It’s Probably Not a Hologram (sorry)

In 1977, Luke Skywalker accidentally stumbled onto a holographic projection of Princess Leia while fixing R2-D2. Those scenes from Star Wars: A New Hope embedded the idea of a “hologram” to be any image floating in mid-air into the lingo of pop culture.

When Dennis Gabor invented holography in  1947, some 30 years before Star Wars, the term had a specific meaning which had little to do with mid-air images. Holography was a technique which could record an optical wavefront into a photographic medium using coherent interference for later reconstruction with a wavefront matched to the original reference beam.

No wonder pop culture has picked up on the “cool, a 3D image in mid-air!” part and glazed past the physics. Holography is a dense subject with lots of caveats, but let’s back up and get the basics. This will help in a moment to understand how real holograms are different than what marketing departments think they are.

The light coming off an object, like the cat in the picture below, comes from many directions. A wavefront, or “wave” for short, depicts all the light at any one moment. When these waves reach your eye, they are focused into an image, and you see the cat. The only thing new here is the terminology.

A hologram uses a second wave, known as the reference, which has some specific properties (which we’ll skip over for now). The object wave and reference wave interfere together; if you put a piece of photographic film so that both waves hit it, you can record the interference pattern (middle diagram below). The key take-away is that you’ve now stored information about how the object wave and reference wave interacted. (Optics nerds: the interaction with the reference stores not just the amplitude but the relative phase between the waves. Photography stores only the amplitude information.) Develop the film and you’ve got a hologram.

Side note: hologram roughly translates as “to write the whole”, referring to the entire wavefront’s phase and amplitude information getting recorded.

Later, you can illuminate your hologram with a beam that’s similar to your reference wave (bottom diagram below). As it passes through the hologram film, the intensity pattern re-shapes it to have the same shape as the object wave. (That physics we’ll save for another post.) When the reshaped wave reaches your eye, you can again see the image of the cat, exactly as you had before. The wave has all the details from multiple angles, so you can move your eye around and see different parts of the cat. If you are lucky enough to have two working eyes, each eye sees a different part of the cat, and you get a sensation of 3D.

Holography, simplified. With cats.

There are a couple key things to remember about holograms. Since the image you see is from light (the reference wave) getting reshaped into a specific object wavefront, it means that:

1. you always need a light source with the right properties,
2. you always need a surface to do the reshaping, and
3. the resulting reshaped wavefront needs to get from the surface to your eyes.

The Star Wars scenes fail on #2 (R2 projected into thin air) and #3 (there wasn’t something to direct the light back towards Luke and Old Ben). (My theory is that the 3D projections in Star Wars are closer to laser plasma displays and may or may not be entangled.)

Side note: holograms are classified into thin (surface) and thick (volume holograms), referring to the thickness of the material used to diffract light. Thin holograms are on the order of a few wavelengths. Volume holograms are tens to hundreds of wavelengths, or about as thick as a piece of paper. That’s a long way from the volume in R2D2’s projector beam.

So let’s take a look at a couple “holograms” with all of this in mind.

There’s a famous moment from the music festival Coachella where deceased rapper 2pac performs alongside Snoop Dogg. Is it a hologram? If you read up on the physics, they were actually using a partially reflecting surface to make it seem like 2pac was standing near Snoop: it’s an illusion known as Pepper’s Ghost, and it’s been around since way before holograms. There’s no need for a controlled light source (#1), and there’s no reshaping of the wavefront (#2). In fact, Pepper’s Ghost should get a lot more credit: it gets used in most of the “holograms” you actually see. Turning your phone into a hologram or projecting a hologram from you phone or presenting a holographic lecture or creating a synthetic pop star or resurrecting a dead performer or winning Indian voters are all examples. They’re cool, but definitely not holograms.

Rule of thumb: any time you see a “hologram” on stage, there’s a 99.8% chance it’s Pepper’s Ghost. There’s no need for stereoscopic views at that distance, and large-scale real-time holograms aren’t feasible yet.

The “hologram” directing foot traffic in London’s Holborn Tube station fails on #2: there’s no reshaping of the wavefront. It’s a projector with a screen cut to shape.

