Iridescence test

[wouter215]

i'm going to do an Iridescence test based on this post.

edit1: right.. so now i have a cube and the iridescence object and a 250 frame animation, the object has 8 faces, each with a different color. they are spread over the cube.
1200000 objects in total + one cube. it takes blender a long time to export each frame..
at frame 125 now an i started 3 hours ago..
edit2: first image

[wouter215]

now with animation!
http://www.youtube.com/watch?v=YzWOpEVKY_Q

[paco]

Looks like its working quite well.... it's just a bit "expensive"...

Perhaps microdisplacement (or even bump?) might also work and be a lot faster?

[slime]

Insanely cool, but there must be some way to do this without millions of octahedrons

Isn't there something called paralax mapping, where the colour of something changes depending on the angle you view it at?

And what about the thin film property of the glass material, how does that do it?

[wouter215]

Looks like its working quite well.... it's just a bit "expensive"...

Perhaps microdisplacement (or even bump?) might also work and be a lot faster?

Insanely cool, but there must be some way to do this without millions of octahedrons

Isn't there something called paralax mapping, where the colour of something changes depending on the angle you view it at?

And what about the thin film property of the glass material, how does that do it?

the test is if you can replicate iridescence like this. not what gives better results imho.

but it sure is expensive, i can't get the octahedrons to form an even grid on the cube(just on one plane..) and if i want to export 1000000+ micro octahedrons on a cube, blender takes forever to export.
can't you just plot one object (the octahedron) and set 1000000 points in 3d space for each of the octahedrons? or use a vector so that the file-size isn't 1TB+?

[neo2068]

Looks like its working quite well.... it's just a bit "expensive"...

Perhaps microdisplacement (or even bump?) might also work and be a lot faster?

Insanely cool, but there must be some way to do this without millions of octahedrons

Isn't there something called paralax mapping, where the colour of something changes depending on the angle you view it at?

And what about the thin film property of the glass material, how does that do it?

the test is if you can replicate iridescence like this. not what gives better results imho.

but it sure is expensive, i can't get the octahedrons to form an even grid on the cube(just on one plane..) and if i want to export 1000000+ micro octahedrons on a cube, blender takes forever to export.
can't you just plot one object (the octahedron) and set 1000000 points in 3d space for each of the octahedrons? or use a vector so that the file-size isn't 1TB+?

You could use a particle system to spread the octahedrons over the cube faces. This way you only need one object. The other particle objects will be instanced.

[wouter215]

Looks like its working quite well.... it's just a bit "expensive"...

Perhaps microdisplacement (or even bump?) might also work and be a lot faster?

Insanely cool, but there must be some way to do this without millions of octahedrons

Isn't there something called paralax mapping, where the colour of something changes depending on the angle you view it at?

And what about the thin film property of the glass material, how does that do it?

the test is if you can replicate iridescence like this. not what gives better results imho.

but it sure is expensive, i can't get the octahedrons to form an even grid on the cube(just on one plane..) and if i want to export 1000000+ micro octahedrons on a cube, blender takes forever to export.
can't you just plot one object (the octahedron) and set 1000000 points in 3d space for each of the octahedrons? or use a vector so that the file-size isn't 1TB+?

You could use a particle system to spread the octahedrons over the cube faces. This way you only need one object. The other particle objects will be instanced.

but i can't get an even spread grid with the particle system, and i already have instances.

[neo2068]

[...]but i can't get an even spread grid with the particle system, and i already have instances.

Sure, you can set the particle distribution to grid.

Image1.jpg (16.58 KiB) Viewed 918 times

[wouter215]

[...]but i can't get an even spread grid with the particle system, and i already have instances.

Sure, you can set the particle distribution to grid.

The attachment Image1.jpg is no longer available

and how do i get them to show?

[neo2068]

[...]but i can't get an even spread grid with the particle system, and i already have instances.

Sure, you can set the particle distribution to grid.

Image1.jpg

and how do i get them to show?

The particles seem to be inside the cube. Check in wireframe mode. I don't have a quick answer for this problem.