PDA

View Full Version : Should I fiddle with hypervoxels or the old dead Dynamite plugin??



prometheus
09-05-2011, 06:45 AM
Uhhmm...what should I play with:D

Dynamite...
http://vimeo.com/28587261

hypervoxels...
http://vimeo.com/28074469

some would say use turbulence instead or use real life footage...
sure yeah, but I want the explosion to be very directable..ergo using particles might be the best way to work on that.

Sure you could mix turbulence and drive the fluids with the particles, I will try that later on, right now the latest turbulence update has removed the single scattering wich means I cant really work with it properly to see shadows etc, waiting for an update of that, and besides voxels are resolution independent.

Now dynamite is all gone and no support, and sadly 10.1 lightwave version is crashing/freezing when trying to start dynamite..so this was rendered out in lightwave 10.0 wich works...something with colorspace perhahaps..dont know?

Michael

erikals
09-05-2011, 07:28 AM
how 'bout... :] OGO Taiki.... :]

http://www.youtube.com/watch?v=oW0ImR2zZ9M

prometheus
09-05-2011, 07:38 AM
how 'bout... :] OGO Taiki.... :]

http://www.youtube.com/watch?v=oW0ImR2zZ9M

Well perhaps if someone could answer the question what the developer of it has been doing since the last "beta release a couple of years ago"
Seems dead in the water too..

But I get your point, ogo seemed to blend nicely..however slower than voxels I guess if you were to compare it on particles.

And I donīt think you can get the same color control in order to create temperature looks and fireshading, or can you create an example of that erikalīs?

Once again..I really wished Newtek could have bought this and polished ogo taiki and replace skytracer, as I mentioned before..ogo was volumetric long before vueīs spectral skies.

Again..dead in the water plugin...and Newtek canīt get hold of the guy?
or donīt think Itīs suitable to put in Lightwave?

For dynamite I know that Newtek tried to reach cantarcan, but they didnīt get any response according to jay roth.

Bakudan..was something that could have taking over the particle engine and replace dynamite, but it seemed like a ballon bursting when Turbulence arrived, would have been nice if they could have worked on the particle voxel engine solely.

Michael

tischbein3
09-05-2011, 08:34 AM
Sure you could mix turbulence and drive the fluids with the particles,

Although I only did some experiments with smoke
I found the results quite rewarding mixing t4d (general look) with hv (details).

speismonqui
09-05-2011, 12:11 PM
wow, dynamite looks really good IMO, too bad is not supported any more (for a very long time).

looking forward to T4D??

prometheus
09-05-2011, 01:43 PM
wow, dynamite looks really good IMO, too bad is not supported any more (for a very long time).

looking forward to T4D??


Im looking forward to a fireshader in hypervoxels and metaball blending for a starter, thatīs what we need to compete with afterburner for max.
At least a simple load of gradients would be nice.

turbulence fluids is another beast, a little trickier to get the same light and fire scattering inside volumes, and youd have to calculate a lot, and you are dependent on the final resolution.

otherwise the settings in the turbulence fire tab, such as white point and setting a good blend between low temp and high temp and together with proper density curves will get similar...but is harder to learn and acheive me thinks.

Im awaiting for a fix of the scattering removed option that was brought in to the last update of turbulence, I donīt find it good to work with unless I can see the full shadowing in viper o VPR wich seemed broken now.


Michael

speismonqui
09-05-2011, 05:37 PM
so, afterburn is definitively easier than T4D?

Mr Rid
09-05-2011, 10:53 PM
Well, Fume is the pyro tool- http://www.youtube.com/watch?v=Mh-OLQV0DEk&feature=player_embedded by a mad genius I've worked with.

Intuition
09-06-2011, 04:55 PM
I remember loving Dynamite. Had that great hypervoxel particle rotation first. Was a simple setup too. Did quite a few fx with it.

Too bad it fell into obscurity. But, doesn't LW's particles now have rotation native now? Like you can rotate the hypervoxels natvely?

prometheus
09-08-2011, 12:08 PM
I remember loving Dynamite. Had that great hypervoxel particle rotation first. Was a simple setup too. Did quite a few fx with it.

