05:01 PM
As you may know, it can be dealt with:

http://www.newtek.com/forums/showthread.php?t=121449

(HC members only).

:)


Why the heavens NT didn't free the CORE forums to everyone !!!!!! HC membership and CORE is gone, continue to split the community after the CORE's death is very bad and nonsense.

cresshead
08-26-2011, 05:17 PM
one day there will be news for the HC members that's why.

erikals
08-26-2011, 05:29 PM
yes, some of the CORE code is still gonna go into LW11, so it makes sense.

Cageman
08-26-2011, 05:38 PM
hm, weell... not really(?)
doesn't show a much bigger number of particles... (or...?)

Well... 1 million particles doing collisions with eachother (which is needed for the intended purpose, which I will not reveal here) calculated in... 4 frames / s... try to do that in LW with the current tools. ;)

erikals
08-26-2011, 05:43 PM
;] hehe, well,... for speed, yes, it looks king... http://erikalstad.com/backup/anims.php_files/king.gif
 

geo_n
08-26-2011, 09:42 PM
As you may know, it can be dealt with:

http://www.newtek.com/forums/showthread.php?t=121449

(HC members only).

:)

I thought he stopped work on that a few months ago.
Its going to be tough to finish it if he is only working on it part time with a full time job in the morning. He and Steve Hurley should combine resources to make a killer phsx engine in lw.

Titus
08-26-2011, 09:51 PM
NT, can this be fixed,...?

Houdini 12> 77 million particles,
Lightwave 10> 700.000 particles,

(or 1 or 1,5 mill if you are lucky..)

P.S. lets make it clear that i'm not asking for 77mill particles here...!


LW was one of the first 64 bit CG programs, why there's a limit?

thomascheng
08-26-2011, 10:03 PM
LW was one of the first 64 bit CG programs, why there's a limit?

There limits in every program. Houdini just happens to be the best at FX. Why is Maya's modeling and native render engine so bad compared to LWs?

Cageman
08-26-2011, 10:04 PM
There limits in every program. Houdini just happens to be the best at FX. Why is Maya's modeling and native render engine so bad compared to LWs?

Good points!

:)

Titus
08-26-2011, 10:12 PM
There limits in every program. Houdini just happens to be the best at FX. Why is Maya's modeling and native render engine so bad compared to LWs?

I understand that, but I remember years ago the selling points of LW 64-bit (according to NewTek's advertising and a BSG ship image) being good at managing millions of polygons with no problems, so why it's a limit so low with particles?

Now, I never needed to create such memory intensive models or simulations. I remember how for Pirates of the Caribbean: At World's End, the company in charge used Houdini for the waterfall at the end of the world :D, they only generated a thin waterfall and duplicated the rest in post.

caesar
08-26-2011, 10:33 PM
one day there will be news for the HC members that's why.

NT can't deliver the monthly newsletter anymore...

...ok, some bitter thoughts about NT last decisions...

Celshader
08-26-2011, 11:06 PM
NT, can this be fixed,...?

Houdini 12> 77 million particles,
Lightwave 10> 700.000 particles,

(or 1 or 1,5 mill if you are lucky..)

P.S. lets make it clear that i'm not asking for 77mill particles here...!
 

Until that day, here's two workarounds I've seen...

Buick - Behind the Beauty (http://www.motiontheory.com/content/434/buick_behind-the-beauty) (2010) -- ParticleFX for the "hero" particles on the ice skater, and a deforming point cloud manipulated by bones/morphs for surrounding particles.

The Guardian (http://www.youtube.com/watch?v=L5bpyaF2tvY) (2006) -- massive amounts of RealFlow particles converted into LightWave partigon object sequences using a Python script.

erikals
08-27-2011, 02:32 AM
there are workarounds for sure,
(thanks for the links by the way, very cool stuff :]

but this is not about workarounds, and it shouldn't be, barely 1mill particles just ain't good.

Cageman
08-27-2011, 09:55 AM
Until that day, here's two workarounds I've seen...

Buick - Behind the Beauty (http://www.motiontheory.com/content/434/buick_behind-the-beauty) (2010) -- ParticleFX for the "hero" particles on the ice skater, and a deforming point cloud manipulated by bones/morphs for surrounding particles.

"The design team used proprietary software to create a unique visualization of form, movement, color, and engineering."

So... LW is "proprietary" now?

