[tom]

p.s. maxwell prism has black reflections due to open room (3 faces), winosi room has 6 faces with 3 unseen faces from camera.

[Thomas An.]

Fantastic
Maxwell dispersion now looks A-OK

[hdesbois]

Yes, fantastic indeed!
HD

[Axe]

Nice! And I like the soft blur between the colours!

[jotero]

WOW...... very nice tom :*)
for all freely for testing

ciao
torolf

[Thomas An.]

I wonder what is the real function of the lens.

Tom, what abbe and ND numbers did you use for the prism?

[andronikos916]

WOOOWW!!!

[Maximus3D]

Nice one Tom

But in the reference scene the dispersion pattern shows more intensity than it does in the Maxwell rendering, and it has that streak in the center.. maybe that's due to lens/material imperfections only but i don't know.
Somehow the reference looks tighter and more focused if i express myself that way

/ Max

[Aldaryn]

Incredible!

Would you please share some scene information with us, couse for example I've tried to recreate this scene in Maxwell, (with geometry imported from the WinOSI scene) but I've only got a bright, a bit dispersed spot of light on the wall, not this wide spectrum.

[lllab]

what software did render the first pic?

cheers
stefan

[j_petrucci]

yes, I agree with Maximus3d...

[tom]

Thank you all !!!

Here is MXS and WRL file that I generated and packed as ZIP for you.
http://www.atillaakin.com/maxwell/maxwellprism.zip

I didn't use any WinOSi data other than reference image.
To be honest I couldn't find any file and technical info about it.

This was a forced test again and was very hard to set it up according
to the reference image. The following material values may not be
real life of course but as you can easily compare the refraction index
of the prism is nearly the same. You can check it from reflections.
Lens is for adjusting the focus of rainbow on the wall. You will all
shout prism abbe is 2 but I think there is no way to catch this much
dispersion with such setup. And I'm not sure about WinOSi test parameters.


Here are mine:

Lens
abbe=37.17 (default bsc1 abbe, no matter)
Nd=2.22 (i manually set for the focus i wanted,
this can be also adjusted by the lens geometry
but i decided the easy way, playing with index)

Prism
abbe=2.0 (forced to disperse more)
Nd=1.4 (according to reference image)

Ceiling Lamp
200W D65

Emitter
10000W D65

Camera
Lens=70mm
fstop=1/8
shutter=1/30

IMHO, accurate physical conditions is not necessary to test the colors
of dispersion pattern because colors are not affected by the parameters.
I only forced them beyond real values to have this spectrum wider.
So, you don't need to worry about physical accuracy of materials.

[jotero]

what software did render the first pic?

cheers
stefan

winosi http://www.winosi.onlinehome.de/

[jotero]

Thank you Tom

[tom]

Torolf, I see you're very experienced at such stuff...
So, can you provide us WinOSi parameters of the test?
Thank you!