Too bad it fell into obscurity. But, doesn't LW's particles now have rotation native now? Like you can rotate the hypervoxels natvely?

I donīt think hypervoxels has the same good rotation options, You can use particle orientation as an option.

Dynamite had/has some good random rotation options and a channel to put on
different gradients such as distance to object,particle age,velocity etc.

The goodies is the internal fireshader definitly, and metaball blending and polyons wich adapt voxel size/placement according to the polygons size and placement.
Also not to forget the open gl preview of dynamite with temperature representation on the display.

I Got to work a little more on the texturing and lighting, I would want a fix
from newtek to get dynamite to work with lw 10.0. vpr crashes ..and lw 10.1...doesnīt work.

It would sure speed up some tweaking and ergo cooler stuff will be produced.

Michael

jwiede
09-08-2011, 03:47 PM
Yep, if there's any easy way to get Dynamite working again on 10.x (from Newtek's side), it would be very welcome. There are still things Dynamite can do which HVs cannot, nor which are easily switchable to Turbulence. It's such a shame to lose this excellent plugin.

Portnoy
09-09-2011, 01:41 PM
Yep, I remember that plugin. And I agree with the other it's too bad it's not still supported that render looked the best.

Downside to 3rd party plugs, you take the chance of the developer just deciding not to continue developing it. And it's a shame when you come across a good one like that one.

jasonwestmas
10-18-2011, 04:08 PM
So yeah, needed to get into Dynamite again today, just doesn't work in 10.1 at all. Locked up my computer. . . Bummer.

erikals
10-18-2011, 04:31 PM
Turbulence4D is the way these days... (or reinstall 9.6)

jasonwestmas
10-18-2011, 06:01 PM
Turbulence4D is the way these days... (or reinstall 9.6)

I didn't see much in regards to roundish T4D explosions. I would suspect it does that like Dynamite did.

erikals
10-18-2011, 06:04 PM
yap, pretty much,...
http://www.youtube.com/watch?v=CJLrJgy7h3Q

(the above video is an early version of the plugin)

jasonwestmas
10-18-2011, 06:24 PM
Like this:

http://www.youtube.com/watch?NR=1&v=fK7kvB4iCC4

Kind of stuff Dynamite was good at, the rolling plumes of fire.

jasonwestmas
10-18-2011, 06:29 PM
This one is ok just need more texture detail in the fire:

http://www.youtube.com/watch?v=o5ZMTPIZLvQ

I guess I should just try the demo.

prometheus
10-19-2011, 12:38 AM
So yeah, needed to get into Dynamite again today, just doesn't work in 10.1 at all. Locked up my computer. . . Bummer.


Yepp it locked my computer completly too..



Turbulence4D is the way these days... (or reinstall 9.6)

You could try the 10.0 version I used that..and it works(beware of color config ui conflicts) and to some degree the vpr works with it too, but on this setup I got particle carrier and trailiers I got crashes using vpr, but on other particle setups I donīt ??
Dynamite and particles..
http://vimeo.com/28587261

Yeah turbulence is the future perhaps, but still particles with voxels are better for some stuff, If I need to design the explosion more explicity, then particles and voxels are more easier to work on, especially for rocket trails, aeroplane trail fire etc.

Sure you could use particles with turbulence too, but thereīs issues with setting proper resolutions and dynamics etc and it will probably render slower too.

Im waiting for a new version update of turbulence from jasha thou, I lost interest of it since the last version where he removed the shadow depth, or single ray scattering? simple put I couldnīt tweak the fire and smoke and get feedback on how the shadows in the volumetric smoke looked like.

Michael

prometheus
10-19-2011, 12:42 AM
I should mention that I really loved the internal rayleigh scattering in the fire shader of dynamite, you could get similar stuff with turbulence, but itīs harder to set up.

Michael

jasonwestmas
10-19-2011, 07:52 AM
Thanks for responding Michael. Yeah I love the way dynamite looks for those big billowing puffs of fire. I suppose I can just use 9.6 or 10.0 and just composite it in later, I should be able to just use the mdd animation as place holders in 9.6 and use that as reference for where to put the fire and smoke and then just render the fire and smoke separately. Treat it more like video FX.

