Skip to main content

Holograms

How Holograms Work​

Imagine you’re playing with a flashlight in a dark room, shining it on an object. The light bounces off the object and into your eyes and that’s how you see it. Now, imagine that you could freeze all that light in place and save it like a photograph.

Instead of a flat picture, a hologram is a light trap that remembers exactly how light would bounce off a real object. To work, it requires a two things:

  • A laser beam, because we need light that is predictable.
  • A beam splitter, so we have the worker beam which shines in the object and then goes onto a special film, and the reference beam, so we know how it all started.

Both beams will interfere, creating a pattern on the film. Instead of a picture of the object, that’s a set of instructions for how the light should bounce to recreate the object later.

Now, if we shine the same laser beam through the film, the patter will “bend” the light in the same way as if the object would be there. So you will see the object, despite not being there. The problem, as usual is a matter of scale: add high resolution, real time motion, real time rendering, large scale manufacturing… things get difficult, quickly.

Scale and Computation​

Holograms work beautifully in small, controlled experiments: a small object, a dark room, a single laser. But that won’t get a lot of people to use it. The problem isn’t just making the hologram bigger; it’s that light is complicated, and simulating how it behaves at large scale requires an absurd amount of computation. A real hologram requires every point of light must be precisely calculated, and the math behind it grows exponentially with size and detail.

Researchers are working on clever shortcuts, like using AI, special algorithms, or even new kinds of optics, but for now, we’re stuck in a trade-off: the bigger and more realistic the hologram, the harder it is to compute.

Breakthroughs Bringing Holograms Closer​

One of the most exciting developments comes from AI. Instead of relying solely on brute-force physics calculations, scientists are now using neural networks to predict how light should behave in a hologram, drastically cutting down the processing time. This is not different as how ray tracing works in games, for example.

Another leap forward is in new materials and technology getting better: special surfaces called metasurfaces can bend light in precise ways without needing complicated setups. These aren’t full holograms yet, but they’re paving the way for thinner, more practical displays that could eventually evolve into true holographic tech. Also, larger and faster spatial light modulators show up, holographic projection systems are also constantly improving. The pixel count on liquid crystal on silicon as well as microelectromechanical systems phase displays are increasing by the millions. This means that new photonic integrated circuit phased arrays are achieving real progress.