PDA

View Full Version : Hypervoxels, blending like in 5.6



erikals
03-25-2009, 06:53 PM
in Lightwave 5.6, Hypervoxels could blend like this,...
http://www.box.net/shared/static/x4ymx09o7r.mov

much smoother, almost like fluids, can we have that again?

http://www.newtek.com/forums/showpost.php?p=788235&postcount=70

Sensei
03-26-2009, 01:29 AM
I don't think so blending mode was changed since v5.6 ages. Blending particles is only done in Surface mode. In Volume mode there is no blending (option is even disabled). To prove it has changed you can't give just links to final renders/anims. You need scene & objects and render it first in v5.6, and then in v9.6, and compare results (I often put renders to Photoshop as layers and turn them on/off or use Subtract blending modes to compare), with simple stuff like 100 particles, without additional textures etc (reduce things that could influence your comparing process).. Otherwise it's just words.. Like in Mr_Rid thread..

I think so results from these videos are so nice because of smart usage of HyperTexture. That's essential for realistic clouds!

HyperVoxels pick up first procedural texture automatically (in alphabetical order)! So after installing f.e. Denis nodes (RenderMan Node Collection) you will end up using completely different default HyperTexture than with fresh LW! (Cells- worser quality and 10x much slower). That's not dpont's fault, but the way LW picks up HyperTexture. I suggest always going to HyperTexture page and pick up manually Turbulence/Crumple each time Volume mode is enabled. I found it accidentally while playing with TrueHair Previewer and volumetrics and wondering "why is it so slow".. it turned out that LW was picking up RenderMan's Cells..

dandeentremont
03-26-2009, 03:34 PM
Man, I would kill to have volume blend like it used to... :(

wildr3d
03-26-2009, 04:41 PM
This screen grab I found shows a metaball blending function in the old 5.6 hypervoxels panel. I remember and have often complained with Mr.Rid about this feature missing in 9.0.

Looking at this screen really makes me miss the way Hypervoxels used to work.

ivanze
03-26-2009, 06:39 PM
If Hypervoxels is a plugin, is there a way to try and see if HV 2.0 works with 9.6? :)

erikals
03-27-2009, 08:30 AM
i'd love to try, but my, 5.6 verion is at my parents house,..
could anyone send the 5.6 plugin this way?
my email is erikals [atention] start [dut] no

wildr3d
03-27-2009, 10:40 AM
If Hypervoxels is a plugin, is there a way to try and see if HV 2.0 works with 9.6? :)

I doubt they still work, I have the legacy plugins installed and Steamer can be added but does not activate. The old Hypervoxels don't show up at all.

erikals
03-27-2009, 11:02 AM
well, i have a Dynamite idea,... let's hope it is,.. Dynamite(!) ehhh... :o

prometheus
03-27-2009, 11:06 AM
I canīt for the life of me figure out why they dropped it?
and why not reimplemented.

I believe the Dynamite plugin has that blending option with Metaballs, not sure thou if it works in the same way as the old hv:s..on the other hand Dynamite doesīnt work properly either does it..and looks very dead.

Wouldnīt it be nice if Newtek actually could be a little flexible about introducing new features in to the existing 9.6- cycle,(in this case an old feature) instead of dropping it and only bugfix and all focus on core.

I believe "tiny" things like this could mean a lot to people since core isnīt production prime time yet for a long time.

Im going to investigate how Houdini is doing this with metaballs, thou Iīd prefere to set things up in Hypervoxels, still much to learn with houdini.

Michael.

prometheus
03-27-2009, 11:12 AM
HyperVoxels pick up first procedural texture automatically (in alphabetical order)! So after installing f.e. Denis nodes (RenderMan Node Collection) you will end up using completely different default HyperTexture than with fresh LW! (Cells- worser quality and 10x much slower). That's not dpont's fault, but the way LW picks up HyperTexture. I suggest always going to HyperTexture page and pick up manually Turbulence/Crumple each time Volume mode is enabled. I found it accidentally while playing with TrueHair Previewer and volumetrics and wondering "why is it so slow".. it turned out that LW was picking up RenderMan's Cells..