In T4D For some reason, I couldn't get the texture stuff to work with the volumes during an F9 or in VPR. Is that a limitation of the T4D learning edition?

prometheus
10-19-2011, 10:12 AM
Thanks for responding Michael. Yeah I love the way dynamite looks for those big billowing puffs of fire. I suppose I can just use 9.6 or 10.0 and just composite it in later, I should be able to just use the mdd animation as place holders in 9.6 and use that as reference for where to put the fire and smoke and then just render the fire and smoke separately. Treat it more like video FX.

In T4D For some reason, I couldn't get the texture stuff to work with the volumes during an F9 or in VPR. Is that a limitation of the T4D learning edition?

Did you check the "generate texture coordinates" in the simulation tab?

Michael

jasonwestmas
10-19-2011, 11:05 AM
Did you check the "generate texture coordinates" in the simulation tab?

Michael

Interesting, I don't have that check box. Now how am I supposed to accurately evaluate this thing without texture. ;)

Possibly I need to turn something else on? I just have a prodcedural texture added to the emitter.

prometheus
10-19-2011, 11:58 AM
Interesting, I don't have that check box. Now how am I supposed to accurately evaluate this thing without texture. ;)

Possibly I need to turn something else on? I just have a prodcedural texture added to the emitter.

Weird...and you cant expand/drag the turbulence UI window to see if thereīs more options beneath the extrapolate in to solids?

Otherwise it might be that something has been moved in the UI between mine and your version (check the other tabs for it), I think you have a later version installed than I have.

I would suggest to mail jascha directly otherwise.

Michael

Elmar Moelzer
10-19-2011, 01:47 PM
I dont really see a point in NT supporting a product that cant even be bought anymore. I would rather see them invest the time in fixing up HyperVoxels and adding the few features that it is missing on Dynamites particle rendering (not the fluid dynamics obviously, but T4D should take care of that).

jasonwestmas
10-19-2011, 04:34 PM
Clearly though dynamite is quite different than T4D in appearance from the examples I've seen. Just saying.

prometheus
10-19-2011, 11:20 PM
I dont really see a point in NT supporting a product that cant even be bought anymore. I would rather see them invest the time in fixing up HyperVoxels and adding the few features that it is missing on Dynamites particle rendering (not the fluid dynamics obviously, but T4D should take care of that).

I almost agree with you, but there is a point, dynamite is working for 10.0 and I donīt believe it would be that hard to fix, not sure thou.

And a big point is that we have no clue if anything will be improved on hypervoxels at all in november or in the close future, and even if it will match dynamiteīs voxel engine.

Simply put it...it would probably require less effort fixing it rather than getting hypervoxels up to date, sure I would indeed wish that to happen to, but I seriously doubt that will happen soon.

So unlike you, I do see a point with checking a fix to it, I suspect some people still would like to work with dynamiteīs voxel engine, Itīs very capable for what it do, (exluded the fluid engine wich turbulence is replacing good)

Michael

jasonwestmas
10-20-2011, 10:09 AM
Jascha pointed me to Sub-Grid Detail tab for all density detailing. I think I'm up to speed with that aspect. Looks like I can get some pretty noisy looking stuff similar to dynamite. Renders fast, Of course that's after the caching takes place, which can take a while but well worth it I think.

I kinda wish I could use the LW procedurals to control the density but that doesn't appear to be part of the design. You can however use LW procedurals to define color patterns.

erikals
10-20-2011, 10:32 AM
looks good... (bit early to tell maybe...)

prometheus
10-20-2011, 11:37 AM
Yeah I would like to use the sub grid detailing too and try it out, but since the shadow scattering was abondoned or broken in latest version of turbulence I choose too still have the older version installed, I donīt feel I can tweak it with the same feedback without it.

I usally set a very thick smoke ..sometimes up to 500, pull up the red values in the fire shader by 200 and lower the other color values to 50 or so...You could also set a clear smoke above value to get the fire more appearant at the start.

for detailing, you can actually use the simulation turbulence settings and perhaps vorticity to add further detail to the sim that makes quite a difference.

