Holography

Holography is a technique for recording and later reconstructing the complete light field scattered from an object, including both the amplitude (brightness) and the phase (wavefront shape) of the light, so that the reconstructed image reproduces the depth, parallax, and perspective of the original scene. An ordinary photograph records only intensity and yields a flat image; a hologram records an interference pattern that, when illuminated, diffracts light back into a copy of the original wavefront. The recording medium is called a hologram, from the Greek holos (whole) and gramma (message), reflecting that it stores the whole of the optical information.

The method was invented by the Hungarian-British physicist Dennis Gabor, who conceived it in 1947 and published it in Nature in 1948, and who received the Nobel Prize in Physics in 1971 "for his invention and development of the holographic method". In virtual reality (VR) and augmented reality (AR), holography is relevant in two distinct ways. Microscopic holograms (holographic optical elements) are used as the see-through combiner optics inside many AR headsets and smart glasses, and true computer-generated holography is being researched as a display method that can reproduce correct depth cues, which conventional flat displays cannot. The word "holographic" is also widely used in product marketing for effects that are not holograms in the physical sense, a point of frequent confusion discussed below.

Origin and history

Gabor conceived holography at Easter 1947 while working in the Research Laboratory of the British Thomson-Houston Company in Rugby, England. He was trying to improve the resolution of the electron microscope, which was limited by the spherical aberration of its electron lenses, and his idea was to take an electron picture that contained the whole information, phase included, and then correct it by optical means. He called the method wavefront reconstruction: the electron microscope would record the interference pattern between the object wave and a coherent background, the "hologram", which would then be reconstructed with light in an optical system that corrected the aberrations of the electron optics. A British patent, "Improvements in and relating to microscopy", was filed in Gabor's name with British Thomson-Houston as applicant on 17 December 1947 and published in 1952, and the principle was published in Nature in 1948 under the title "A New Microscopic Principle". Gabor's original arrangement, now called in-line holography, placed the object, the recording plate, and the light source on a single axis, because the short coherence length of his light source allowed nothing else. It suffered from the twin-image problem: each object point produces two images, one at the original object position and one at a conjugate position, and in the in-line geometry the two lie on the same axis and can be separated only by focusing, never completely.

Because Gabor had no coherent light source, he used a filtered high-pressure mercury lamp with a roughly 3 micrometre pinhole to obtain spatial coherence, which limited him to holograms about 1 cm across with exposures of a few minutes. Gabor later wrote that holography lay dormant from about 1955 until 1963, when the first successful laser holograms of Emmett Leith and Juris Upatnieks were published; the helium-neon laser they used had a coherence length about 3,000 times that of the mercury lamp. The laser itself had first been reported by Theodore Maiman, using ruby, in 1960.

In 1962 Leith and Upatnieks, of the University of Michigan's Willow Run laboratories, published an off-axis (carrier-frequency) method in the Journal of the Optical Society of America. A second beam, the reference beam, passed around the object and reached the recording plate at an angle, so the two reconstructed images were separated by angle as well as in depth; this removed the twin-image problem of Gabor's in-line scheme. High-quality three-dimensional display holograms followed in 1964: using the laser, Leith and Upatnieks made holograms with diffused illumination and holograms of three-dimensional scenes using reflected light, showed them at the spring 1964 meeting of the Optical Society of America, and reported that the reconstructions kept the parallax between near and distant objects and required the eye to refocus at different depths. Also in 1962, the Soviet physicist Yuri Denisyuk published a reflection method that combined holography with Gabriel Lippmann's technique of colour photography: the object wave and the reference wave enter the emulsion from opposite sides, and the developed hologram reflects the image back when it is lit by the reference wave. Because the recorded layers reflect only a narrow band of wavelengths, such holograms can be viewed in white light, although Denisyuk, who had no laser in 1962, could produce only an "existence proof"; according to Gabor, a two-colour reflection hologram viewable in white light was first made in 1965 by G. W. Stroke and A. Labeyrie. In 1968 Stephen Benton, then completing a doctorate in applied physics at Harvard and also participating in Edwin Land's vision research laboratory at the Polaroid Corporation, created the white-light transmission rainbow hologram, also called the Benton hologram. The technique gives up vertical parallax: in Benton's US patent, filed in 1969 and assigned to Polaroid, a narrow horizontal aperture limits vertical parallax, which matters less because human binocular vision is horizontally oriented, and the viewer looks through a real image of that aperture as if through a window. The image shows a rainbow spectral coloration in place of vertical parallax. Rainbow holograms are the type most often seen on credit cards and magazine covers, and the credit-card form is also known as a "Benton hologram". Benton later directed the MIT Center for Advanced Visual Studies and was a founding faculty member of the MIT Media Lab. Gabor was awarded the Nobel Prize in Physics in 1971 for the invention.

How it works

Holography is an interferometric and diffractive process carried out in two stages, recording and reconstruction.

During recording, coherent light, in practice a laser, is split into two beams. The object beam illuminates the subject and the light it scatters travels to the recording medium. The reference beam is sent directly to the same medium. Where the two beams overlap they interfere, producing a fine pattern of bright and dark fringes whose spacing and shape encode both the amplitude and the phase of the object wave. Interference maxima occur wherever the phases of the two waves agree, so the reference wave acts as the standard against which the phase of the object wave is recorded. The pattern is recorded on a photographic plate, still the most common medium in 1975, or on other materials such as dichromated gelatin, photoresist, thermoplastics or photopolymers. To the naked eye the recorded pattern is unrecognizable; Leith described it as "a hodgepodge of specks, blobs and whorls".

