PDA

View Full Version : Cockpit Canopy Swirlies



Richard Hebert
03-04-2014, 12:46 PM
Hi everyone!

This goes beyond any tutorial I can find. I'm looking for avenues to achieve this effect in test renders for plexiglass canopies. I've tried using a texture map with luminosity applied but not sure how to have a gradient only effect back facing polys. to recreate the backscatter look. In real life, I realize this is refraction from the 'swirlies' in the plastic but not sure how to go about this... any takers?

Thanks for all the advice,

Richard

spherical
03-04-2014, 05:42 PM
It's forward scattering and dispersion created by polarization and interference in the material. I seem to remember a method to use this principle for situations like these. LW's dispersion is less than great but it may get most of the way there. I'll see if I can dig it out. Odd that we were discussing this in the studio just this morning.

Richard Hebert
03-04-2014, 05:55 PM
Thanks for the quick response! These projects get involving really quickly. It's easy to do a little reflect, refract and transparency but when you start getting into what makes a material look real... wow.

spherical
03-04-2014, 06:02 PM
Heh. I hear that! It's the subtle things that make all the difference... and consume the most time.

XswampyX will probably come up the one of his signature node setups and blow all our minds.

Richard Hebert
03-04-2014, 06:15 PM
Let's encourage him!

XswampyX
03-04-2014, 06:24 PM
Hello! :)

My first thoughts were just to stick a poly behind the plexiglass and have it unseen by camera, you can then adjust it to give whatever effect you want, without ruining the rest of the render.

spherical
03-04-2014, 06:29 PM
Oh, come on! You can do better than that. :)

XswampyX
03-04-2014, 06:31 PM
Not at 1:30 in the morning I can't . :P

I'll sleep on it.

spherical
03-05-2014, 06:14 AM
I still haven't hit the sack. Slacker. :)

djwaterman
03-05-2014, 08:00 AM
Here's a big discussion about dialectric glass/plastic, and this page where Dave Jerrad joins in and hints at all sorts of things related to your question. Scroll down the page to his post and check out the picture of back scattering on a canopy.

http://forums.newtek.com/showthread.php?135678-Dielectric-issues/page6

Tobian
03-05-2014, 08:04 AM
I'd ask Dave Jerrard, he built a shader like this for the plane canopies in that show he did. I'd probably recommend using a combination of backscatter shader and Dielectric, but it's very tricky because you can't mix SSS with transparency properly in LW.

RebelHill
03-05-2014, 01:13 PM
Well this is my 60sec idea... Im using dielectric, but that's by no means necessary, you could be making up your own perspex shader via "bits" if you wanted... The main idea here is simply shoving a bump on the backside of the polygons to create the "scattering" for light coming through, rather than reflecting off.

Cheers.

shrox
01-31-2015, 03:39 PM
Waking this thread up out of status to say:

http://www1.cs.columbia.edu/CAVE/databases/dirty_glass/texture.php

I am dirtying up a canopy too.

Richard Hebert
01-31-2015, 04:59 PM
I've been busy doing other things since this I started this thread... have you seen any examples of backscattering anywhere out there?

spherical
01-31-2015, 11:55 PM
A bit OT of this thread but just thought it may be good to cite that "swirlie" is not the same as "swirly". The latter is free-flowing forms, while the former is stuffing someone's head in a toilet and flushing.

shrox
02-01-2015, 12:10 AM
A bit OT of this thread but just thought it may be good to cite that "swirlie" is not the same as "swirly". The latter is free-flowing forms, while the former is stuffing someone's head in a toilet and flushing.

Oh, there's another definition as well.

spherical
02-01-2015, 12:19 AM
Which is....?

shrox
02-01-2015, 12:27 AM
Which is....?

And you claim to live in California...

spherical
02-01-2015, 12:39 AM
I don't claim. I do. Have the tax bills to prove it. What's your point? And, I'm not what would be termed as clueless and/or exclusionary.

shrox
02-01-2015, 12:51 AM
I don't claim. I do. Have the tax bills to prove it. What's your point? And, I'm not what would be termed as clueless and/or exclusionary.

Well, this is family forum. It involves using one's tongue, that's all I can say.

Richard Hebert
02-01-2015, 08:09 AM
Any ideas on how to create them? I mean the family friendly non-toilet version!

