Forget 32K, are you ready for light field displays?

This is a shot taken looking at a Creal3D Light Field display headset. The focus differential here is real as the camera taking the shot saw it, not burned into the footage

It will eventually be one of the most transformative display technologies since the invention of the television. The race is currently on to produce the world's first practical high resolution Light Field display.

Adrian, please forgive me for hijacking the title of your recent article: “Forget 8K, are you ready for 32K?”, but there is more to the advancements coming in display technology than you had time to delve into in any depth. I am referring to the light field display which offers the promise to create natural 3D viewing without eyewear or any of the drawbacks of stereoscopic or auto-stereoscopic display solutions.

So, what exactly is a light field display? A true light field display is kind of like a simplified real hologram (more on the use of the term hologram later). In the ideal case, it is very close to seeing a person, object or scene as your eyes and brain would perceive them naturally with all the subtle nuances and depth cues we normally experience. As you move your head or body around the display, the perspective changes with new textures and reflections coming from each light point in accordance with the angle it is viewed from. That means each pixel not only must display luminance and colour, but it must display it differently in each angular direction – left/right and up/down. Think about that. Doing this with high density, so you have a smooth continuous image in all directions, is incredibly hard but would come close to simulating how we would naturally perceive the image.

Are we there yet? No, but we are making good progress. In fact, an entire event focusing on the Light Field display and the full end-to-end ecosystem is being organised on Oct. 8-9 at the CableLabs facility in Colorado (2019 Light Field and Holographic Display Summit).

Light Field Processing

In addition to Volumetric capture and light field display solutions, the event will also take a close look at Light Field processing. Today, we have two very different types of content. Physical camera-based content for TV, theatrical and other uses captures images with a grid of sensors that is conventionally displayed in a grid of pixels.

Game-based content is computer generated by building a 3D model of the characters, objects and scenes that are then represented as a series of primitive elements with textures and other properties assigned to each primitive. Sometimes photographic elements can be assigned to primitives, but the fundamental way the image is represented is different. To see the image, a virtual camera renders out a conventional 2D or stereoscopic 3D view from a particular perspective. One can move around in this 3D world, but these primitives do not have the full angular attributes that a light field image would. New techniques like ray-tracing are a great step in this direction, however.

These philosophies of content creation, representation and distribution are being fiercely debated today with proponents on each side arguing their path is the best to ultimately delivering the light field image to end-users.

Finally, we need to understand that many displays are being called “holographic” or “light field” when in fact they are neither. It is still early days for true light field display solutions, but there is also a lot of work underway directly in this area and/or in related fields and applications which will help advance the field. And 32K pixels may look like a starting point in terms of resolution if you count all the angular variations each pixel must deliver.