PDA

View Full Version : Can hypervoxels contribute to radiosity illumination?



jeric_synergy
07-27-2015, 02:18 AM
If an HV surface is set to 100% Luminosity, can it add to a scene's illumination?

EG: I have an HV emmitter spitting out persistant particles that form a solid line above the surface of a sphere. Can the HV's thus generated be illuminating the globe underneath?


Thanks.

Every4thPixel
07-27-2015, 02:36 AM
Why don't you just try it before posting a question?

prometheus
07-27-2015, 03:29 AM
yes I think it can, I think you need to check volumetric radiosity in the gloal illumination panel.
If that by any chance it wonīt work...you can always use partigons instead of hv emitters, if you already have made the scene with hv emitters, you can use the particle2partigon command to convert hv emitter to partigons.

Michael

prometheus
07-27-2015, 07:54 AM
yepp...couldnīt test before..but I knew I once tested it, and hypervoxels can illuminate surfaces, but you need to check the volumetric radiosity in the global illumination tab, and raise your luminosity way above 100, also be carefull to not have too small particle sizing, that donīt seem to do much,(unless having a lot of them perhaps, or increase luminosity to extreme values like 10000 or 20000), otherwise make them big enough so you can see it illuminate the surfaces around it or below it or whatever, use VPR for interactive feedback.

Hv surface, hv volume and hv sprites ..all of those works with volumetric radiosity/GI in 11.6.2 at least.

jeric_synergy
07-27-2015, 10:10 AM
Why don't you just try it before posting a question?
BECAUSE I DID. But there's a bazillion options and it's easy to miss one, like "Volumetic Radiosity", thank you prometheus.

Here's a question: if you couldn't contribute, why did you post? Did you have the answer?

jeric_synergy
07-27-2015, 10:17 AM
yes I think it can, I think you need to check volumetric radiosity in the gloal illumination panel.
If that by any chance it wonīt work...you can always use partigons instead of hv emitters, if you already have made the scene with hv emitters, you can use the particle2partigon command to convert hv emitter to partigons.
Michael
Thanks promethheus: it would have been a lonnnnnnnnnnnnng time before I ferreted that switch out.

#aflw, but I can confirm that JUST making HV's 100% Luminous and turning on Radiosity does not do the trick.

I may try partigons, which I've never quite understood, instead of HVs anyway, because I'd like the emitted items to have a different aspect ratio..... although now that I've typed that I think I may be able to flatten out the emitted HV's to form more of a ribbon than a tube, using the Scale controls in the HV panel.

re: number of particles: these are set to create a solid tube of particles, as per the "light trails" thread, hattip to souzou. I'm hoping to get the entire trail to illuminate the underlaying globe.

prometheus
07-27-2015, 10:53 AM
Thanks promethheus: it would have been a lonnnnnnnnnnnnng time before I ferreted that switch out.

#aflw, but I can confirm that JUST making HV's 100% Luminous and turning on Radiosity does not do the trick.

I may try partigons, which I've never quite understood, instead of HVs anyway, because I'd like the emitted items to have a different aspect ratio..... although now that I've typed that I think I may be able to flatten out the emitted HV's to form more of a ribbon than a tube, using the Scale controls in the HV panel.

re: number of particles: these are set to create a solid tube of particles, as per the "light trails" thread, hattip to souzou. I'm hoping to get the entire trail to illuminate the underlaying globe.


Youré welcome.

Iīm in need of some sleep now, so I canīt go further and try your light trail version today anyway.
I did stumble on some experimental ways of using hv and radiosity with particles, to create lightning on a poly with a cloud image, so it gives a very nice flash sequence by just setting the particle birth right and particle size too of course...there are of course many ways to do that, not sure if I will continue on it, but it was interesting.

Michael

Sensei
07-27-2015, 01:20 PM
If an HV surface is set to 100% Luminosity, can it add to a scene's illumination?

100%?? Set it to 10000%....

Always worked without problems.
Because it doesn't matter what is in luminosity.
It's total color that matters.

Try:
Const > Scalar, 100.0 (=10000%),
plug to Diffuse Shading (or any other .. Shading),
of course in Node Editor.

spherical
07-27-2015, 01:37 PM
Why don't you just try it before posting a question?

His post count wouldn't increase to the level it has if he did that every time. :D

spherical
07-27-2015, 01:43 PM
if you couldn't contribute, why did you post?

Let's not go there, OK? Things go snarky and augur in pretty quickly around here enough as it is. Ya gotta agree that some people use this as their personal manual; posting question after question, when the answer can many times be found on one's own with a little digging. How many times has someone cited a LightWave manual page as the contributed answer?

jeric_synergy
07-27-2015, 01:53 PM
I started to get interesting/usable results with a 20,000% (20K) Luminosity setting, after setting the Global Illum/Volumetric Radiosity switch. That's a lot of Luminosity. Render times are not too bad, yet, but it's just a testbed.