:)

Red_Oddity
08-27-2011, 10:14 AM
There limits in every program. Houdini just happens to be the best at FX. Why is Maya's modeling and native render engine so bad compared to LWs?

If you're going to compare render engines, then at least compare it to a render engine that is still being developed, like say, MR or Vray.

And seriously, Maya's modeling tools ain't that bad, at least i can model in CC subs for example.


Still, your point is valid, every program has its limits, (to put it into terms of production experience, in LW you'll be wasting 75% of your time finding workarounds, in Maya you'll be wasting 75% of your time fixing reference and render bugs.)

Red_Oddity
08-27-2011, 10:18 AM
Also, XSI's Ice tree is amazing to play with (note i said play, as this program sure brought back a lot of the fun in doing dynamics), the amount it can push around is just silly.
And at this point (or previous point really, 2011.5 Advantage pack), it is absolutely rock solid, in the last couple of days i have not been able to make it crash even once (and i have been doing some stuff that would have made LW or even Maya explode by just thinking about doing it.)

thomascheng
08-27-2011, 01:19 PM
I just want to clearify. I do like Maya and enjoy many of its strengths over LW, but I do find modeling in LW and its render engine to be better than Maya's modeling tools and MR's engine. I'm definitely not saying their tools are bad either.

I won't compare Vray, because it would be better to compare that to Kray. As they are both not included with the package.


(to put it into terms of production experience, in LW you'll be wasting 75% of your time finding workarounds, in Maya you'll be wasting 75% of your time fixing reference and render bugs.)

:agree:

zarti
08-28-2011, 06:27 PM
NT, can this be fixed,...?

Houdini 12> 77 million particles,
Lightwave 10> 700.000 particles,

(or 1 or 1,5 mill if you are lucky..)

P.S. lets make it clear that i'm not asking for 77mill particles here...!


hi ,

seems you have seen the latest video from H12 in SIG , right ?

if so , there were other more ' deep in Core ' changes which NT can take as a good reference , imho .

particles benefit a Side Effect from the Core Changes ( Wow ! what a sentence :D )

believe me , i have nothing against particles personally ..


=)

jasonwestmas
08-28-2011, 08:05 PM
Damn Houdini 12 must be good. It's the only software that was expensive to begin with but the price keeps going up :D

Sensei
08-28-2011, 09:30 PM
77 million particles.. ? Maybe it's a fake? Illusion dynamically created..
Let's calculate how much memory is needed- 3 floats * 4 bytes = 12, position, size, direction and color 12 * 4 = 48, at least, * 77 million = 4 GB..
I can't imagine either baking it to file (2 TB disk enough to keep 500 frames of raw data without compression.. ?), nor loading it, nor running simulation (weeks of calculations? so must be baked to file)..

Titus
08-28-2011, 09:55 PM
I don't find it impossible. Blender implemented LZW-compressed .obj to store their meshes for fluids, they load fast. I know, a baked mesh isn't the whole domain cache but the info is being transferred in front of the user, maybe they are using algorithms like a KD-Tree, or whatever. Also, in my experience using a domain size (before a Blender crash) of 1.5 GB is common.

zarti
08-29-2011, 03:38 AM
77 million particles.. ? Maybe it's a fake? Illusion dynamically created..
Let's calculate how much memory is needed- 3 floats * 4 bytes = 12, position, size, direction and color 12 * 4 = 48, at least, * 77 million = 4 GB..
I can't imagine either baking it to file (2 TB disk enough to keep 500 frames of raw data without compression.. ?), nor loading it, nor running simulation (weeks of calculations? so must be baked to file)..

im not good at calculating memories but considering the ' audience ' in that presentation ..

.. i wdnt say something like " fake " .. =)



.cheers

creacon
08-29-2011, 07:50 AM
I put it in the fridge for the time being, I only did this as a test for several reasons:

- I wanted to know if it could be done.(it can!)
- Nobody was considering PhysX for dynamics and it seems like a well documented mature engine. (it is!).
- I wanted to check if the LW sdk was really as bad as some people told (it isn't)
- the weather was bad so I couldn't go out :-(

Now Steve H is implementing PhysX (among others) and (I think) he's a full time developer, so it will only take some time before he starts implementing liquids into his solution (not much work because PhysX will take care of the hard work).
I really didn't do anything special, I spent more time on getting into the LW sdk than I did on PhysX. A newtek developer could have done this in 2 days.

