The Leith-Upatnieks patent's assignee is Battelle, not a university

What you see

A patent record is not a demonstration; it is a legal description of what the inventors were entitled to claim. Opening the patent document itself shows the invention as text and line drawings, not photographs: figures of beams, plates, and reconstructed “virtual” and “real” images, described in the specification's own words as letting a viewer move their eye around the reconstruction and see “new points coming into view as others go out of view,” so that structures behind the foreground become visible. That description matches an off-axis transmission hologram: a viewer positioned to intercept the reconstructing beam sees a three-dimensional scene with parallax, not a flat photograph, and no headset or screen is described anywhere in the claims.

How it works

The patent, US 3,506,327, titled “Wavefront Reconstruction Using a Coherent Reference Beam,” lists Emmett N. Leith and Juris Upatnieks as inventors, “assignors, by mesne assignments, to The Battelle Development Corporation,” filed April 23, 1964 and patented April 14, 1970, per the document's own printed cover text. Its abstract of disclosure describes directing coherent radiation onto an object to form an object-bearing beam, and directing a second coherent beam as a reference at a finite angle onto the same detector, producing an interference-fringe pattern — a hologram — that is later illuminated with coherent light to reconstruct the image. This is the interference-recording method later associated with off-axis holography, stated here as a filed and granted claim rather than a review's paraphrase.

Viewing conditions and limits

A search of the patent database record confirms the application number, the assignee, and the classification, but a patent specification is not a performance report: it does not state brightness, contrast, or how many people can view the reconstruction at once, and none of those figures should be read into it. The document does specify that a divergent illuminating beam can enlarge the image relative to the recording beam, meaning magnification and image size in this design depend on the geometry the operator chooses at reconstruction, not a fixed spec.

What it is not

This is not the inventors' Journal of the Optical Society of America papers, which report experimental results rather than legal claims, and it is not evidence of a shipped product; a granted patent secures rights to an invention, it does not certify that anyone built or sold a device matching every claim. The assignee is also worth stating plainly: the patent record names Battelle Development Corporation, not the University of Michigan directly, a detail easy to assume incorrectly without opening the document.

• Is the number and grant date being cited for this patent drawn from the patent office record itself, or from a secondary summary?
• Does the patent's assignee match the institution usually credited with the underlying research?
• Do the claims describe an interference-recorded hologram, or a different reconstruction method entirely?

Read this way, the patent complements rather than duplicates the inventors' published papers: it fixes, in dated legal language, the specific apparatus and method they sought to protect, including the assignment chain that a citation of “the Leith-Upatnieks patent” often glosses over.