Panoramas

Panoramic perspective refers to capturing or displaying extremely wide-field images of spatial reality.

Panoramic Perspective – Capturing

We can define four ways of capturing a panoramic perspective view/image:

Fixed Viewpoint or single ‘Snapshot’ Panorama (fixed viewpoint): A wide-angle snapshot of 3-D object space is taken from a fixed position and viewing angle, with the eye/camera existing within a visual sphere/cylinder of potential viewing directions. Uni-angular, monocular representation of a spatial scene/object: A 2-D/3-D image/view of a spatial scene/object captured from a single viewpoint or viewing angle. A 2-D linear perspective image of a spatial scene/object projected onto a 2-D surface or picture plane is one example, captured from a fixed viewing angle.
Multiple Viewpoint Panorama (combination of multiple viewpoints): 3-D object space is observed from multiple directions and combined into a single picture. Multi-angular, monocular, or binocular representation of a spatial scene/object: A 2-D/3-D image/view of a spatial scene/object captured from multiple viewing angles, using either a monocular or binocular method. Viewing a hologram image is an example of multi-angular binocular 3-D, employing depth cues such as monocular perspective, focusing plus (partial) natural scaling and variable resolution effects, binocular vergence and parallax, plus 3-D shape changes due to (a narrow range of) multiple viewing angles, etc. In terms of monocular multi-view panoramic images, one example is the work of Dick Termes, who captures multi-directional images from inside a sphere of vision and then projects them onto a sphere that can, in turn, be viewed from different angles.
Unlimited Viewpoint / Motion Panorama (model a spatial scene from any/every angle): Viewing/exploring a digital model (CAD computer modelling), or a Virtual Reality world, etc.
Revolving / Motion Panorama (sphere of revolution about a 3-D object): A 3-D object is scanned from multiple directions using a revolving eye/camera position.

Panoramic Perspective – Observing

We can define four ways to observe a panorama as follows:

Natural View Panoramic Perspective (direct wide-angle view of an expansive scene): A wide-angle perspective view of a spatial scene (e.g. natural scene and visual perspective (2nd type).
Internal View Panoramic Perspective (image on a large-format circular/spherical screen): The representation of objects upon the walls of a rotundo, or circular building, or else circular/ spherical screen, and the spectator either sitting or standing at the centre of the screen, or else moving around the centre of the same. We have two kinds of internal view panoramas: A) Multi-angular view of uni-angular image of spatial scene/object: The term “3-D” is used for a volumetric display that shows uni-angular images of a spatial scene taken from multiple viewing angles (camera changes direction of observation), whereupon the images taken from different angles have been ‘stitched together’, and are notionally viewed in a 3-D space, but without experiencing true observer-based angular perspective changes. Only (uni-angular) perspective depth cues are available (from observer perspective). Example: multiple flat or 2-D images viewed/projected onto a circular screen and viewed internally. Many projection systems have been used to produce panoramas from wide-screen cinema formats such as CircleVision, IMAX and the Sphere theatre in Las Vegas. B) Multi-angular view of multi-angular image of spatial scene/object: The term “3-D” is also used for a volumetric display that generates content viewed from multiple angles, i.e. multi-angular images captured/generated by/for viewing a spatial scene from multiple viewing angles, whereupon the onlooker (may) experience true observer/camera-based multi-angular perspective depth cues (on a directional screen canvas). Example: spatial or 3-D images/views that are projected onto a circular screen that creates an apparent 3-D space, and said views are (apparently) presented/viewed from multiple or changing viewpoints (or viewing angles) depending upon observer/camera position and viewing angle. One example is an LED volume screen used in a virtual production system, allowing in-camera visual effects, realistic lighting and reflections on actors, and enabling immediate, camera-synced perspective shifts (phenomena).
External View Panoramic Perspective (image projected onto the external surface of a sphere): One example is a terrestrial globe of Earth that allows the user to rotate the view.
Flat View Panoramic Perspective (spherical / cylindrical / curvilinear perspective): Any wide-angle perspective image, of any type, displayed on a flat picture plane.

Expanding the Visual Field

A fundamental feature of all perspective methods/instruments/systems relates to the concept of field of view (FOV), which, for an outward-looking optical instrument, is a measure of the proportion of the total environment (180°/ 360°) captured in a single perspective image taken from a particular viewpoint. Often, we relate such images to the natural monocular FOV of the human eye, known to be around 135 degrees in the horizontal direction (HFOV), and 180 degrees vertically (VFOV).

Over the centuries, there have been attempts to go beyond these limitations in natural vision, specifically to facilitate viewing, matching/ modelling and representation of a broader ‘captured’ spatial reality. Artists have explored cylindrical and spherical surfaces to depict their paintings across a wider FOV. In the 20th-century, photography has contributed to these developments. A wide range of lenses was produced, from extreme wide angle (84°-179°) to wide angle (63°-83°) to normal (34°-62°), etc.

Wide-field developments occurred on three fronts: fisheye, multi-camera (polydioptric), and catadioptric cameras. In the open air, a fisheye lens can capture up to 180 degrees. Alternatively, using another second method with multiple cameras in a dome, we can compress spatial reality to capture/represent up to 360 degrees for a total view of the environment. Simple versions typically use symmetrical numbers, such as 2, 4, or 8 cameras, resulting in cylindrical panoramic images. These methods have been complemented by spherical configurations such as the Google Street View and related cameras, leading to ever more dramatic omnidirectional perspective images.

The third method involved catadioptric lenses, which typically use mirrors: parabolic, hyperbolic, elliptical, or planar. These principles are used to capture/ project omnidirectional perspective images, which can subsequently be translated into cylindrical/spherical panoramic images (for consumption / display).

Spherical Projection

In addition to the original scene’s shape (captured visual field) and the recorded scene’s shape (displayed visual image), there is the question of the projected surface (or display surface), being a shallow/deep concave bowl, or a hemispherical or full spherical dome, which can in turn use its interior or exterior surface.

Television faced similar issues, with larger, higher-resolution screens becoming a trend. On the one hand, there are ever more dramatic 8K, 10K, and even 18K resolution screens (Sphere Theatre, Las Vegas). On the other hand, new kinds of stereoscopic screens are being introduced, in which technology expands the ‘local’ viewing field for 3-D objects (ref. sphere of revolution perspective). In sum, panoramic perspectives are set to remain popular, as we seek to match/expand the natural visual field.