Itīs not really the grain or detail that is a match for turbulence to match dynamite.

It is NO 1..the internal rayleigh boost and light scattering and the cooling in the Dynamite plugin that is a little bit difficult to match in turbulence, I am however convinced that if you use it often or learn it well, You will probably be able to match it, I guess Itīs a matter of setting proper density curves between smoke and fire and also good settings for the white point values within turbulence.

Im not sure, but when we talk about rayleigh illumination model that Is in
fluids called multiplied scattering I believe?

afterburner for max and dynamite for lightwave has it called rayleigh illumination I guess, Iīm Interested to get more knowledge regarding this thou.



It is NO2 ..the fact that you can design the particles, motion life, direction and particle size in a way you really canīt with turbulence.

It is NO3...The fact that You are not in need of calculation in many cases working with particles and dynamite voxels and resolution independent, and renders faster than fluids (I think?)

Image attached...just a quick test, to low resolution 45...upresed a little..but I havent set a good turbulence setting or vorticity wich would probably give better detail...I just more wanted to check basic explosion flow direction and the smoke vs fire density, not as good as the illumination, fire and cooling of the dynamite plugin thou. (wich erikals pointed to in the turbulence thread)

Michael

prometheus
10-20-2011, 12:20 PM
äuhhm.. felt like playing today actually..so Iīm testing this.

changed color to darker smoke color.
Showing difference between 0 turbulence and 50 mm turbulence setting.

container voxel size 45 mm..object emitter size 50 mm (yeah I know what a tiny explosion, I need to learn setting the right scale from beginning:))


So just small settings in the turbulence tab can add some good detailing too.

Michael

jasonwestmas
10-20-2011, 02:11 PM
That last image is starting to be looking pretty good. Thanks for the help.

jasonwestmas
10-20-2011, 04:36 PM
You guys ever get the particle emitter stuff to work with T4D? I followed the quick start but no luck.

http://help.jawset.com/en/lw/quick_start

Never mind, it works =P Just won't show up in OGL for some reason. Probably a button some where. . . .yah had to turn the viewport preview channel to match the fire shader channel type.

prometheus
10-21-2011, 12:38 AM
Why? I know that you can use geometry's vertices to build your volumetrics and stuff, but so can you in TFD, no? Of course you can't animate really weird animations of a cloudball walking down the street, which you can with HVs and Dynamite, then again, how often is that the case? ;)
If you use those particles you designed for TFD, all you get additionally to Dynamite/HVs is the option of a much, much better behavior for the dissipation of the volumetrics. Don't quite get what's so different about "direction, motion life" and so on - you mean the gradients for shading the voxels? Of course you're right then. But I'm fed up with HV's shortcomings (and Dynamite has the same flaw in "non-fluid mode") in regard to fading out - it always looks wrong - either you scale it down, or you blend it away, or whatever (never seen a render with smoke disappearing nicely) but apart from the birth and a nice billowing motion, that's it, and that's only 2/3 of the way, always has been. Better hope you never have to show smoke drifting away in swirls... moving convincingly, too. In regards to the often-wished for inter-particle gradient, this was one of the roots as well... :D

Wrong in most cases as far as rendering is concerned. The moment you have the volumes of close particles intersecting, HV's render time goes up significantly. Dynamite was a little better in that regard, iirc, but calculation wise it will always take longer than rendering a grid, where each voxel can only be occupied by one dataset. Ever noticed that the more particles you put in a volume, the longer it renders, even if the screen space stays constant?
No idea what you mean with resolution independent...
Of course you're absolutely right on the simulation aspect. However TFD's real fluid behaviour is exactly what I have waited for quite some time... it's just wonderful. :)

Of course I would have loved it if there would have been all 3 fluid systems available - Dynamite, Bakudan and TFD. But things are as they are, and I agree with Elmar: it's highly unlikely that LW10.1 "is broken" - Dynamite is. You'd need that developer to fix it. Asking the developers of LW to probably revert to some older code (most likely it was simply something that has been updated and which was necessary for several parts of the software) because of an outdated plugin is a little bit... illogical, sorry. Working on an explosion right now myself, but only in my spare time, which is... little short at the moment. :(


well your sort of right...the additional fluid dynamics can be nice and reallisticly, but You might not always want that but instead more straight forward voxel volume only controlled by particles, particles and fluids requires much more setup and effort to work with, but sure ..you could probably get the most realistic stuff with fluids.