And yeah that one annoys me to..the cell procedural slows things up and is annoying having to switch all the time, would be nice with some kind of fix..or update to that denis p texture so it wonīt change the standard turbulence unless selected.

Michael

erikals
03-27-2009, 11:20 AM
well, i don't like the way particles render in Dyna,
so the idea is to try to use dynamics to emitt from actual geometry,..

i should test it right now actually....

prometheus
03-27-2009, 11:51 AM
well, i don't like the way particles render in Dyna,
so the idea is to try to use dynamics to emitt from actual geometry,..

i should test it right now actually....

I have no Idea what your thinking of using dynamics to emitt from geometry to get a blend of hypervoxels? but go ahead make a scene:)

and Yeah the render of dynamites voxels has errors in the old versions and
the latest doesnīt work properly with 32 bit systems.

apart from that there is a lot in hypervoxels shading system that canīt be done in dynamite so..sure a proper meta blending system for volumetric mode and a similar fire shader but with standard hypervoxels options and hypertexture effects would be nice, I would like to see even some more hypertexture effects and abilities to mix them and control them over time so that you could freeze the hypertexture effect or slow it down by particle age and time.

Heck why not be able to blend that so the billowing or turbulence effects gradually mixes and turns over to the dissolve expand hypertexture effect over time or particle age.

one more ..introduce the implicit surface volumetrics mode, that is apply hypervoxels to the whole volume of an object, not only points. procedurals or noise functions are then applied throu out the volume with an option to use volume offset so that noise/procedurals could be more detailed and stronger either at the center of the object or at the outside leaving a smoother inside volume, I can imaging a whole lot of different offset nodes for thickness,procedural detail etc fully controllable via nodes.

Michael

wildr3d
03-27-2009, 11:53 AM
And yeah that one annoys me to..the cell procedural slows things up and is annoying having to switch all the time, would be nice with some kind of fix..or update to that denis p texture so it wonīt change the standard turbulence unless selected.

Michael

I think Hypervoxels should just remember the last texture selected.

prometheus
03-27-2009, 11:56 AM
Ohh..forgot..
wasnīt there also a load gradient in the old hypervoxels..
I sure would have some of those to:)
sometime you really like the hypertexture setup but you would like to swith colour schemes on the fly.

Donīt know if theres a true color spectrum index over different gases, would be cool to load different schemes like argon gas,xenon etc.
better yet different material shaders, that corresponds to temperature etc.

Michael

erikals
03-27-2009, 12:39 PM
actually, the Dyna test is kinda interesting,
calculating now...
probably done in urm... 40 minutes or so, on this oldie PC....

wildr3d
03-27-2009, 01:15 PM
To bad about Dynamite, the metaball function in the voxel shader was the closest thing to the 5.6 voxels. It crashes instantly when I try to activate it. Fluid sims still work though.

erikals
03-27-2009, 01:17 PM
check that no SubD objects are in the scene, or try to reinstall the plugin...
i had some weired stuff happening, could be one of those...

erikals
03-27-2009, 01:23 PM
ok, here's a test,

30 min calculation on a Duo 2.3,.. so kinda slow PC,...
this was testing using several small poly, just to see,

the setings here are far from optimal, almost haven't touched Dyna, so i assume much better can be done,...

first video, the cloth object polygons (1 object) that were used to simulate the fluids
second video, the fluids

(p.s, as of the polygons flying through the air at high speed and the settings used in Dyna there is some blockiness, i'm gonna test if it can be removed... plus the polygons are flat, so sideways movement causes changes)

Sensei
03-27-2009, 02:24 PM
or update to that denis p texture so it wonīt change the standard turbulence unless selected.


That's not possible in dpont's node. He can only make Cell node not being recognized as HyperTexture by not providing flag LWTEXF_HV_VOL.. But other texture will be catched in alphabetical order, unless he will use this flag in all his textures..

prometheus
03-27-2009, 02:42 PM
Erikals!
Ahha oki ..I get it..your using dynamites fluids combined with cloth fx dynamics..I thought your were to do it with hypervoxels.