+: Renders seem to be fine, but VPR gets flakey a while into the animation: frames render as black, and the location of the illumination seems to be jumping around. But, render is fine. Huh.

++: Wrongo: lots of flickering in the final render, I just didn't happen to hit F9 on the frames that had conspicuous flicker, and of course it's a lot easier to see in an animation.

jeric_synergy
07-27-2015, 02:39 PM
Here ya go, if you want to play with it. Of course, including luminosity is just taking away from the basic trick that souzou created, which is kinda foolish, since it didn't need that. If you turn the lights back on and disable radiosity you can get a nice trail thru the sky (I may have dropped the birthrate below what it should be, tweeking around...).

I'll be delving to find out what can be done about the flickering.

129064

souzou
07-27-2015, 03:14 PM
You could try the instancing approach with luminous geometry? Wouldn't need volumetric radiosity, may still flicker though.

129065

jeric_synergy
07-27-2015, 03:26 PM
I'll have to reread the instancing approach: I suppose I should know, at least theoretically, the roots and causes of flicker, although I find such information incredibly boring and fiddly, and the lack of interactivity (you HAVE to have a sequence to see it) frustrating. Dave Gerrard used to post long articles about such things, and AA, and for me it's a sure cure for insomnia. (But appreciated, since I'd never in a million years figure it out myself.)

Ignoring the radiosity issue, I'd like to develop a more of a 'ribbon' look than 'thin tube'. I'm hoping Scale will do the trick-- you used HV's, right?

jeric_synergy
07-27-2015, 10:36 PM
Now, HERE'S a surprise to me: I was assuming there would be some trouble in getting HV reflections, but instead, I'm getting TOO MANY reflections.

Regardez cette JPG:
129067
The black bits are 100% reflective, the blue zero. I've chosen frame #24 because the reflection of the polygonal ball on the end of the particle stream is straddling a reflective and non-reflective zone. The little bit of the red 'handle' below the white particle stream is a reflection also.

Here's the kicker ( "oh yay" ). There's no trouble making the HVs reflect, in fact, it seems they reflect even where they shouldn't. The lower particle stream is only a reflection. You can see that polys (the ball, the handle) are both correctly reflected and NOT reflected, but the particle stream is reflected in the blue area when it shouldn't be.

Is this analysis correct? Why are the HVs reflecting where they should not?

129068

Danner
07-28-2015, 02:26 AM
something with 20k luminosity will reflect even if your reflection is as low as .001

I don't know if it's possible to have hypervoxels unseen by camera and rays, if that is so, use two sets of HV, one to be seen by the camera and reflecitons and another to light the scene.

prometheus
07-28-2015, 03:51 AM
something with 20k luminosity will reflect even if your reflection is as low as .001

I don't know if it's possible to have hypervoxels unseen by camera and rays, if that is so, use two sets of HV, one to be seen by the camera and reflecitons and another to light the scene.

yes...you make hypervoxels or partigons unseen by camera in the particle emitters render tab, it will still be active to illuminate surfaces if volumetric radiosity is checked.

Michael

jeric_synergy
07-28-2015, 08:51 AM
Thanks Danner, prom: I should never have gone down the radiosity rabbit-hole. It's filled with tears.

prometheus
07-28-2015, 12:57 PM
Thanks Danner, prom: I should never have gone down the radiosity rabbit-hole. It's filled with tears.

You wanted it, all I offered was the truth and you swallowed the red pill :)

JoePoe
07-28-2015, 04:02 PM
Turn reflection down to zero in main layers panel. Seems, oddly, that then you are dealing with just the texture driving the reflection.
And turn down the 20000 lum in HVs as stated.

129072

jeric_synergy
07-28-2015, 04:18 PM
JoePoe: I >always< expect the Texture to wholly define the channel, but it seems that sometimes it does, sometimes it don't. But the texture works correctly with the "handle" -- it must be the 20,000.

The 20,000% was just to get an adequate coverage from my early experiments.

JoePoe
07-28-2015, 04:25 PM
.....But the texture works correctly with the "handle" --

I'm not so sure about that.....
...for the ring neither.

129073

jeric_synergy
07-28-2015, 04:53 PM
Then what about my render in this post?
http://forums.newtek.com/showthread.php?147390-Can-hypervoxels-contribute-to-radiosity-illumination&p=1438208&viewfull=1#post1438208

JoePoe
07-28-2015, 05:19 PM
Good point. Don't know where I got turned around.