Transparency problem

I created a transparency scene, in which there is only one cube. Its attributes are set as: Alpha: 0.35, IOR: 1.55. The scene is rendered by only using the funcitonality of “Ray Transp” (no Ray Mirror)

The first image shows the result when the Camera is set to real

http://www.cs.auckland.ac.nz/~jli023/images/transparency_0.jpg

The second image shows the result when the Camear is set to blend the blue and white color.

http://www.cs.auckland.ac.nz/~jli023/images/transparency_1.jpg

In the second image, it seems there are two cubes, one is inside another.

Do you have an idea why the Camear setting can affect the transparency rendering?

Thank you very much,

Jing Li

what you changed is the world background

it looks like the first image is not rendered with raytracing enabled (or the background is too dark) but the second one is: the ‘second cube’ probably is the self-refracted cube

test this by changing IOR, the inner cube should grow or shrink

thank for you reply. I tested the scene, the inner cube changed size based on the IOR.

I do not understand what the self-refracted mean. Could you explain it more detail?

Thank you very much,

Jing

He means self-reflected; the object is reflecting a reflection of another reflective object in which it itself is reflected. Confusing?

Object 2 reflects object 1. Object 1 reflects object 2 and object 1. Object 1 is self-reflecting.

%<

But on my scene, there is only one object - the Cube. How can it reflect itself to its inner? It is possible that the world background is counted as a object in Blender 3D?

Thanks,

Jing

Can you upload the .blend or mail it to me at (dotdotdashh@yahoo.com) so we can look at it? Otherwise we’ll all just be guessing.

%<

Jing, the 3D world isn’t really an object.

When you set the cube to trasparent with reflections, it’s not only the outer faces that reflects - the inner faces also reflects. In Blender, when something has a reflection setting, every faces act as a mirror whether it’s the backside of the cube or the front.

Maybe a picture might help.

http://img.photobucket.com/albums/v639/Ryu_Kaiser/CubeRef.jpg
(Depth of 2)

http://img.photobucket.com/albums/v639/Ryu_Kaiser/CubeReffracdepth1.jpg
(Depth of 1)

http://img.photobucket.com/albums/v639/Ryu_Kaiser/CubeReffracdepth0.jpg
(Depth of 0)

Notice how the monkey being inside is reflected. The inner faces reflected the monkey. The inner cube you see there isn’t really a cube, it’s just the reflection of the other faces inside the cube. The only meshes I have there is a cube, a monkey’s head and the plane which the cube sits on ofcourse.

If you really don’t want to see the ‘cube’ reflection inside your mesh cube, then turn the Depth (under the Ray Trasnparent button) to 0 or 1. A Depth of 2 gives you that cube reflection which you don’t want unless you want to make diamond things ^_~

I hope this clears things up for you,

Wolf

First of all, thank you all for helping me to understand the self-refraction.

I did some test with Blender 3D, and found out the inner cube would be disappear if the “Traceable” shader of the cube were set to off whatever you set the refraction depth or the IOR values.

Does anybody know what the functionality of Traceable is? The function is located under the Material’s Shaders tab.

Thanks a lot,

Jing

I just happen to be working on a tutorial on TRACEABLE at this very momnet.

OK I spelled it wrong, but here’s an example…

I did more test on Refraction in Blender 3D and POV-Ray. The test scene is just a white cube, which materail attributes are set to { Alpha 0.35, Reflection 0, IOR 1.55, Traceable on, Shadow off, TraShadow off, Bias off, Radio off } in Blender 3D, and { Filter 0.35, Reflection 0, interior{ior 1.55} } in POV-Ray. The background color of the scene is set to sky blue. All of the blend files and pov files and images can be accessed via http://www.cs.auckland.ac.nz/~jli023/images/Refraction/

The difference between the renders done by Blender 3D and POV-Ray are obvious.

When the depth is set to 0 in both of tools. Here is the render done by Blender 3D (its correspoinding blend file is cube_1.blend)

http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_1.jpg

Here is the image done by POV-Ray

More images:

Depth = 1:

Blender 3D

http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_2.jpg

POV-Ray
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_1.bmp

Depth = 2:

Blender 3D

http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_3.jpg

POV-Ray
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_2.bmp

etc…

Then I tested the scene using Blender 3D again with turning on the Shadow.

Depth = 0
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_1_shadown.jpg

Depth = 1
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_2_shadown.jpg

Depth = 6
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/cube_3_shadown.jpg

I wonder why the Blender 3D and POV-Ray give different results when rendering a transparent object. Which result is more correct or realistict?

Thank you in advance,

Jing

Oops, it seems that the bmp images can not be parsed.

Here is the images rendered by POV-Ray:

Depth=0

http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_0.jpg

Depth=1
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_1.jpg

Depth = 2
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_2.jpg

Depth = 4
http://www.cs.auckland.ac.nz/~jli023/images/Refraction/box_refraction_4.jpg

When the depth is either 0 or 1, the inner cube is hard to see or there is no inner cube at all in POV-Ray rendering. After the depth is great or equal to 2, the inner cube becomes visible.

The Blender 3d does the different rendering from the POV-Ray. If the Shadow is turned on, the inner cube completely is invisible whatever the depth values. On the other hand, if the Shadow is set of off, the inner cube is visible even the depth is 0.

I am so confused which render is correct.

I do appreciate your help.

Thanks,

Jing

I got your email and your .blend but I would rather reply here. In an email it might sound like I know everything when i don’t really and here there are people who can correct me. Other users can in the future also find this thread with a search and it will help them.

Can you please explain what it is you expect to see with RayTransp? If you can explain this then maybe we can get some answers that help you.

%<

Hi Fligh, I do not understand why the inner cube appears in Blender 3D when rendering a transparent cube even though the depth of the Ray Transp is set to 0.

I just wonder there may is a bug in Blender 3D’s Ray tracing as the results rendered in Blender 3D are different from the results rendered in POV-Ray.

Thanks,

Jing

why the inner cube appears in Blender 3D when rendering a transparent cube even though the depth of the Ray Transp is set to 0.

The short answer is because the IOR (incidence of refraction) is set at 1.55.

If you really want the definitive, technical answer as to why blender uses render algorithms different to PovRay and what the basis for these are then you will have to email or PM eeshlo, cessen and/or Ton (you will find their profiles at the top in the “Memberlist”). Eeshlo won’t be around for about 2 weeks.

%<

Actually, I had nothing to do with the raytracer in Blender, so I don’t know any better about that than anyone else. I could look into it, though.