I think I tried something similar some time ago..but as you sad..had problems
with blockiness in the resolution.
I donīt have the full version either to get decent renders from it.

by the way..in dynamite fluids...have you tried using procedural textures in the x,y,z wind direction channels?

that could yield some interesting explosive effects, but you have to tweak it right and start to set the texture value or layer value very high so you can see that it is taking effect.
the textures scale is important to.

if you are doing a fire flame rising up, you could ofcourse set a direction wind upwards, but you can still use textures on those channels if you set blending mode to additive..that can give some additional direction variation, not using additive layer blending mode will kill/override the other basic wind settings.

Michael

erikals
03-27-2009, 02:59 PM
procedural textures in the x,y,z wind direction?
nope : ) i haven't tried that,... : ) that's an interesting idea though,..

there might be good to use a combination of several methods, what I'm afraid of in the end though is the blockiness, gotta fix that somehow...

i'm going to simulate some more on the "offline" Quad Core 2.8, see how things turn out
(damn Motorola modem driver...)

btw, you wouldn't know of any method on how to attach spheres to a particle emitter without using that Linker thing? what's it called again,.. FXlinker,...
or freeze the FXlinked objects motions into one object and one object.
(so it can be used for Dyna)
maybe MDD it somehow...?

prometheus
03-27-2009, 03:55 PM
procedural textures in the x,y,z wind direction?
nope : ) i haven't tried that,... : ) that's an interesting idea though,..

btw, you wouldn't know of any method on how to attach spheres to a particle emitter without using that Linker thing? what's it called again,.. FXlinker,...
or freeze the FXlinked objects motions into one object and one object.
(so it can be used for Dyna)
maybe MDD it somehow...?

Exactly ..really, try that out with procedurals on the wind direction..ive even
got some cool tornado swirl by using cyclone or coriolis procedurals..it was a little tricky to find out using the correct axis for the texture and such. and at the end I couldnīt think of a decent way to do the typical swaying of the tornado..but the swirl looked cool.

There is no other way to redirect the fluids direction other that using other objects as wind effectors but that has not the same control as
lightwaves native wind dynamics and those do not work with dynamite fluids..Would be nice if they worked as in maya or houdini.

so the other way to redirect the wind is to use the texture channels on the wind settings, I would recommend almost turning of the Dynamites swirl settings so it would be easier to see how the texture effects the wind.

and for attaching spheres on particles..Nope..fx linker is the only way I know of..not sure if mdd would work do get it as one object in the end.

I was also thinking of metaballs in modeler, not sure if it would work with dynamite, but you could try making some points and convert to metaballs..
do a morph on the points and use morphmixer in layout, you would need to set the geometry level to after morph to get layout to recognize the meta morph..
I think there will be a problem if you will try and use it for dynamite fluids thou..not sure but I donīt think dynamite recognizes the object as polygons to emitt fluids from.

Ohh..my computers been knocked out for almost two weeks now..but
I just plugged it in and are doing some testing with it to check the health.and my brothers been checking it out for me..maybe Ill be up and running soon.

canīt really do much with a 866 processor and and old geforce card compared to the other pentium 3 with quadra fx 1100..

Actually that quadra fx 1100 is getting old to.

Michael

erikals
03-27-2009, 05:11 PM
heh... :) don't say that, i just "upgraded" from a 7300LE to a 8400GS.... :o
i'd look into upgrading the cpu if i were you,... Pentium 3,.. damn...

thanks for the advice, i've been really hoping to make this work,
didn't get the time for more tests, will get on it soon i hope...
need to have a word with Mr. Blocky. [ugh]

prometheus
03-27-2009, 05:32 PM
Oops..well ..I stated wrong specs..itīs a pentium 4.
3 ghz..2 gig ram and quadra fx 1100 graphics card.

well yeah I need to get me a 64 bit system, and probably a new nvidia card, but that will take me some time, im only working extra from time to time at my old job, and I actually need a wacom tablet right now so that comes first.

And Im forgetting about core for the moment, itīs not for me at current state..could have been interesting to help develop and make my points heard..hopefully someone with a core membership can speak up about a new
hypervoxels system with distance between particles options, metaball blending and some of the nifty stuff mentioned perhaps.

