back
6 comments
Funny to see this item on HN. I taught holography with Steve Benton at MIT, and put together a few of the documents available at MIT Open Courseware referenced in the document. The material is pretty specific to the optics of diffractive elements. I have a bunch of great Tcl/Tk holography demo apps I used with the course that I need to port to something that isn't Tcl/Tk.

Sadly, much of Steve's pioneering work in holography at Polaroid and later with the Spatial Imaging Group at MIT predated the blossoming of the world wide web (and "pre Google") and thus has been lost to history.

I am pleased to see work in autostereoscopic displays, computer-based optical predistortion and light fields - all of which Steve and Spatial Imaging did some neat stuff to develop - continues.

Steve would have loved the Looking Glass display. "Lovely!"

I just emailed someone in holography a question, maybe you wouldn’t mind answering...

Say I have a transparent hologram plate and shine a laser into the “front of the plate” to get a virtual hologram (the hologram appears “inside” the plate). Now let’s say we break the plate in half (or otherwise have a clone) and we place the duplicate in the same plane (like dominos) only the 2nd plate is facing front out (so laser, plate 1 front facing laser to get “virtual” hologram, an inch of space then plate 2 with back of plate facing plate 1/laser) what happens? Does plate one give the “virtual” hologram and plate two give a “real” hologram (project)? So in other words, is this right:

Laser —— | (virtual hologram) | (real hologram) ?

Is the goal to have a "walk around hologram": one that you can look through from one side and see a real image projection on the other?

If so, what you propose won't work for a couple of reasons. First, you're flipping the plate, which means left-right or top-bottom is going to get flipped. Second, with a single laser you'll be illuminating one version as a transmission hologram, and the other as a reflection hologram. That's not going to work, I'm pretty sure.

Another way to attempt to do what you describe is to use two lasers to illuminate the same hologram: one, say, at 45 degrees to the plate coming from one direction, and the other 180 degrees opposite, shining on the plate from the other direction. If the two beams mimic the original reference source (but with one being the perfect conjugate: same angle but opposite direction), you'll get both a real and a virtual image.

However, and this is the big however, the hologram "knows" only about light from the parts of the object that are visible to it at exposure time, and it faithfully images them back into space exactly where they were. That means that the real or projected image will be pseudoscopic. It will be viewed from behind it's original location and will appear inside out or depth-reversed.

An interesting effect, but probably not what you were looking for.

Breaking the plate in half isn't like cloning the hologram, although it's a common myth. Imagine a hologram like an ordinary window. If you look through it, you can see objects on the other side from many different perspectives. Now cover up half the window. You can still see many of the same objects, but from a restricted range of views. And if a fly was climbing up one side of the window, you probably can't see the fly from the uncovered piece: the optical information about the fly is gone. To clone a hologram, you can either recreate the interference pattern (e.g., stamped foil holograms from a master) or reconstruct the original image optically and make a new hologram of the original. Rainbow or "Benton" holograms use this technique (can be used for both techniques, actually).

Hope that helps.

I know HN rules don’t want simple “thank you posts” but I really appreciate your thoughtful response.

In my proposal I did know the real hologram (if it worked) would be “inside out”/“depth reversed” but I left that part out. These are very difficult concepts to try to visualize and when googling them I tend to come up with academic papers (like this very interesting article) or cartoonish representations of the concepts as opposed to say pictures/video of the specific hologram properties.

I did not know setting up the hologram using two lasers to mimic the original reference source (with one from the back of the plates would give both real and virtual from the single plate (though it makes perfect sense). It just creates more fascination and questions. I really need to get my hands on some master plates and start playing to satisfy these curiosities.

Doing what you describe (getting an off-axis laser transmission hologram and playing around with it) is one of the best ways I know to build intuition about light fields. A hologram is essentially a dense surface of light field emitters.
Awhile ago, for a different article, I wrote "Best article on HN ever." I take that back. That "award" goes to this article. Best article on HN ever. Makes me want to study Holography. Absolutely fascinating!