dispersion bsdf

[anamous]

I just tested the scene. Dispersion doesn't show up in the caustics, and it seems that the photon behavior has been changed yet again (9.5.260) - shaders that output only Cf aren't considered worthy of photon generation anymore, and you get the "not enough photons.." error message, which currently means that we can't use the raytracing version of the shader to color the caustics. I'll ask SESI about this.

cheers,

Abdelkareem

[jason_slab]

o man...

thanks for checking it out anamous, let me know what SESI's rely is if you don't mind.

thx

jason

[somesanctus]

https://vimeo.com/95755327

[jumper]

Hey Alex,

Great stuff  - have you got a scene file you could share?

Thanks!

[somesanctus]

ok

dispersion.hip

dispersion.otl

dispersion1.otl

[jumper]

Hey Alex,

Thank you very much!

Stephen

[tfreitag]

..some test of mareks amazing shader with alex hipfiles. big thanks to alex and marek. I took alex prism and tessalate a lot with tiny little curlnoise.

Again, I'm loving this shader!

 

Edited January 27, 2015 by tfreitag

[sebkaine]

Gorgeous threads guys ! You Bring us some Maxwell Rendering inside Mantra ...

 

Tom would it be possible to share the triangle .hip test scene with your exact setup ? I'm really curious about this nice dispersion that actually look REAL and not a dirty tricks like this one :

http://www.psyop.tv/psyop-releases-jf-nested-dielectric-open-sourced-shader/

Edited January 28, 2015 by sebkaine

[tfreitag]

Here is the hipfile. For getting all the nice details I pushed the photoncount to 50.000.000. The photoncaching takes around 10-20min.

dispersion_prism.rar

[sebkaine]

Thanks a lot Tom !

[sebkaine]

I would have one question about dispersion.

 

in many explanation i find on the web it say that dispersion is cause by a modification of the ior according to the wavelength of the incident ray.

but they only talk about refraction , only the refracted ray looks to be affected by dispersion.

 

in Mario shader both reflection and refraction are affected by dispersion.

So i would like to know what the rendering expert think about this ?

 

Does dispersion occur both for reflected and refracted ray in real life ?

It seem to me that is is logical that IOR variation modify both behavior , but i would like to be certain on what happen in real life !

 

Thanks for your help !

Edited July 15, 2015 by sebkaine

[somesanctus]

but they only talk about refraction , only the refracted ray looks to be affected by dispersion.

TRUE

in Mario shader both reflection and refraction are affected by dispersion.

I dont think so. Every rainbow you see in my renders is result of multiple refraction, or reflection of previosly refracted(dispersed) light.

[sebkaine]

Thanks for your answer Alex !

 

well in Mario shader the code is the following :

...
for(i=0;i<samples;i++) {         
	fresnel(-wo,n,1.0/ior,kr,kt,wr,wt);         
	w = getfilter(ramp,wl);         
	         
	Ft += specular(wt)*kt*w;      // transmit   
	Fr += specular(wr)*kr*w;      // reflect 

	wsum += w;         
	ior  -= dior;         
	wl   += dwl;      
}      

vector iwsum = CONE / wsum;      
F  = Ft*iwsum*Ktrans + Fr*iwsum*Krefl;
...

In fact in your shader the code is the following :

...
for(i=0;i<samples;i++) {         
	fresnel(-wo,n,1.0/ior,kr,kt,wr,wt);         
	w = getfilter(ramp,wl);         
	         
	Ft += specular(wt)*kt*w;      // transmit   
	Fr += specular(wr)*kr*w;      // reflect 

	wsum += w;         
	ior  -= dior;         
	wl   += dwl;      
}      

vector iwsum = CONE / wsum;      
F  = Ft*iwsum*Ktrans; //+ Fr*iwsum*Krefl; 
...

in case 1

- Fr and Ft are computed with the ior variation and then sum together

 

in case 2

- Fr and Ft are computed but only Ft is use , Fr is set as a comment 

 

In this post eetu offer an other simpler method to fake dispersion

http://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=16567

 

He offer 2 hip , again

in case 1 you have dispersion for reflection and refraction

in case 2 you have dispersion only for refraction

 

eetu himself is not sure about what is correct ...

 

So i just need to be sure ! So if you are 100% sure that's great !

 

Dispersion phenomena doesn't affect direct reflected ray , only the refracted ones !

 

Cheers

 

E

[Skybar]

https://www.boundless.com/physics/textbooks/boundless-physics-textbook/geometric-optics-24/reflection-refraction-and-dispersion-169/dispersion-rainbows-and-prisims-611-6326/

[sebkaine]

Thanks for the link David, so it confirm what Alex say !

 

"The law of reflection doesn't depend on the light wavelength"

Edited July 15, 2015 by sebkaine