I just have a feeling of that it will not be implemented for a long time
yet, either for the end of the 9 cycle or for a new core...and man the request for a new hypervoxels system has been going on for a long time now, so if it would take a year or two to do something of this I would rather spend time learning houdini..(takes time too)..wich also seems to gain market in sweden too.

Michael

Mr Rid
03-27-2009, 05:50 PM
in Lightwave 5.6, Hypervoxels could blend like this,...
http://www.box.net/shared/static/x4ymx09o7r.mov

much smoother, almost like fluids, can we have that again?

http://www.newtek.com/forums/showpost.php?p=788235&postcount=70

Part of what makes that cloud chamber-like example work is the hyper texture was set to world. But yeah, LW 5.6 HV (HV v2) at least had 2 levels of blend, and could also shrink in size as particles spread apart, or grow as particles came together. That simple aspect went a long way to making more convincingly organic renders.

5.6 also had control to make HVs matte black, or invisible in the alpha for holdout passes, like what I am desperately needing right now. It can be difficult to fake in a black-n-white color pass and means you have to waste time setting up another scene and compositing another element. But cloud-like HVs just plain looked better in 5.6 and it became more difficult to get away from the puffball thing after LW6. I dont know what happened. The HV interface improved a great deal, but some very useful aspects disappeared. And NT has never listened.:bangwall: They didnt need to overhaul all of HVs, just associate a gradient with the 'distance between particle' data that is already there, in the blend and particle size, and get the the 'DBP' grads working correctly in the rest of the channels. How hard can this be?

This is where someone plays the worn out record, 'they are rewriting the code and the next version is going to be heaven.'

erikals
03-27-2009, 06:11 PM
if you're looking at buying a wacom, i'm actually quite pleased with my bambi (whoops, the "i" i stoo close to the "o" on the keyboard here)... Bamboo Fun is the name.
surprised at the good resolution it has, that is A5 size though, not A4... but i like A5 better, less dragging, faster,
i also tend to draw at small scales, like some other concept artists, it varies though what one likes.

nah, no new HV's any time soon i think, tested Fluids in FumeFX and Maya, Maya is ok, but FumeFX is 5-10 times faster.
might wanna look into that.
tricky to find a great app these days, all of them have big drawbacks.
LW is actually quite allright, if only we could have ok dynamics and fluids.
Jay3D made some dynamics plugin tests, so hope to see that, that was posted in the 9.5 beta forum forum though...

you have no idea how much i've posted misc ppl and developers to try to get fluids in LW ;) but hey, one day, it must arrive... one day...

don't think i'll be touching Houdini any time soon,... can't see it is worth the cash, not unless it is used as an additional package.
(though i assume it usually is)

the beta forum is at an early stage, hecticall, (so that's good in a way, NT workng hard)
if you want to though, post me a PM some time in mid Q2 and i will post a feature request for you over there.

erikals
03-27-2009, 06:23 PM
@ Mr Rid

hehe, weell,... ;)
that's how it is, and how it has been for a loong time now,...

question, can't recall i saw any posts from you testing Dynamite, did you test it at one point?

Mr Rid
03-28-2009, 01:10 AM
I plan to evaluate FumeFX and Vray in coming weeks.


@ Mr Rid

hehe, weell,... ;)
that's how it is, and how it has been for a loong time now,...

question, can't recall i saw any posts from you testing Dynamite, did you test it at one point?

Never touched it. I saw that it had issues early on and was essentially unsupported.

erikals
03-28-2009, 06:09 AM
FumeFX,... hmm...might have some good links for you then,... ;)

Allan McKay,..
http://www.turbosquid.com/FullPrevie...cfm/ID/384802/
http://www.allanmckay.com/FFXDVD2.html

prometheus
03-28-2009, 10:37 AM
I wrote a mail several months ago to sitnisati about the possibilities for develop fume fx for Lightwave, he had no current plans for it, that was before core thou....

Maybe Ill send him a mail again and ask more specific if the new code for core could make it more interesting, or if it is something else behind the reason not to develop it for Lightwave.

Michael

erikals
03-28-2009, 11:00 AM
hm, interesting,...

maybe we could send one mail, with names or/and e-mail,... from users who want to see it in Lightwave,...