creacon


I thought he stopped work on that a few months ago.
Its going to be tough to finish it if he is only working on it part time with a full time job in the morning. He and Steve Hurley should combine resources to make a killer phsx engine in lw.

archijam
08-29-2011, 09:00 AM
creacon: thanks for the summary, very concise, yet enlightening :thumbsup: !

ps. awesome work on your website, love the intro vid! :D

prometheus
08-29-2011, 09:45 AM
Dream quest images wrote the proprietary software for the gas dust stuff of the armageddon asteriod within 6 months.."Hookah" was the nickname of it.
http://digitalcontentproducer.com/mag/video_closeup_armageddon_bruckheimer/

Also used on mission to mars..
http://www.cgw.com/Publications/CGW/2000/Volume-23-Issue-4-April-2000-/MISSION-POSSIBLE.aspx

http://scribblethink.org/Work/Notes/hookahVariance.pdf


I can wait six months for something like that Newtek:D

Edit...
DreamQuest's animation team created the vortex using the Hookah renderer that they initially designed for the particle "gas jets" shooting out of the meteor in Armageddon. "The technique of creating the gas itself was procedural," he elaborates, "but our animators have been working on Hookah to allow us to control those procedural elements via animation to create a character. It takes two different types of animators: a traditional animator animates the character, and then the Hookah procedural animators called TDs actually translate that into the final 'look' of the vortex. In other words, a traditional animator moves a series of concentric tubes to create the motion of the vortex, without knowing exactly what's happening inside the vortex. Then that position geometry goes to the Hookah animators, who add these multiple concentric tubes, which work within the Hookah software to control the forces inward, the forces outward, and the turbulence, to get the look or the behavior of the outside Hookah force. There is no solid form; using Hookah, we created this amazing, surf-like vortex that consists completely of a hundred million particles or more. It takes forever to render, and it's a mess to light, but the aesthetic is interesting and spectacular. I've never seen anything like it, which is pretty cool."

Michael

creacon
08-29-2011, 10:59 AM
You can see the new commercial here:

http://www.mm.be

Click on the crab and then on the link below the picture. (for those who don't speak dutch, which would be 99.99999% on these forums :-)

And btw I'm not making those, there's a whole team here ;-)

creacon


creacon: thanks for the summary, very concise, yet enlightening :thumbsup: !

ps. awesome work on your website, love the intro vid! :D

tischbein3
08-29-2011, 02:57 PM
a little in defence for the old lady:

If you really need to deal with such an hudge amout of particles, you would, no matter what program, even houdini, try to build your scene up in layers:
a smaller workeable amount of particles (displayeable in realtime) wich later controls a larger amounts of particle subsets / get clones, all baked in smaller manageable files.

Although lw 10.1 has a 1million particle per emitter cap, by using multible instances / bake them. you are able to archive higher particle counts and also get the following advantages:

- Better control (subsets can be subject to different forces / interactions)
- Better Scene buildup (a "calculate all in one shot" technique is a time waster when it comes to refinement)
- HV might render significant faster depending on the scene
(BB precalculations)
- The possebillity to use Multible cache files.
- The possebillity to turn off particle visibility for partial (or almost all subsets) in viewport greatly reduces memory usage and higher performance, without actually lose the ability to work with the scene.
- Easy to splitup scenes to render in layers.

as allways, just my 2 cents on this,
chris

(but I agree: in general particle ram usage is quite high)

erikals
08-29-2011, 03:11 PM
i know, i use this technique, and it's a good one, but is it really that hard to make room for more particles?!

i had a vague hope that buy doing so we would have more accurate simulation of particles.
(probably not... http://www.newtek.com/forums/showthread.php?t=121921)

Cryonic
08-29-2011, 03:57 PM
"The design team used proprietary software to create a unique visualization of form, movement, color, and engineering."

So... LW is "proprietary" now?

:)

now? more like has been from the beginning, heheh.

manholoz
08-29-2011, 04:09 PM
- Nobody was considering PhysX for dynamics and it seems like a well documented mature engine. (it is!).