About resolution independent...turbulence requires setting a resolution grid wich affects simulation time and render time I guess, hypervoxels or dynamite voxels doesnīt have to worry about that and the detailing
wich is constant in hypervoxels or dynamite for the render detail.

Not sure but render times..using fluids with proper shadow depth and high detail enough makes render time long too..in my experience on tweaking and rendering dynamite voxels vs fluids...I really believe that voxels are faster...If you get it faster..then either you or I am doing something wrong..you could render the fluids very fast with Itīs initial state without the multiple scattering or single scattering, and much faster rendering ordinary fire without the smoke, but as soon as you work with smoke and the more advanced illumination methods, you get much longer render times.

About dynamite being faster in render than hypervoxels, actually no I donīt think it was faster...but it depends on what you work with and quality settings etc.

of course more particles with voxels gives longer rendertimes..thatīs pretty logical, also think about that using many particles with smaller particles size doesnīt give that long rendertimes compared to a somewhat less particle amount with bigger voxel size, ergo You can get particle trails render quite fast or particle plumes.

Ps...Like to see your work on the explosion when you get there.

http://vimeo.com/user680656/videos

Michael

prometheus
10-21-2011, 12:45 AM
That last image is starting to be looking pretty good. Thanks for the help.

I got it a little better in terms of shading last nigh, Im trying to match the dynamite fire illumination, so when I get back home around 9 hours or so...Ill try and render out animation too.

the tricky part is to get a smooth blend between the temperature fire shader and the thick black smoke, but to still retain the redish scattering beneath the thick black smoke.

Michael

prometheus
10-21-2011, 12:51 AM
About voxel and fluid resolution, while tweaking grain detailing in a volumetric handler, in fluids you either need to set a resoluion and simulate..oh didnīt like it, you need to set another resolution and simulat..oh not quite there..simulate again, even with upresing it isnīt tweaking friendly.

voxels doesnīt work that way, you simple set your scale of the detail or frequencies in the texture tab and render or viper preview it, no need simulate the particle action or anything, and you can always work with the same frame all the time if you want.


So workflow wise I sometime think voxels are easier and faster to work with.
Michael

bazsa73
10-21-2011, 07:06 AM
Hello guys, I just found this thread and share pretty much the same concerns.
I played with blender fluids but somehow blender feels awkward to me so I abandoned it and did not give attention to fires because I would like to see what LW11 offers if at all. So, until that time: Happy tweaking I guess.

prometheus
10-22-2011, 05:26 AM
hereīs a little test again, mostly checking the blend between reddish fire and black smoke, the sim is in need of better detail thou, not sure where I should approach that, either in a hight voxel grid wich currently is only 45...or add more detail through setting hight turbulence wich currently is 50 mm.

I need to improve the beginning too, with hight temp,ignition and not start with a puff ball :)


http://vimeo.com/30915803

Michael

jasonwestmas
10-22-2011, 09:33 AM
yeah, you need a lot more noise. Maybe try lowering the subgrid smallest scale and the small power? You may need to get the latest version, not sure.d

Very cool voxel animation though.

prometheus
10-22-2011, 11:34 AM
yeah, you need a lot more noise. Maybe try lowering the subgrid smallest scale and the small power? You may need to get the latest version, not sure.d

Very cool voxel animation though.

well ..I cant use subgrid since Iīm not using the latest version, and I wonīt install a newer until jascha getīs his latest build finished, I just canīt stand trying to tweak smoke illumination in the current last build since that was removed from viper and vpr.

Have to wait for the next build.

jasonwestmas
10-22-2011, 11:54 AM
well ..I cant use subgrid since Iīm not using the latest version, and I wonīt install a newer until jascha getīs his latest build finished, I just canīt stand trying to tweak smoke illumination in the current last build since that was removed from viper and vpr.