[edit] i guess it might be better to wait untill Core is released, Q4....

prometheus
03-28-2009, 12:15 PM
I think I know to little about the core to make statements perhaps, Itīs just a gut feeling somewhere that we wonīt see any type of fluids for a long time in Lightwave core, I might be wrong.

donīt see any harm in sending request thou, I could start a request thread and if/when thereīs a lot of people requesting it on that thread, I could send a mail to sitnisati with the link to the thread.

Michael

erikals
03-28-2009, 12:24 PM
yes, hm, maybe a pool (sorry,.. poll) would be in order,...

erikals
03-28-2009, 12:51 PM
btw, this is a somewhat new manual o Dynamite 1.2,...
says it renderes metaballs, didn't know that...

http://www.cantarcan.com/v12/files/dynamitedoc_1b2.pdf

did you ever experiment with regular .jpg pictures in the wind options?
hmm... it does something, i just can't make it work correct for some reason...

prometheus
03-28-2009, 01:35 PM
btw, this is a somewhat new manual o Dynamite 1.2,...
says it renderes metaballs, didn't know that...

http://www.cantarcan.com/v12/files/dynamitedoc_1b2.pdf

did you ever experiment with regular .jpg pictures in the wind options?
hmm... it does something, i just can't make it work correct for some reason...

Yepp it renderes metaballs in volumetric mode or polyons..the voxel size adapts to the size of created poly..I did some interesting test by some simple flat polygons facing the cam.
( a rectangle divided and deleted top corners and bottom corners so you get a cross)

.and I then did some morphs, size and slightly movement to expand the geometry..so when using morph mixer you get and expanding geometry
and dynamites voxels follows that pretty nicely and blends nice to..the problem with the older version is that it didnīt render properly, some fire bug.
this teqnique is similar to the 24 nuke using geometry instead of particles and dynamites voxel engine on the mushroom top.

check the firebug image here..
http://www.cantarcan.com/v12/download.html


And no..donīt think I tried images on wind direction..could be interesting with animation clips aswell if it works.

I did some Image test or procedural? in the density channel thou..on a simple plane..trying to get a cloud plane similar to Mayas fluid cloudmodels
using noise..didnīt work out that good..to blocky.

Havent got my comp up and running yet...have to switch harddrive first, but my brother couldnīt fix the comp, it still just freezes sometimes..once or twice a day:( and itīs obviously a hardware issue since we tested different drives and system.

Have to live with that for sometime I guess..donīt know what to do..new graphics card,new processor/motherboard or what?
I could always do the spring time cleaning..but it was still snowing yesterday:)


Michael

erikals
03-28-2009, 02:03 PM
hehe, was snowing here too :)

i mailed Cantarcan about images in x,y,z... they seemed to be locked to being linked
so no independend x,y,z, channels...

do you still get blockiness in high resolution?
i haven't got it confermed, but it seems to help, if it removes it however, i'm not sure.

for the PC, try switching the graphics card temporarily with another one...
then it is only the CPU/motherboard left...
heh, maybe you have to get a Quad core after all... :)

prometheus
03-28-2009, 02:47 PM
I doubt that you will get any answers from cantarcan thou.

I was trying all kind of resolutions but I couldnīt really get the blockyness away..but it was also kind of hard to see with demo versions..with noise on the renders ..

yeah thanks..I will try that if I can get my hands on some other graphics card..my brother did a test with another card, but that card wasnīt ok from the beginning either so he really couldnīt test on that..and I donīt wanna mess with this card on the old computer right now..have a lot of jobs to apply for.

But thanks for the advice...Ill see if I can dig a card up before buying new stuff.

Back to hypervoxels topic..do you know if thereīs been any information or discussion on the hypervoxels for the new core?