Tom Hank’s A Hologram for The King follows the same tack as Star Wars, projecting a “hologram” into mid-air — so it also fails on tests #2 and #3.

Rule of thumb: if you see a projector mentioned anywhere, there’s a very good chance it’s Pepper’s ghost, a simple projection, or computer generated imagery.

Microsoft has been getting press for its HoloLens device. It displays an image in front of each eye with a slight shift to give an illusion of depth. It’s again Pepper’s Ghost, but a slightly fancier version. So not a hologram. The only caveat is that Microsoft could be using a hologram to aim the rays towards the viewer (as a holographic optical element), but the image you see is not a hologram.

Researchers at Queen’s University in Canada have a flexible cell phone display they call a HoloFlex. This one’s tricky and requires some reading: it fails on #1 (light source with the right properties) and #2 (surface reshaping a wavefront). It’s actually a light field display, which uses tiny lenses to redirect light from each display pixel to a different direction. (It’s one step up from lenticular displays: you know those post cards which look 3D or animated as you tilt them side-to-side? Same idea!) The physics is different from a hologram, but your eye gets a similar effect of being able to see different views of an object. It’s not a hologram, but a cousin. I’m just disappointed that the researchers called it a holographic display when “light field display” sounds pretty awesome.

Security holograms on credit cards and the gift wrap pictured here are real holograms.

If you Googled HoloFlex, you might have found Holoflex Ltd, makers of holographic stickers and tapes. Is it a hologram? Yes! Distant light (#1) gets reshaped by the microscopic grooves in the foils (#2) and is reflected towards your eye (#3). One hint is that with replicated holograms, you usually see rainbow effects as you look from different directions. The hologram on your credit card, the manufacturer’s stickers littering your computer, the security strip in your large denomination currency, or most of the sparkly gift wrap out there actually are real holograms.

Rule of thumb: if the colors change with angle, there’s a reasonable chance it might be a hologram. If it has a reflective background, there’s even more reason to think it might be a hologram: that reflector is reshaping the wavefront (#2) and reflecting the light towards your eye (#3).

Zebra Imaging makes large-scale holograms, including a hologram of Bob Marley. Is it a real hologram? Yes! They use controlled illumination (#1), a sheet of material reshapes the wavefront (#2), and the wave is reflected towards your eye (#3). One extra hint is that, as you change your viewpoint you can see different parts of the scene. The high resolution and need for controlled lighting are good hints that this isn’t a light field.

So in the end, most of the “holograms” in the news headlines aren’t holograms in the scientific sense, and the few that are may not be obvious as holograms. To me, that’s an indication that the distinction between the science and the effect isn’t understood by writers and marketing departments–and that we need different terms to represent either the optical science or the sensation of seeing something that isn’t physically present.

We also need to give Pepper and his ghost a lot more credit.

 

Blog Motivation

The MIT Museum is one of my favorite places. There’s always something interesting and inspiring. Take, for example, a display they had highlighting Oliva et al’s work (e.g., Monroe Einstein) that combined two faces into a single image. One face was visible up close, the other was visible from a distance. Having just taken a class in Fourier optics, it made sense that one was high-pass filtered and the other low-pass filtered, and your eye would select an appropriate frequency band depending on your viewing distance. I was inspired to try making my own hybrid images, one which hung in my apartment until the day I left Cambridge. I was motivated to use the tools I had and curious to try developing the methods.

Me and my buddy, Joseph Fourier.

Fast forward six years. I’m still fascinated by image processing and the ways that a human eye can pick out new meaning from an image of something else. (Pointillism, half-toning, and photomosaics, for example.) I’ve been having fun coding random ideas when they hit me, and posting results to a personal online gallery. But now I’m branching out and creating a public blog in the hopes of either inspiring other people to tinker, or giving other people some ideas. I’ll also admit that many of the ideas presented in this blog were inspired by other works (science or art), and that I’m really in it for the coding challenges.

In addition to image processing and image representation, there are other topics in Matlab, computer vision, machine learning, and my field of optics that I’m planning on writing about. For my own fun… and reference, in some cases. If other people find those posts interesting or helpful, that’s great.

Monkeys and cogs,

Nick