During reconstruction, the developed hologram is illuminated with light similar to the original reference beam. The recorded fringe pattern behaves as a complex diffraction grating and diffracts the illuminating light so as to recreate the original object wavefront. An observer looking into this reconstructed wavefront sees a three-dimensional image that shows parallax and changes with viewing angle, because the eye receives the same light field the object would have produced. This is the property that distinguishes a hologram from a stereoscopic display: a hologram presents a continuous light field, so the eye focuses (accommodates) on points at their true depth, whereas a stereoscopic display shows two flat images on a fixed screen plane.

Computer-generated holography

A hologram does not have to be made by physically interfering two light beams. In computer-generated holography (CGH) the interference pattern is calculated numerically from a digital model of the desired scene and then displayed, most often on a spatial light modulator (SLM), a pixelated device that alters the phase or amplitude of light passing through or reflecting off it. Coherent light reaching the SLM is diffracted by the computed pattern and focused to the object points that form the image, so the displayed image can be changed electronically and animated. CGH is computationally demanding: the most common way to compute phase-only holograms is iterative optimization, such as the Gerchberg-Saxton algorithm, which takes roughly 10 to 100 iterations per hologram. The SLM also limits the optics. The etendue of a holographic projector, the product of its eyebox (exit pupil) area and its field of view, is set by the area of the SLM and its maximum diffraction angle, so unlike a conventional display it is tied to the SLM's pixel resolution, and magnifying the image to widen the field of view shrinks the eyebox. Modern research uses machine-learning methods to compute high-quality holograms in real time and to correct for optical imperfections in the display itself.

Relationship to AR and VR

The term "holographic" appears throughout AR and VR marketing, but it covers two different things that are easy to confuse.

Holographic optics in AR headsets

Many AR headsets and smart glasses use a holographic optical element (HOE) as part of their see-through optics. An HOE is itself a hologram, often a thin volume hologram recorded in a photopolymer, that acts as a wavelength-selective mirror, lens, or grating. In a waveguide display, a pair of such holographic gratings couples light from a small microdisplay into a transparent lens, guides it across the lens by total internal reflection, and then couples it back out toward the eye, so the user sees the projected image superimposed on the real world. Here the hologram is a passive optical component inside the lens; the image the user sees is still a conventional, mostly stereoscopic picture. The Denver-area startup Akonia Holographics, which described its product as volume holographic reflective and waveguide optics for transparent smart-glasses displays, was acquired by Apple in August 2018, a sign of how central holographic optics had become to AR headset development.

The "hologram" misnomer

The virtual objects that headsets such as the Microsoft HoloLens place in the user's environment are commonly called holograms, but they are not holograms in the physical sense. They are produced by real-time head-tracked stereoscopic rendering, the same principle used by VR headsets, in which two flat perspective images are shown to the two eyes. The holographic part of the HoloLens is the volume-hologram waveguide inside its lenses, not the floating image the wearer perceives. Because a stereoscopic image is fixed at one focal plane while the eyes converge on objects at varying apparent distances, such displays produce the vergence-accommodation conflict, a mismatch between where the eyes focus and where they converge that can cause eye strain. True holographic and light field displays are studied partly because they can present correct focus cues and avoid this conflict.

Holographic displays for AR and VR

A separate line of work aims to use real holography, usually computer-generated, as the actual display in a headset, so that virtual content is reconstructed as a true light field with continuous depth. In May 2024 a team at the Stanford Computational Imaging Lab led by Gordon Wetzstein published in Nature a full-colour 3D holographic AR display that pairs inverse-designed metasurface waveguide gratings with an AI-driven CGH algorithm and a phase SLM, fitting the optics into a glasses-like prototype that overlays full-colour three-dimensional images on the real world. The British company VividQ, founded in Cambridge in 2017, develops CGH display software and, with the waveguide maker Dispelix, announced in January 2023 a 3D waveguide combiner intended to show variable-depth holographic content in AR glasses while reducing the vergence-accommodation conflict; the company raised further funding in 2024 and continued to demonstrate the technology with game integrations through 2025. These holographic approaches differ from light field display devices such as the Looking Glass Portrait, which create a glasses-free 3D image from many discrete views rather than by reconstructing a wavefront through diffraction; Looking Glass content is stored as a "quilt" of usually 48 to 100 perspectives of the same scene, depending on the display. Maimone and colleagues describe light field displays as approximating the wavefront coarsely through a limited number of ray bundles, with a diffraction-limited trade-off between spatial and angular resolution, while holographic displays control the wavefront directly.

Current status

As of 2026, optical holograms in the form of holographic optical elements and waveguide combiners are in widespread commercial use inside AR headsets and smart glasses, while full holographic displays for near-eye AR and VR remain largely in research and early prototype stages. The main obstacles are the limited resolution and space-bandwidth product of available spatial light modulators, the heavy computation required to generate holograms at video rates, and the difficulty of delivering a wide field of view and an adequate eyebox in a wearable form factor. Work since 2024 has concentrated on AI-accelerated hologram computation, metasurface and waveguide optics that shrink the hardware, and better colour and occlusion handling, with the goal of a thin, all-day-wearable display that reproduces true depth.