Cameras for Computer Generated Holograms (CGH)

[holorefugee]

If you have an optics background, this patent will help you understand the problem better.

http://www.google.com/patents?hl=en&lr= ... ms&f=false

[jnhong]

Gigabit film is a high-resolution B&W 35 mm film that's reportedly better than Tech-Pan film, but just as hard to expose and develop properly. Photography professionals who work on document copying or high-resolution imaging would know much more about how to deal with this film and the cameras involved for document capture. You will get better technical information from document-photo professionals regarding camera bodies, lenses, lighting, and development chemistry. You should try to find a consultant in that area, maybe someone with experience in microfiche or microfilm. While we are all holographers here, there may not be many who understand technical photography. It's a totally different field versus traditional holography. Please note, it's also a very different field from standard professional photography, e.g., portrait or landscape or product photography. You really need to find someone who has experience in document capture.

Now, that is only the start of your troubles. Even if you were able to transfer your digital interference pattern perfectly onto a (small) piece of film, the best resolution you are likely to achieve would be 2.5 um to 5 um. Compare that with the 40 nm grain of the very popular PFG-01, a holographic film that is readily available, fairly affordable, and easy to develop. This affordable holo film can produce way better holograms than your multi-core-computed Fresnel or Fraunhofer diffraction pattern. It's very easy to see why: let's say you use 632.8 nm for recording, your digital pixel spans 8 lambda, while the holographic film grain is less than 1/12 of lambda. You probably want to ask your professor what kind of holograms you want to record and reconstruct, and about the techniques to overcome that disadvantage.

[xavislow]

What we want to do is creating an hologram so that when Gaussian light passes through it, we obtain Laguerre-Gaussian modes (it's like Gaussian light but with a torque, in case you are not familiar with this notation). We will use these LG modes for our experiments.
I'm sure that this method of printing on a paper an interference pattern and then taking a picture of it is not the most efficient method to do it, but it works. However, I'm interested in improving it if it's possible. Maybe I could try to use this PFG-01 films, yet they seem to have a much better resolution of what I can reach with my method. But I don't know how I could grow the hologram there without creating the interference pattern on a computer....cause I guess it wouldn't work to directly print the interference pattern on the film...

Thanks for replying me guys!

[Ed Wesly]

The Nikon and Canon f/1.4 or f/2 50 mm lenses are the ones to look for. Their macro lenses might also be up to that standard.

That business of f/5.6 is a little bit erroneous. The best resolution for a "diffraction-limited lens" is at its biggest opening! (Diffraction Limited meaning the wave-like nature of light limits the size of the smallest resolvable feature.) Most lenses aren't diffraction limited, with the dreaded "Seven Deadly Aberrations" contributing to the lack of resolution at the biggest openings. Conventional photographic wisdom states that a lens works best a couple of stops from the biggest, so I would suggest when you do photograph your patterns, shoot at every stop from biggest to smallest to see what you can capture.

I am available for consulting or kibbitzing, PM me c/o this forum so we can work out the details.

[Jeffrey Weil]

My rule of thumb for that kind of thing is the middle of the aperture range. Thats usually around f5.6-f8 or so depending on the lens. I've never noticed any difference between the middle and one up or down.

Jeffrey Weil

[Ed Wesly]

Here is a lens test that I use for an example in my optics class. Notice the best results are one or two stops down from max, and with these f/2.8 lenses that turns out to be f/5.6! Golly! So it depends where you start.

http://www.mamiya-usa.com/assets/pdfs/6 ... sChart.pdf

Although these are high quality medium format lenses, they are only peaking at 70 -80 lp/mm. that's all you need for general purpose photography. For nuttier stuff, you might need something like this:

http://holography.ning.com/photo/steppe ... ntext=user

[Dinesh]

What we want to do is creating an hologram so that when Gaussian light passes through it, we obtain Laguerre-Gaussian modes

This might help you, if you're not already familiar with these papers:

N. R. Heckenberg, R. McDuff, C. P. Smith, and A. G. White,
Opt. Lett. 17, 221 ~1992.
J. Arlt, K. Dholakia, L. Allen, and M. J. Padgett, J. Mod. Opt.
45, 1231 ~1998

On the other hand, if you can generate an LG beam, eg by using an external cavity, then you can always interfere a Gaussian beam with the generated LG beam. Then, when the hologram is hit by the Gauissian, it'll reconstruct the LG beam. However, I suspect you may want to do this "in-line", in a geometry known as Gabor geometry. You can do this by using a partial mirror to reflect one of the beams in-line with the other beam. In this case, unless the efficiency of the hologram is 100% (unlikely!), the zero order Gaussian will be mixed in with the reconstructed LG beam.