Noooo please no goodie tied to specific hardware! I am not a hardcore :D member, so I have no idea how the well Bullet physics engine was being implemented. But I hate when I am tried to be forced into any software/hardware/etc by anyone/anywhere. That is why I use Lightwave, maybe it is not as hyped, but pretty useful and open. :newtek: Please keep it open!

creacon
08-29-2011, 04:36 PM
If you want the job done, you buy what's necessary. And anyway you don't need an NVidia card to use PhysX.
If you want eg rigid bodies coliding with fluids you could wait a few years until bullet gets that or use PhysX now. Nobody would be forcing you, you have the choice to use it or not, isn't that great :-)

Nobody is forcing you to run windows or OSX but if you want to run LW you'll have to.

creacon




Noooo please no goodie tied to specific hardware! I am not a hardcore :D member, so I have no idea how the well Bullet physics engine was being implemented. But I hate when I am tried to be forced into any software/hardware/etc by anyone/anywhere. That is why I use Lightwave, maybe it is not as hyped, but pretty useful and open. :newtek: Please keep it open!

archijam
08-29-2011, 04:46 PM
You can see the new commercial here:

http://www.mm.be

Click on the crab and then on the link below the picture. (for those who don't speak dutch, which would be 99.99999% on these forums :-)

And btw I'm not making those, there's a whole team here ;-)

creacon

Gek en leuk ook :thumbsup: ..

Veel Komplimenten aan julie Kameraden!

(Ik mis het Nederlands, zo een mooie Taal .. Vlaams is natuurlijk ook wel iets ;) .. beetje mooelijk voor mij)

erikals
09-01-2011, 02:22 PM
Although lw 10.1 has a 1million particle per emitter cap, by using multiple instances / bake them. you are able to archive higher particle counts...

a minus here is that the particles won't blend when doing fluids...

Bill1955
09-02-2011, 11:15 AM
So... LW is "proprietary" now?
:)

You can get an answer to to that question only in the HC forums.

Bill1955
09-02-2011, 11:17 AM
Damn Houdini 12 must be good. It's the only software that was expensive to begin with but the price keeps going up :D

Don't forget Realflow, it went from expensive to "forget about it."

erikals
09-02-2011, 11:21 AM
strange realflow has no competion,... just glu3D, but that's Max only, C4D also has some i think.

wish they could push Blender fluids a bit more... (more particles, more accurate, less flicker)

Titus
09-02-2011, 11:40 AM
wish they could push Blender fluids a bit more... (more particles, more accurate, less flicker)

Blender fluids are being improved a little every version. Next update will work in realtime:

http://www.blendernation.com/2011/01/03/fluids-in-real-time-with-opencl/

EDIT: I've talked with Ton and the other Blender guys to improve fluids. I'm really interested in this topic, actually was looking into pay someone to
implement a better memory management for fluids.

zarti
09-02-2011, 11:42 AM
Damn Houdini 12 must be good. It's the only software that was expensive to begin with but the price keeps going up :D

chronologically , No . ( afair )

.. or are you talking about its features and performance ' going Up ' ? :D

erikals
09-02-2011, 12:38 PM
Blender fluids are being improved a little every version. Next update will work in realtime:

http://www.blendernation.com/2011/01/03/fluids-in-real-time-with-opencl/

EDIT: I've talked with Ton and the other Blender guys to improve fluids. I'm really interested in this topic, actually was looking into pay someone to
implement a better memory management for fluids.

that looks sweet :] <3
(and free...)

jasonwestmas
09-02-2011, 12:50 PM
chronologically , No . ( afair )

.. or are you talking about its features and performance ' going Up ' ? :D

Oh yeah, i must've been looking at a multi-licensing plan or something for 12K. Master is still 7K like it has been for a while now. My mistake.

Red_Oddity
09-02-2011, 03:01 PM
strange realflow has no competion,... just glu3D, but that's Max only, C4D also has some i think.


Yes it does have competition, it's called Naiad (http://www.exoticmatter.com/), and when Realflow is out of your price range, you can just forget about Naiad.

Titus
09-02-2011, 03:08 PM
Yes it does have competition, it's called Naiad (http://www.exoticmatter.com/), and when Realflow is out of your price range, you can just forget about Naiad.

Not only that, you need an IT expert to install and use Naiad.

erikals
09-02-2011, 03:24 PM
i'll just buy a big pool...

50one
09-02-2011, 03:57 PM
Also a Flowline VFX - German company, got it's own solution for water VFX, but not available to mortals like us lol.