Michael

erikals
03-28-2009, 03:02 PM
strange you should mention blockiness, check this test...
as you can see upping the reso helps quite a lot...
this is usign the exact same settings, all i did was render 200% instead of 50%

as for HVs and Core, nope, i can say this without breaking any 'code'...
the first ride is as stated by NT, focusng on modeling tools, thought as they also said, as modeling tools and anim features are connected in Core as of stacks and stuff, there are some anim features as well.

as for SFX stuff, i wouldn't count on it, but as I said earlier in the thread, I'm hoping to see Jay3D's dynamics plugin sometime... looks allright.

erikals
03-28-2009, 03:21 PM
just a miniupdate on blockiness,...
...so there seems to be 2 kinds of blockiness types in Dynamite,

-too low Dynamite resolution
-too low Camera resolution

can't confirm this 100% yet, but at the moment it looks that way...

erikals
03-28-2009, 05:28 PM
shoot,... agh!!
increasing the resolution doesn't help... that's strange...

this is testing an object with gravity though, if you use cantarcan's smoke preset and most other presets the smoke builds upwards.

that's the only solution to get rid of the reso. blockiness that i can think of...

Dynamites blend mode doesn't give good results,

here's the test, looks more blocky than before, even if resolution of this specific test was over twice as high,... back to the drawing board...

erikals
03-28-2009, 09:04 PM
No,.. Dynamite doesn't cut it I'm afraid,... jumping over to Maya fluids or FumeFX...

prometheus
03-29-2009, 09:23 AM
No,.. Dynamite doesn't cut it I'm afraid,... jumping over to Maya fluids or FumeFX...


I believe some people comes to that conclusion at the end of the day, and then therés all the support unsecurity and update unsecurity.

A little bit sad thou, the basics are good and the fire shader both in voxels and fluid mode together with viper made it actually much faster and easier to tweak than what ivé seen with Mayas fluid texture and ipr render..

However Mayas opengl and calcultation and overall fluid properties is way beyond what dynamite has, and top of that..its damn simple to setup by simply choosing kind of fluid container adding emitters and controlling by other winds etc.

In dynamite setting it all up with nulls first and geometry and no control interaction with winds makes it kinda dull.

Ive also been trying out the houdini fluids, and when talking about the smoke and fire solvers they feel kind of slow compared to Mayas, even in 2d mode, but probably more power under the hood.

To sad I canīt continue trying out Houdini for a while now..have to recover my harddrive where I had all my 3d stuff, a system error accured when the computer freezed the last time and I cant start the system with that hard drive..so I got some recovery tools and have to work with that for a while and transfer the data to a new drive if I can.

Well I heard that Mr. Rid Is supposed to Evaluate fume fx anytime soon?
Would be nice If you could share some of your thoughts about it later on Mr.Rid

I remember I got me some basic course introduction in 3d max in 1999
at school that gave summer courses..a couple of weeks only And I got a simple certificate of the course, but it was not really recognized as more than a simple course.

Then of course I never fell in love with 3d max as I did with Lightwave...not very keen on pickin it up again..but what can you do..when thereīs no other good solutions for vfx stuff like this.

Michael

prometheus
03-29-2009, 09:29 AM
No,.. Dynamite doesn't cut it I'm afraid,... jumping over to Maya fluids or FumeFX...


I believe some people comes to that conclusion at the end of the day, and then therés all the support unsecurity and update unsecurity.

A little bit sad thou, the basics are good and the fire shader both in voxels and fluid mode together with viper made it actually much faster and easier to tweak than what ivé seen with Mayas fluid texture and ipr render..

However Mayas opengl and calcultation and overall fluid properties is way beyond what dynamite has, and top of that..its damn simple to setup by simply choosing kind of fluid container adding emitters and controlling by other winds etc.

In dynamite setting it all up with nulls first and geometry and no control interaction with winds makes it kinda dull.

Ive also been trying out the houdini fluids, and when talking about the smoke and fire solvers they feel kind of slow compared to Mayas, even in 2d mode, but probably more power under the hood.

To sad I canīt continue trying out Houdini for a while now..have to recover my harddrive where I had all my 3d stuff, a system error accured when the computer freezed the last time and I cant start the system with that hard drive..so I got some recovery tools and have to work with that for a while and transfer the data to a new drive if I can.

Well I heard that Mr. Rid Is supposed to Evaluate fume fx anytime soon?
Would be nice If you could share some of your thoughts about it later on Mr.Rid

I remember I got me some basic course introduction in 3d max in 1999
at school that gave sommer courses..a couple of weeks only And I got a simple certificate of the course, but it was not really recognized as more than a simple course.

