Rainbow rings (the secret revealed)

[Planapo]

See my correction: I meant chromatic aberration, of course.
But I guess this doesn't make the answer much better.

--Betty

[rjlittlefield]

See my correction: I meant chromatic aberration, of course.
But I guess this doesn't make the answer much better.

Alas, this is true. Though as I said, one small but important part is exactly correct. If it will help, I'll say that it's a very important part.

--Rik

[abpho]

Rik. You are such a tease.

[rjlittlefield]

Well, I have been working on various aspects of this problem for weeks, and I am nowhere near done yet.

Truly this is an embarrassment of puzzlements (a new collective noun?)

It seems only fair to share the riches!

--Rik

[DQE]

I still think its images of inertial confinement fusion! (insert more grins here)

I wonder if the math models of that process and Rik's optical process have any similarities? This is left as a homework assignment for the interested reader.

[TheLostVertex]

Your predicted image intrigued me. Looking aorund I discovered a program called Maskulator. I wonder if this, or something like it is playing a role in the creation of this mystery.

So I will take a shot in the dark at this. How about an infinity objective without its tube lens, sitting on a corrective eyepeice with an aperture placed between the objective and eyepeice or eyepeice and camera(I am trying to hedge my bets as much as possible )

[rjlittlefield]

Looking aorund I discovered a program called Maskulator. I wonder if this, or something like it is playing a role in the creation of this mystery.

Well, sure, for some definition of "something like it". My images are generated entirely by my own software, and there were no masks involved, but otherwise I imagine there are a lot of similarities.

How about an infinity objective without its tube lens, sitting on a corrective eyepeice with an aperture placed between the objective and eyepeice or eyepeice and camera

I am honored by the complexity, but in fact the setup I used is far simpler than that. As I wrote earlier, "The optics setups for these shots are actually the rear ends of classic setups involving Mitutoyo objectives, with two unusual adjustments. " This description is quite literal. The sequence and orientation of all optical components is exactly what I use routinely with Mitutoyo objectives, minus the objectives. The difference is in what's being looked at, plus those two unusual adjustments I mentioned.

Earlier, I asked "So now: what is the setup, what are the two unusual adjustments, and what makes the rings colored?" I've told you a lot about the setup, and I'll confirm that the rings are colored because of diffraction. But I'm still hoping that somebody can spell out what the two adjustments are, and maybe say something about why the colors don't appear in the normal spectral order. Surely these are small things to ask?

By the way, the images show "100%" because these are actual pixels from a Canon T1i, 15 megapixels APS-C.

--Rik

[g4lab]

Do the variations in the set up involve polarizing filters or diffraction gratings?

[rjlittlefield]

Do the variations in the set up involve polarizing filters or diffraction gratings?

Ah, sorry, no polarizing filters or gratings in the setup.

As I write this, I'm recalling an old TV program in which contestants could win something merely by saying a "secret word". The host would often tell the contestants "It's a common word, something you see every day." That's what I feel like here.

To make the images of oddly colored rings, my "unusual adjustments" are just extreme variations of common adjustments that surely would be mentioned in any introductory photography class.

Actually both of them have been mentioned already in this thread, but not in a way that would allow reproducing the images.

So let me rephrase the puzzle.

Background: The "rainbow rings" images were made with a setup commonly used for infinity objectives, minus the objective. The setup was pointed at a distant bright pinhole and some easily accessible adjustments were made so that the rings could be captured directly by the camera's sensor.

Question: what adjustments are required to reproduce the rings?

Here's a picture of the setup with the objective, before the adjustments.



--Rik

[g4lab]

So you must have removed the objective and racked the tube length back and to focus it at infinity where the star/ aperture was located. ? ?

[rjlittlefield]

So you must have removed the objective and racked the tube length back and to focus it at infinity where the star/ aperture was located. ? ?

What you've described is part of the normal setup. When properly configured for use with the objective, the tube lens is supposed to be focused at infinity. A simple way to set or check that is to remove the objective and focus on a distant object.

But if the distant object is a bright point, then all you'll get from that process is the standard Airy disk, with a bright center and rings in standard spectral order if you expose enough.

So again the question: what does one do differently to produce the odd rings shown here, sometimes with a dark center and often with the colors not in spectral order?

I see several folks nibbling around the solution, but so far it just hasn't come together as a coherent story. When it finally does, I think you'll be intrigued to see how short the story is!

(Oops, I need to be careful saying "coherent". There are no lasers involved here either. )

--Rik

[Pau]

In your setup I can see a diaphragm aperture just behind the objective, likely the one made by that chinese seller after Javier's request. I know that those diaphragms can close to a very small aperture so...
Is just that diaphragm full closed? (well far of the normal use of your setup...)

[rjlittlefield]

Is just that diaphragm full closed?

This is definitely on the right track...

shot with a lens, aperture diameter 2.5 mm.

So, not full closed -- that iris goes to under 1 mm -- but definitely stopped down more than normal, around effective f/80 instead of f/18 like it would be with just the objective in place.

Looking at the focused point source in live view, it is easy to see the Airy disk get wider as the aperture gets narrower.

But it's still just an Airy disk, albeit wider and simpler to see.

Which means we're still left with those odd concentric ring patterns, some of which have dark centers and unusual color orderings.

What is the second adjustment that produces those? (And why does it do that?? )

--Rik

[DQE]

Is the light source unusual in any way(s)?

In particular, I am thinking about light sources with complex spectra.

Also, is the intensity profile of the light source relevant?

Putting complex light source spectra together with complex light source intensity profiles is certainly a challenge.

Interestingly enough, during my career in medical imaging, we had to cope with such things all the time, as a part of the projection radiography specialty I worked in. Both the x-ray spectrum of the source and the intensity profile of the source (usually, an x-ray tube) would have large effects on pinhole images of the source.

http://en.wikipedia.org/wiki/X-ray_tube

[rjlittlefield]

Is the light source unusual in any way(s)?

In particular, I am thinking about light sources with complex spectra.

Also, is the intensity profile of the light source relevant?

The experimental light source is an ordinary "white" led (blue diode + yellow phosphors).

It's actually one of the units shown in THIS POST, wrapped in a piece of aluminum foil with a small pinhole more or less centered over one of the chips.

Surely this particular LED will have some peaks and valleys in its spectrum.

However, I have no reason to think that the nonuniform spectrum is relevant. Likewise for the intensity profile.

The simulated result uses a uniform distribution of 91 discrete wavelengths, multiplied by an idealized sensor having gaussian response channels:



I did a manual levels adjustment on the model output to roughly match the overall color balance of the experimental images.

--Rik