Got Question and this will be the best place to ask..since all the guys are here hehe, anyway recently updated Lw to v.10, I've always been using LW for small tasks like getting shaded object and then doing paintover, but want to jump into the particle field (I've got a sh*tload of wordware books - but never read the particle / F parts hehe) so...the questions is, well in fact few Q:

1) will Lw play nice with other packages?importing the particle data from different apss 3ds max / Softimage, If I've got exporter?

2) What is a partigon? What's the benefit of using it, if any.

3) I know that Dynamite is dead, but anyone knows if dynamite was using it's own shader? or was that HV itself - just driven smart by the plugin itself

Ta in advance!

erikals
09-03-2011, 02:04 AM
1 the plugin point oven will work, or try to export particles directly with LWs collada/mdd or LW10geocache plugin
2 a partigon is 2 point that are merged into a polygon. few benefits, search this forum for partigon.
3 yes, it was it's own shader, just like Turbulence4D for Lightwave.

erikals
09-03-2011, 11:19 AM
a minus here is that the particles won't blend when doing fluids...

sorry, have to correct myself here, they do, make a group and set all emitters to use that group (set to additive).
might not blend super, but they do blend...

jasonwestmas
09-03-2011, 11:29 AM
sorry, have to correct myself here, they do, make a group and set all emitters to use that group (set to additive).
might not blend super, but they do blend...

They blend like oatmeal blends =]

Celshader
09-03-2011, 12:53 PM
1) will Lw play nice with other packages?importing the particle data from different apss 3ds max / Softimage, If I've got exporter?

A co-worker told me that if you have the RealFlow5 plug-ins, you can convert particle data into RealFlow *.bin sequences and then back again into the native format of another package.

The plug-ins are not posted on the Next Limit site. In the past, if you signed up for the RealFlow demo version, you had a 1-day window to download the RealFlow demo and all of the RealFlow plug-ins. I'm not sure if that is still the case.


2) What is a partigon? What's the benefit of using it, if any.

Short version: partigons render faster than HyperVoxels, but partigons have less individual detail than HyperVoxels.

A partigon is a one-point polygon. It renders as a dot in space. In Modeler, you can create a partigon by selecting a single point and typing "p" to make a one-point polygon. You can also use "Points to Polys" in Modeler to convert a point cloud into a cloud of partigons.

In the Object Properties panel, you can control the pixel size of these dots by entering the pixel size in the Particle/Line Thickness setting of the Edges tab. With positive values, the dots render with scanline and do not respect raytracing, only shadow maps. With negative values (introduced in the 9.x series), the partigons render as actual spheres of geometry, and these spheres will respect raytraced lighting in addition to shadow maps.

When rendering massive amounts of particles, sometimes 1-pixel-wide dots are all you need, because the camera never gets close enough to inspect individual particles. The Guardian (http://www.youtube.com/watch?v=L5bpyaF2tvY) used LightWave partigons to render most of the RealFlow-generated effects. The waterfalls at 0:26 in this video (http://www.youtube.com/watch?v=G-bZ5BJqAqw) are RealFlow data rendered as partigons with a negative thickness in LightWave.

-+-

In LightWave dynamics, you have a choice of creating an "HV emitter" or a "Partigon emitter." If you replace a particle emitter with a one-point polygon, it will become a "partigon emitter." If you replace a partigon emitter with a null, it will become an "HV emitter" again.

jasonwestmas
09-03-2011, 01:06 PM
That's a great point, when rendering fast moving and spraying water, it's just a white material at that point. Partigons would work great in that case. What does the negative thickness do exactly? Create a more believable shadowing of the water?

erikals
09-03-2011, 01:21 PM
short, it creates a sphere :]

jasonwestmas
09-03-2011, 01:25 PM
short, it creates a sphere :]

hmph, I would have thought it would be a sphere by default, and not just a flat dot.

erikals
09-03-2011, 01:31 PM
nope, it's flat....

http://www.newtek.com/forums/attachment.php?attachmentid=97914&stc=1&d=1315074686

Celshader
09-03-2011, 01:35 PM
hmph, I would have thought it would be a sphere by default, and not just a flat dot.