Then of course I never fell in love with 3d max as I did with Lightwave...not very keen on pickin it up again..but what can you do..when thereīs no other good solutions for vfx stuff like this.

Michael

prometheus
03-29-2009, 01:16 PM
with regards to the smoke and fire fluids mentioned here with dynamite, fume fx and maya fluids..
hereīs some interesting stuff from the knowing film..havenīt seen anything thou.

http://vfxworld.com/?atype=articles&id=3941&page=2

interesting stuff about why they choosed digital firefluids with maya instead
of practical elements..and some interesting info about required high resolution for it.

Michael

erikals
03-30-2009, 08:26 AM
well,... you might find this interesting,
but the release date?...
http://www.spinquad.com/forums/showthread.php?p=246223#post246223

prometheus
03-30-2009, 10:09 AM
Oto_taca? I knew Tim was up to something..but I have missed that one..Nice
It gives hope for the core, still I wonīt jump in the core program right now.

interesting thou..maybe he could have two versions..for different machine specs.

Michael

http://www.vimeo.com/videos/search:michael+ivarsson/sort:newest/format:thumbnail

silviotoledo
04-17-2009, 08:37 AM
this absent blending mode is what I hate about Hypervoxels.
Anyway DYNAMITE have cool blending and it also uses hypervoxels.

prometheus
04-17-2009, 12:51 PM
this absent blending mode is what I hate about Hypervoxels.
Anyway DYNAMITE have cool blending and it also uses hypervoxels.

Well Dynamite doesnīt use hypervoxels, it has itīs own voxel system but not to be confused with or named as hypervoxels.

and yepp it has cool blending, but I have no hopeīs for seeing Any updates for Dynamite at all, if cantarcan proves me wrong..just fine.

And we can also forget about updates to the native hypevoxels since no new features are supposed to be implemented on 9.6 series or up.

And for core, well I just have a hunch we wonīt se any hypervoxels stuff in there until maybe one year ahead, if Newtek proves me wrong..just Great.

Common ..couldnīt hypervoxels be improved with the old code, I have a hard time believing that it couldnīt..statements from newtek saying they really pushed lw 9.6 as far as they could ..donīt sound creditable to me, just sounds like excuses to carry on with the core.

you can tell there was some old good stuff in the earlier hypervoxels versions, and of course as we can see with Dynamite there sure are things that could be done and improved in hypervoxels.

Michael

erikals
04-17-2009, 01:06 PM
...Common ..couldnīt hypervoxels be improved with the old code, I have a hard time believing that it couldnīt..statements from newtek saying they really pushed lw 9.6 as far as they could ..donīt sound creditable to me, just sounds like excuses to carry on with the core.
Michael

absolutely, actually, it was done once...
http://www.spinquad.com/forums/showthread.php?t=8802

erikals
04-17-2009, 01:30 PM
Quicktime examples,...

you can see however, that "RV_smoke.mov" deals with the same blockiness problem.
"nicesmoke.mov" has somehow solved the problem...

adamredwoods
04-20-2009, 11:19 AM
Did you create that smoke? Is that LW hypervoxel smoke?

erikals
04-20-2009, 11:44 AM
absolutely, actually, it was done once...
http://www.spinquad.com/forums/showthread.php?t=8802

nope, it is an abandoned in-house Fluids generator that was integrated into LW. i actually mailed (pm) the guy once again this week to see if there was any chance he could change his mind and sell it. so far no luck, didn't hear back from him...

Jockomo
04-21-2009, 02:30 PM
HyperVoxels pick up first procedural texture automatically (in alphabetical order)! So after installing f.e. Denis nodes (RenderMan Node Collection) you will end up using completely different default HyperTexture than with fresh LW! (Cells- worser quality and 10x much slower). That's not dpont's fault, but the way LW picks up HyperTexture. I suggest always going to HyperTexture page and pick up manually Turbulence/Crumple each time Volume mode is enabled. I found it accidentally while playing with TrueHair Previewer and volumetrics and wondering "why is it so slow".. it turned out that LW was picking up RenderMan's Cells..

Thank you for pointing this out!

silviotoledo
04-21-2009, 06:44 PM
RV Fluids are great!

Whys isn't it in commercial use?