What is a holographic display?

 
 
 

Holograms were actually discovered by mistake by Dennis Gabor back in 1948. He was attempting to improve on the quality of electron microscopy when he discovered the concept of holography, which relied on the constructive interference properties of light. He coined the term hologram from Greek, which means “full information” or “full record”. The word hologram has evolved since then to capture the imagination of the world, thanks in particular to special effects in movies, television and gaming. Now, a holographic display means many things, and the term is abused badly in the industry. But a holographic display, in the truest sense, is a re-creation of light in the same way that we experience it in real life.

A particular type of holographic display is called a light field display, which is one of the most efficient technological approaches that retains the critical properties for a compelling viewer experience. Similar in concept to a magnetic field, a light field describes all of the visible light in a volume of interest. If a light field display is of sufficient quality, it is appropriately described as a holographic display.

DennyG_scale.png
 
AHBG3.png

What is possible and what is not.

 
 
 

Science fiction has pushed the imagination of holograms beyond the boundaries of what is possible with current physics.

Unfortunately, you can’t get light to simply change direction at a point in space, just because you want it to. Therefore, light must either come from a light source, or interact with a physical surface to change its direction, colour, or intensity. The only known way to make light appear in front of you out of thin air is by focusing lasers to generate a plasma event, which is essentially setting the air on fire. Because of this, the holograms in movies that appear out of nowhere are impractical and dangerous - if you want to do that, prepare to be fried!

BurnedHologram.gif
 

The concept of a hogel

 
 
Light from an RGB pixel is collimated, then focused into a Ray (which is really more like a cone)

Light from an RGB pixel is collimated, then focused into a Ray (which is really more like a cone)

Holograms are indeed possible, but require a greater degree of control over light, and a larger amount of much smaller pixels, than normal displays currently support. Normal pixels in a display control colour and intensity, but do not control direction. The light from those pixels emits in all directions. That’s why images that you see on a 2D screen appear flat. Both eyes, no matter where you look on the screen, see the same array of pixel colours and intensities, regardless of the angle from which you view the pixels or the display. But if you can control the colour, intensity, and direction of light from a single point, you can create a hogel (a.k.a. holographic pixel), which is a cluster of directional pixels that can present a different colour and intensity from a single point on the display, based on the viewing angle. Once you have hogels, you can make a holographic display.

AHBG.png

 

Light field display requirements

A high quality light field display requires two things: a lot of hogels, and a lot of rays per hogel. To achieve this, what is first required is a large and very dense array of underlying pixels, which can then be clustered into particular hogel configurations.

HogelGraphic2.jpg

Rays per hogel

HogelRPHB2.jpg
 

To replicate the natural light field you need a lot of rays per hogel, because that’s how light is perceived in the real world.

You need at least one ray per degree, and preferably a lot more. The angular density of these rays is what ultimately determines the depth fidelity of the display (i.e. the depth both into and out of the display surface at which you can clearly focus on content). This is also sometimes known as the Depth-of-Field of the display. You also need a usable Field-of-View (or viewing angle) for the display to create a reasonably large viewing region, which requires a lot of rays, both vertically and horizontally. While different applications have different viewing region requirements, the general rule is “bigger is better”.

 
AHBG2.png

 

Hogel resolution

Once you have that hogel, you need enough of them to form high fidelity images. If you don’t have enough hogels, maybe you have depth, but the spatial resolution of the image itself is too low, and appears blurry or distorted.

HogelRes.png

 The Bandwidth hog

BandwidthHog_07.gif
 

OK, now let’s say you have a lot of rays per hogel, and a lot of hogels. That’s a lot of pixels! How do you control those pixels at reasonable frame rates? Those pixels create a huge bandwidth hog! 

And now in comes his brother, the compute hog!

Next, what happens if you want to interact with real-time content?  Now you’ve got a second pig in the mix - the compute hog! A high quality light field display has enough pixels to drive the equivalent of hundreds or even thousands of 2D displays.  Your standard GPU running classic rasterization or ray tracing algorithms isn’t going to keep up with that.

 
AHBG.png

 Is there a better way?

Breakthrough_03.gif
 

You gotta get those hogs under control! That’s where some secret sauce comes in.

If you put a lot of math, physics, hardware, and software together, you can make a system that works today!