Have to wait for the next build.

Ah so that's why I can't get smoke to look good in VPR. The smoke illumination thing. It always renders flat in vpr.

prometheus
10-22-2011, 12:01 PM
Ah so that's why I can't get smoke to look good in VPR. The smoke illumination thing. It always renders flat in vpr.

It probably is..I really cant figure out why jascha left that out in the last build, only to make it faster? But He mentioned he will fix that to the next build again, at least for VPR, but I hope he fix it back for viper too, I prefer to tweak with viper since Itīs still faster and more stable and direct animation preview is easier to set up.

Michael

jasonwestmas
10-22-2011, 11:23 PM
I thought this looked pretty cool and noisy enough. I'd have to render out an animation to be sure. After messing around a bit I'm finding the controls for the noise and shading to be rather simple so F9s are ok to get started I guess. But yeah I hope that viper previews for the smoke get fixed. The multiple scattering illumination settings are rather important for animation tests.

prometheus
10-24-2011, 02:54 AM
Little test - don't like the result of the simulation, I know what went wrong... but further playing has to wait for at least a week. :(

Of course HVs are awesome, I would not want to be missing them. Dynamite also has a few options that are quite cool and unique, it's really pity that it's gone.
I also prefer those options for other stuff than this. :) Snowflakes, underwater fluff, bacteria, I once used 4 or 5 giant HV spheres without any shading to add a more irregular atmospheric haze to a scene - perfect, fast renders, looks awesome. And it's not that anyone does not like Dynamite, it's just that - sad as it is - it has run its course. :( Whatever people still get out of it, it's running on reserve tank... and then again, lots of people keep old versions of LW around for specific tools.

This is a fun exercise, will try again... :)

Yeah I will try something similar too with turbulence with my basic setup I had for the dynamite voxels on particle tank explosion...(I actually have a collision object sphere over the tank to force particles to the side)
http://vimeo.com/28587261

The difficulties would be in getting it to look like a plume, not just volume trails, and also to get a rolling plume wich might be hard.

This has a little of that rolling explosion on it, but harder to control where the flow goes when Itīs based on object emitter with a texture, some expansion and pressure and normal force from the round sphere in order to have it pressure out at the sides.
http://vimeo.com/30915803

If I only had the thime I would rework this with the tank and do more work with the initial start explosion.

But the dynamite fireshader looks better than the ol hypervoxels I think..this one is hypervoxels and not as good as the dynamite setup.
http://vimeo.com/28074469

and hereīs a little more ordinary hvīs..
http://vimeo.com/27734204

Michael

jasonwestmas
10-24-2011, 08:40 AM
Hypervoxels look way too flat and blobby for this kind of work, I would never use em on anything that required a thick volumetric look based on what I've seen. You can make the voxels tiny and use a lot of them but that shoots your render time through the roof.

erikals
10-24-2011, 09:15 AM
Hypervoxels are good for mud, for the monster in the Green Lantern, or for some other VFX.

limited, but has it's use...

jasonwestmas
10-24-2011, 09:59 AM
Hypervoxels are good for mud, for the monster in the Green Lantern, or for some other VFX.

limited, but has it's use...

good luck getting that many HV into your scene. Of course I suppose it depends on the material you are dealing with.

erikals
10-24-2011, 04:41 PM
...for many arms like in that shot, geometry can be the solution,
for only one arm closeup, Hypervoxels could do the work.

rendering in passes can also do the trick. always cheat.

prometheus
10-25-2011, 01:08 AM
Hypervoxels look way too flat and blobby for this kind of work, I would never use em on anything that required a thick volumetric look based on what I've seen. You can make the voxels tiny and use a lot of them but that shoots your render time through the roof.

It depends on partly the size, the thickness of hvīs, the density of hvīs etc, you can acheive smoother look by setting those lower and correct the appearance of volume within gradients on different channels, and also set a proper hypertexture that cuts enough in to the volume, also important with setting a good contrast that isnīt to high so the hypertexture blends better.

