Help me render telescope lenses? (Twilight Render V2)


#1

I’m trying to render a view through a telescope. Looking through 2 lens elements works okay (aside from mild polygon distortion or something), but the 3rd element is opaque in the render. How do I render through 3+ lenses? Are there certain render settings I need to use?

3rd element is closer:

Telescope.skp (765.1 KB)


#2

I used exterior daytime progressive render and the 3rd element is now see through :grinning:, but the distortion is pretty bad. Maybe if I make it high poly…


#3

there may be tracing, transparency, reflection depth settings you can adjust. For speed these are often limited to avoid unnecessary ray-tracing for details you may not see. In Thea Render for example these can be increased - For example, I had done a hall of mirrors where the multiple reflections cap at 5 and beyond which it renders black, but the setting can be increased - not as far as infinity of course !!! but enough to where you cant see any more as they get too tiny.


#4

Thank you for the reply. I figured out how to render the 3rd element by changing the render preset. I actually tried a bunch of other render presets before posting, but none of them worked. I didn’t try the progressive render preset and figured it out after posting. Then I made it high poly to correct some of the distortion and it improved it, but it still has some distortion in the middle…

Bidirectional Path Tracing Preset:

My two remaining goals are to:

  • Remove the distortion
  • Make the 3rd element magnify

I have a feeling that my model might need to be made with exact dimensions from a plan and I might have to go higher poly… :sweat:

My end goal is to be able to design telescopes in SketchUp.


#5

For your interest



#6

I think the distortion problem might be because of my model because the distortion rotated 45º when I rotated the lens 45º…

I tried changing n-gons into tris this time around too, but no luck.


#7

How did you arrive at the shapes and locations for the lenses?


#8

I found this image and just made the general shape without using any real dimensions. I tried to make it semi-accurate though. In my model, I think the lens is only 1"-2" wide.


#9

I don’t think your model is accurate enough.


#10

I totally agree. That’s probably what it comes down to.


#11

You need to sort out the index of refraction for the “glass” you’re using and determine the required foal lengths for the lenses. You also need to sort out the eye relief for the camera location.


#12

Thanks for all the help so far.

I used a refractive index calculator (http://www.endmemo.com/physics/snells.php) to calculate the angle of light rays passing through glass & air to design this lens.

Edit: Oh no… I accidentally reversed the IOR’s of the air and glass, so my diagram is totally wrong.

I managed to magnify the background with this lens.

I’m hopefully going to add another lens to increase the magnification.

The distortion is terrible though… Hopefully going high poly isn’t the only fix because last time my lenses had over 100,000 edges. :sweat:
Any help with the distortion would be greatly appreciated. Even at 100,000+ edges, there’s still distortion.

Edit: Changed IOR from 1.520 (crown optical glass) to 1.6 (flint glass)

P.S. Most telescopes try to focus the light, but I think dispersing the light will help you see smaller rays of light by getting the peripheral information out of the way. I think the goal should be to gather the smallest rays of light possible, then enlarge them into an image.

Edit:
Interesting… I found out that geodesic spheres and revolved spheres have different types of distortion.

Meshes:

Revolved Sphere

Geodesic Sphere:

The distortion is terrible…


Mathematically accurate curved surfaces
#13

Like I know anything, but it looks like you are experiencing a moire pattern due to the faceting of the geometry. Just a thought. And perhaps your issue with the third lens is some sort of polarizing?

The only reason I can guess at this is I’m a photographer, and this sometimes happens in that field also.


#14

It started working and I’m not sure why, but the problem is solved at the moment. Thanks for all your help!
@jimhami42’s model

My lens model

There’s no more distortion.


#15

Here’s an update:

Reference Image
NewtonianLightPath649499

Model

View

I think some mechanical plans with dimensions might help, so I can draw it more accurately… :sweat_smile:

I had difficulty finding any on Google though.


#16

I created a nominal design using a 6" parabolic reflector with a 25" focal length. The 2D parabola is given by:

y = x^2 / 100 from -3 to +3.

NOTE: The model is scaled 1000x to eliminate any “tiny edge” issues.