Fiber optic taper

[bpepz]

Just a crazy idea i'm throwing out here, I am not sure how many people here are familiar with Fiber optic tapers, but I see great potential for them in photographic use. Here is a description from edmund optics.


"A Fiber Optic Taper is a coherent fiber optic plate that transmits either a magnified or reduced image from its input surface to its output surface. All small end diameters are optimized for 1/2" or 2/3" sensor sizes. Magnification is a ratio of the diameters of the large and small end of the tapers. Typical applications include image magnification or reduction, CCD coupling, fluoroscopy, and light sensors."

Basically I want to make a fiber optic taper that would allow me to take advantage of the full image from a medium format camera like a fuji gx or rz67 and reduce it to say, my leaf aptus 6-II, I would be able to retain the field of view of the larger format, and experience a size able gain in light gathering ability since light gain would be equal to the magnification factor.

As far as feasibility, there are some things I found out that make it feasible in my mind and try to clear up some misconceptions with people I've brought this up before.

1- fiber optic tapers can be rectangular on either side, so you are not limited to a circle,
2- even though most ccd coupling is with 2/3 sensors, they are also used with much much bigger sensors in x-ray machines, some of these sensors are 60x60mm, so a 44x33mm sensor like mine would be no problem.
3- image quality, fibers can be made as small as 3 microns, and normally at 6 microns, which would eliminate the "chicken wire" effect seen in lesser solutions.


Anyway I have contacted a Chinese company that manufactures items like this and do custom orders. I am thinking of ordering a prototype and posting the results here, also if the results are good I was wondering if how many people would actually be interested in buying something like this? if there are alot of people interested we can proably get it done drastically cheaper. anyway, just throwing the idea around, I am interested to see what people think.

[EricWHiss]

Interesting idea.  I wonder if this would also help with lens color casts?

[Kagetsu]

I'll be interested to see too.

[bpepz]

Glad to see some people interested, I have done some further reading and I am fairly sure it would eliminate color castes or problems with wide angles, even if the sensor had micro lenses, since each pixel would be seeing the light from the optimal angle. I guess they use a fiber optic plate, which is just like the taper but it is a 1:1 so not difference in size on either side on lcd panels to achieve perfect viewing angles, since the image projected on the other side is only being viewed by the optimal angle. 

I get the feeling that the company I contacted is only interested in working in bulk, not really an option for me right now. I am thinking of trying to buy some "broken" x-ray or medical imaging equipment and salvaging the fiber optic taper and using that as an experiment. I will try to post results if I ever get ahold of one.

[EricWHiss]

Yeah these things are expensive!  Into the thousands of dollars for the size needed which probably explains why they are not used much.

[cunim]

Mating a fiber optic to input and output devices is quite technical.  No problem placing a taper on a flat surface and viewing the large or small end for moderate magnification or minification.  They are too expensive to be commonly used this way but lots of engineers have surplus tapers as magnifier toys.  They are just so cool.  However, quality imaging requires careful interface design and does involve compromises in MTF, light transfer, etc.  Tapers work best bonded directly to the sensor, though I remember a patent that described using an air gap (microns) to actually improve the quantum efficiency of an intensified camera (used a fiber optic taper).  Generally, though, the larger the gap the greater the degradation in optical quality.

[degrub]

isn't this the same thing as the micro lenses on the sensor ?

[ondebanks]

isn't this the same thing as the micro lenses on the sensor ?

I guess a 1:1 transfer plate would have a similar effect to microlenses. But a taper adds an additional effect - shrinking the entire image, and preserving the light flux, as it transfers the image from one plane to another.

In the early '90s, the astronomy research group I had joined worked with an imaging detector called the RAL-PCD (Rutherford Appleton Laboratory Photon Counting Detector). This used an optical taper to shrink the image from the front photocathode intensifier stage, onto the smaller rear CID detector array. Without the taper, the field of view captured would have been too small.


If bulk is no limit, there are other ways to demagnify the image from a large MF camera onto a smaller sensor. An afocal arrangement with a longer focal length lens on the camera rear side and a shorter focal length lens on the sensor side will do the trick. The front elements of the lenses face each other, and the light between them is collimated. The coverage of the lenses must be matched to their respective image planes. So the longer focal length lens must have enough coverage for the camera's image format (6x7, 6x9 or whatever) and the shorter focal length lens must have at least the coverage of the MFD sensor at the end of the chain. The shorter fl lens must also be set to a faster f-stop so that its entrance pupil matches the exit pupil of the larger lens. It would probably work best with long focal length lenses acquiring the camera image. I reckon an extra field lens at the original focal plane would be necessary to herd diverging rays from a wideangle aerial image back into the afocal arrangement.

This description may sound a bit complicated, but it is very simple in actual implementation, as it can be done with off-the-shelf lenses. We used this approach in both variants of our 1990s TRIFFID camera, which believe it or not used Mamiya 645 and RZ67 lenses for the afocal demagnification in its later configuration. We brought TRIFFID to various large research telescopes for high-speed, high resolution imaging and photometry.

Ray