- - - Updated - - -

And with that we're back on topic!

djwaterman
02-01-2015, 08:24 AM
I'm glad I deleted my post because I went way off topic. Seriously, did anything from the Dave Gerrard discussion help, I'm pretty sure he even had a render of canopies back-lit with the sun.

shrox
02-01-2015, 12:19 PM
Look at this, it's from what I posted earlier.

http://www1.cs.columbia.edu/CAVE/databases/dirty_glass/code.php

MG artist
02-02-2015, 11:07 PM
Just refraction blur on dielectric. The fist image has exactly the same material without bluring. I tried a few networks with kappa, sigma and sigma 2 but I didn't like the result. They are good if you need visible dust though. 126842126841126840126839

spherical
02-02-2015, 11:49 PM
Not the same thing. What I believe the OP is asking about (rainbow colors, usually stronger when seen though a polarizing filter) is seen in this shot:
126843
Most easily visible in the aft fixed canopy. Yes, that's me.

MG artist
02-03-2015, 05:59 AM
He's asking for back scattering effect


I've tried using a texture map with luminosity applied but not sure how to have a gradient only effect back facing polys. to recreate the backscatter look.

For rainbow colors, dispersion would be adequate. Or a gradient with rainbow colors having as an input a procedural, plugged to the color input of a reflections node with tint reflections enabled connected to reflection shading. I'll give subsurface scattering another shot and post any results.

EDIT: In the thread dj waterman posted there is exactly the same backscattering effect he is asking for made with material mixer from Mike Wolf, I'll try that.

MG artist
02-03-2015, 09:58 AM
Kappa and dielectric this time. You need db&w tools to load the surface. The surface is changing with angle of view, from the other side this effect isn't visible, the way it should. 126868126867126869

spherical
02-03-2015, 02:55 PM
Backscattering doesn't exactly translate to "swirlies". Overnight, I was thinking that he might be meaning the circular micro-scratches that inevitably appear on the polycarbonate until they're polished out. Those would add to what actually would be forward scattering.

JoePoe
02-03-2015, 03:21 PM
Why not mix in a bit of Soap Bubble Texture (http://3dxyz.pro/?s=soap). :hey:

http://3dxyz.pro/wp-content/uploads/2014/10/[email protected]

shrox
02-03-2015, 03:47 PM
Backscattering doesn't exactly translate to "swirlies". Overnight, I was thinking that he might be meaning the circular micro-scratches that inevitably appear on the polycarbonate until they're polished out. Those would add to what actually would be forward scattering.

I was thinking smudges and smeared fingerprints. Tiny scratches from polishing sound more like it too.

JoePoe
02-03-2015, 04:41 PM
Fingerprints and micro-scratches :stumped: :D


I'm just having some fun with a rainbow and wood pattern.
.... just a little something I like to call.... a "swurlee". ;D

126883

126884

MonroePoteet
02-03-2015, 05:56 PM
Backscattering doesn't exactly translate to "swirlies". Overnight, I was thinking that he might be meaning the circular micro-scratches that inevitably appear on the polycarbonate until they're polished out. Those would add to what actually would be forward scattering.


I was thinking smudges and smeared fingerprints. Tiny scratches from polishing sound more like it too.

It could also be the natural polarization that polycarbonate and other plastics acquire during manufacturing interacting with a polarized light source. The long polymer chains in the plastic are bundled with more or less consistency, producing a polarizing material. With a polarized light source, and two layers of polarizing material (i.e. looking through it from the side), you might get an interference pattern similar to those described here:


http://www.glitterins.com/introduction-to-photoelasticity.html

I'm sure cockpit windscreens are manufactured to very high quality requirements, perhaps even including the alignment of the polymer chains, but there still might be a type of moire pattern as a result. Just a thought.

mTp

Pensart
02-03-2015, 06:43 PM
Nice result JP

spherical
02-03-2015, 07:44 PM
It could also be the natural polarization that polycarbonate and other plastics acquire during manufacturing interacting with a polarized light source. The long polymer chains in the plastic are bundled with more or less consistency, producing a polarizing material.

Yes. That's what I indicated earlier and showed in the shot of the F-16.

JoePoe
02-04-2015, 09:42 AM
Nice result JP


Thanks :)

one more....

http://forums.newtek.com/attachment.php?attachmentid=126890&d=1423068097

MG artist
02-07-2015, 09:42 AM
Something I built in my free time:
126942126943126944
Feel free to use the normal map wherever you want. You need db&w tools to load the surface.