Methods for capturing a Fabry-Perot Interference Pattern

[undefined314]

I'm planning on doing an (undergraduate-level) experiment to study Zeeman Splitting in Cadmium.

There's no complete set of instructions for the lab, but after seeing the materials, it appears that I will attempt to use a cylindrical Fabry-Perot Etalon to resolve the wavelength differences between light emitted from certain Cadmium transitions with/without a supplemental magnetic field, and then use that information to estimate the shifts in the energy states.

For preliminary reading, my instructor posted a paper describing a computational method for Fabry-Perot fringe analysis (Least squares algorithm for rapid analysis of Fabry–Perot fringe patterns, by O'Hora, Bowe, and Toal). This leads me to suspect that casting the interference pattern on a piece of graph paper and making measurements this way will not be sufficient to resolve the wavelength difference between light emitted from transitions with vs. without the presence of the external magnetic field. My instructor usually indicates optional readings quite clearly, and I'm aware that the change in the energy levels will be quite small, given the limits to the magnetic fields the equipment can generate. (Though I'm not sure how much the Fabry-Perot patterns can exacerbate small wavelength differences.)

I spent a summer working in an optics lab and worked on image processing and camera testing. There, we had cameras (ThorCam C1285R12M) that would output multipage TIFF files that I could then split and process with Mathematica. (Each individual image was then treated as a 1280x1024 matrix populated with 8-bit grayscale values.) We would frequently shine laser light into the camera sensor (with a filter screwed on) and capture images this way, rather than shining a laser onto a screen and taking a picture of that.

For my current experiment, I have no such specialized cameras, but I saw that Smartphone mounts were available for use with the optical breadboard. I don't think I'll be able to direct light directly into the smartphone camera lens, and I'm also uncertain if the presence of that extra lens of unknown characteristics could distort the image. The alternative would be casting the pattern on a screen and simply using the camera to take a picture of it, but then I'd have to deal with distortion due to the image being taken off of the optical axis. I believe I could use OpenCV to compensate for distortion of the image, but I'd prefer not to rely on that.

I do have my own camera (Sony Alpha 6000) with a removable lens. I'm wondering if I should remove the lens and place the camera in the path of the light (with the room lighting off). I have heard that shining a laser directly into a camera sensor without filtering can cause damage, but this is not a laser. It is a cadmium lamp, and the light will interact with a quartz Fabry-Perot etalon before reaching the camera.

My camera can take RAWs without automatic corrections for vignetting and other such things. I think I should be able to work with these similar to how I did with TIFFs in Mathematica, except these will not be in grayscale.

I'm mainly hoping to get much closer to having the sensor normal to the optical axis. I've also thought about using photosensitive paper and developing it as a photo to scan, but I believe that would take too long.I don't expect to be able to figure out all the answers before I actually get into the experiment, but I'd appreciate any guidance to avoid damaging equipment. Thanks for any advice.

 

[Andy Resnick]

Summary:: I'm thinking about using a consumer camera without the lens to capture a Fabry-Perot interference pattern that will be processed in Mathematica to extract wavelength information. I'm wondering about concerns for damaging the camera.

[snip]

I do have my own camera (Sony Alpha 6000) with a removable lens. I'm wondering if I should remove the lens and place the camera in the path of the light (with the room lighting off). I have heard that shining a laser directly into a camera sensor without filtering can cause damage, but this is not a laser. It is a cadmium lamp, and the light will interact with a quartz Fabry-Perot etalon before reaching the camera.

[snip]

I don't expect to be able to figure out all the answers before I actually get into the experiment, but I'd appreciate any guidance to avoid damaging equipment. Thanks for any advice.

Most likely, you will not damage the sensor (other than dust!). However, I suspect the Bayer filter will complicate use of RAW 'images'. In addition, you may find additional fringe structures arising from the multilayer construction of the sensor (anti-aliasing/optical low-pass filter and IR cut filter) and if there is some kind of auto-focus beamsplitter present, that will cause additional fixed-pattern noise.

If you want to use a lens, try focusing it to infinity and see if you get an in-focus fringe pattern.

 

[sophiecentaur]

This sounds as if it would work. The only consideration is the spatial frequency of the fringes compared with the APS-c (?) sensor which is 23.5X15.6 mm. This is equivalent to the so called Prime Focus method for astrophotography with an SLR. If the fringe spacing is too great for this method, you could use the camera lens to image the fringes. This is another method used for astrophotography and can produce different fields of view, in addition to what you get with a given objective focal length and sensor size.
As for exposure time, you won't hurt the sensor as long as the fringes you actually see are not ridiculously bright - just choose the best exposure to give from near white to near black level. Good idea to avoid shining a laser directly on the sensor but even then , you should be able to mask it on the sensor or just tilt the camera away from the direct line of the laser. (You would also be careful doing the measurements direct, of course)
Any such method would be preferable to trying to make live measurements, once you have calibrated the scale of the images you see on the computer screen. You can alter the exposure and contrast during post processing to get things just right.