Sometimes faking it with sprites could work..
http://vimeo.com/27101544

This one is indeed to blobby, but I know I could tweak a away it quite a bit more if I had the time setting the right texture etc, post blurring would also help a little.
mushroom top geometry with hypervoxels on vertices, stem hypervoxels sprites.
http://vimeo.com/26391236

Michael

jasonwestmas
10-25-2011, 08:18 AM
http://vimeo.com/27101544,

I actually do like the texture and shading on that one, very good. What exactly is hypertexture? It just looks like a bump map on the surface voxels for gels and liquids. I was hoping it would make the actual size of a single HV more Asymmetrical. Could be different for volume and sprite, I can't remember.

I can see that blobby works ok for bomb clouds, not bad at all.

jasonwestmas
10-25-2011, 08:28 AM
...for many arms like in that shot, geometry can be the solution,
for only one arm closeup, Hypervoxels could do the work.

rendering in passes can also do the trick. always cheat.

Yeah I think what I was doing something similar to the Green Lantern monster wrong the other day. I was trying to use the HVs to create texture based on the HV size and subdivision of the underlying geometry points. That's not a good idea. I had 5 million extremely small HVs that grow off the verts on the model. This took 15-20 minutes to render 1 frame which is ok but too long for what I'm doing. I should be messing with the Hpertexture settings I think and to let the HV's be more soft, with less of them. It's a tough balance though because you have to increase the size of the HV to get good coverage and then your thickness of the material gets to be too much and too blobby.

prometheus
10-25-2011, 08:31 AM
http://vimeo.com/27101544,

I actually do like the texture and shading on that one, very good. What exactly is hypertexture? It just looks like a bump map on the surface voxels for gels and liquids. I was hoping it would make the actual size of a single HV more Asymmetrical. Could be different for volume and sprite, I can't remember.

I can see that blobby works ok for bomb clouds, not bad at all.

image attached shows what hypertexture is.

and I used sprites on that clip sample, It should be faster velocity thou, and it has a little to much cut out in the particle birth and some render clipping artifact from sprite mode.

the bomb cloud could be tweaked a lot better, but I donīt have the time now to do that, thatīs just a sample to show voxels on geometry and rotating it.

Michael

prometheus
10-25-2011, 08:35 AM
hypervoxels are a volume pixel, around a single particle or null or vertice, you donīt control itīs size with terms of asymmetrical shape really, but it give you an asymmetrical shape wich is a different thing really.
volume pixel= vo in the word "volume" and xel in the word "pixel"= voxel

random size in the hv panel is good to use for the hv particle size, to break up uniformity.

Ps..my avatar has tiny tiny hv sprite pixels on geometry on the flames.

Michael

prometheus
10-25-2011, 08:41 AM
Yeah I think what I was doing something similar to the Green Lantern monster wrong the other day. I was trying to use the HVs to create texture based on the HV size and subdivision of the underlying geometry points. That's not a good idea. I had 5 million extremely small HVs that grow off the verts on the model. This took 15-20 minutes to render 1 frame which is ok but too long for what I'm doing. I should be messing with the Hpertexture settings I think and to let the HV's be more soft, with less of them. It's a tough balance though because you have to increase the size of the HV to get good coverage and then your thickness of the material gets to be too much and too blobby.

wouldnt normal adaptive displacement be more efficient?? on tentacle monster arms? instead of voxels?

jasonwestmas
10-25-2011, 08:49 AM
Ps..my avatar has tiny tiny hv sprite pixels on geometry on the flames.

Michael

Yeah I did something similar to that a while ago with a displaced model using the same mdd file. I liked the technique. It was LW sprites. Render time wasn't too bad.

http://vimeo.com/12110295

I noticed with the hyper texture it apply the same exact shape to every HV particle. I tried to use world coordinates but that only turned off the effect. But maybe I'm just not familiar enough with world coordinates.

jasonwestmas
10-25-2011, 08:57 AM
wouldnt normal adaptive displacement be more efficient?? on tentacle monster arms? instead of voxels?

oh yeah sure, but at the time it was a more direct solution to use HV. But I'm more likely going to use displacement procedurals like the sprite example above.