Positive values create scanline dots; negative values create raytraced spheres. Scanline dots can only see shadow maps for their shading, but raytraced spheres will respect raytraced shadows and radiosity shadows in addition to shadow maps.

erikals
09-03-2011, 02:11 PM
Jennifer knows :]

(nice to see you back here Jennifer :]

Celshader
09-03-2011, 02:50 PM
Jennifer knows :]

(nice to see you back here Jennifer :]

Thanks. :)

Thanks also for posting the renders (http://www.newtek.com/forums/showpost.php?p=1179164&postcount=55) of partigons with positive and negative thicknesses. That's a great illustration of the difference between the two types.

erikals
09-03-2011, 04:46 PM
you're welcome :]
that wasn't much though, check my youtube channel, LoL :]

dwburman
09-03-2011, 09:09 PM
Also, partigons are surfaced using the surface editor, and they turn to lines/streaks when you turn on on particle blur in the camera options.

prometheus
09-04-2011, 06:34 AM
Also a Flowline VFX - German company, got it's own solution for water VFX, but not available to mortals like us lol.

Got Question and this will be the best place to ask..since all the guys are here hehe, anyway recently updated Lw to v.10, I've always been using LW for small tasks like getting shaded object and then doing paintover, but want to jump into the particle field (I've got a sh*tload of wordware books - but never read the particle / F parts hehe) so...the questions is, well in fact few Q:

1) will Lw play nice with other packages?importing the particle data from different apss 3ds max / Softimage, If I've got exporter?

2) What is a partigon? What's the benefit of using it, if any.

3) I know that Dynamite is dead, but anyone knows if dynamite was using it's own shader? or was that HV itself - just driven smart by the plugin itself

Ta in advance!

Dynamite used It´s own volumetric voxel shader, nothing to do with hypervoxels.

Dynamite had It´s internal fireshader wich had temperature controls and rayleigh scattering much better than Hv´s rayleigh..you had control of how much rayleigh scattering there would be..and alos light and rayleigh boost controls.

you also had cooling control..wich made it easy to set up a nice internal shading to control how much red fire would be inside of the voxels or if you would like to cool it of more.

Dynamite also used metaballs blending wich we still are waiting for with hypervoxels..right!
Dynamite could also take use of polygons for size and placement of the voxels and rotation control.

There were downsides thou, I didn´t like the noise functions very much..only the cell smoke noise..you could use lightwave´s own fractals but they were not working properly.

Some render artifacts can occour and it was in my meaning slower than hypervoxels.

And of course since It´s dead in the waters..no support, and I found it not
working to install with lightwave 10.1..the whole computer system freezes.
Lightwave 10 works, but I get crashes when working with vpr and particles with dynamite..a single null object seems to work thou.


Michael

jasonwestmas
09-04-2011, 08:57 AM
Not sure I'd want to make a hobbyist programmer app my main method of working full time.

jasonwestmas
09-04-2011, 09:02 AM
With Core in mind, I found that statement quite hilarious.... :D

Oh is that what Newtek hires? Hobbyist programmers? Yikes! ;)

Titus
09-04-2011, 10:59 AM
I would not be so sure of that. ;)
Old problem with most of independent and student Blender development

"Out of the 17 students who joined our Google Summer of Code project, 15 made it for the midterm and all had their projects approved, delivering results exceeding our expectations in many cases!"

http://code.blender.org/index.php/2011/08/google-summer-of-code-round-up/

Not a bad year for Blender. Always the most active contributors for blender are students, many even since highschool.

jasonwestmas
09-05-2011, 08:58 AM
Absolutely. It's a very healthy, active and often cutting edge community. Don't quite get your point, but you probably didn't get mine. :)

Titus could be reacting to my pessimism with hiring hobbyist students to work on professional full time software.

tischbein3
09-05-2011, 09:41 AM
i had a vague hope that buy doing so we would have more accurate simulation of particles.
(probably not... http://www.newtek.com/forums/showthread.php?t=121921)
Not sure what to answer on this...
subframe calculation settings would be nice to have for such stuff...
some better force damping options for inter particle collisions... and so on, its really a can of worms you are opening for me with this.

jasonwestmas
09-05-2011, 10:06 AM
Well, once you pay a hobbyist/student, he's just become a professional. ;) We all had to start somewhere...

Yeah, well I know what Students can be like. I was one once. ;)

Titus
09-05-2011, 10:11 AM
Absolutely. It's a very healthy, active and often cutting edge community. Don't quite get your point, but you probably didn't get mine. :)

I got the point, and never took it personal. I'm happy to know the success of another GSoC.

It's sad to know the realtime fluids developement is stalled. But now we have camera tracking :D.