Mastering Digital Panoramic Photography

A History of Panoramic Image Creation

Adapted from Mastering Digital Panoramic Photography (Rocky Nook)

By Harald Woeste



The First Panoramas

The word “panorama” is derived from the Greek words pân (= everything) and hòrama (= to see, that which is seen, the appearance, the view), and thus stands for the all-around view. Panorama images can be created in a variety of ways, from the first round paintings in the 18th and 19th centuries, to the first photographic panoramas in the middle of the 19th century, right up to modern, digital panoramas created using computer-based image processing techniques.

Historical Development

Panoramas first became known when in 1787, Robert Barker (an Irishman) patented his plans for a cylindrical building that was to be erected around a large, panoramic painting. The building served the exclusive purpose of allowing the cylindrical painting to be created and subsequently displayed. The special painting technique, combined with sophisticated lighting, produced a new experience for the viewer, who stood on a special platform in the center of the circular room. The goal was to produce the perfect illusion of a real scene.

Barker erected his first permanent circular panorama building (or “rotunda”) in Leicester Square in London in 1792, where one or two new panoramas a year were exhibited for the next half-century.



Panorama etching, after the first great panorama painting of London by Henry
Aston Barker, 1792 Source: Oetterman 1980: Leporello 1 (click to enlarge).


These panoramas showed attractive landscapes, famous battles, or important cities. The new art form was a source of great excitement and became hugely popular, breeding many imitators not only in England, but also in Holland, Denmark, the USA, France, Switzerland, and Germany, where Johann Adam Breysig and his colleagues painted the first known German panorama in 1800 for exhibition in Berlin.

These circular images (or better, circular paintings—the images were at this time still painted on canvas) measured up to 15 meters in height and were often as long as 100 meters. In order to reinforce the illusion that the observer was an integral part of the scene, the panorama exhibitors began (around 1830) to experiment with placing three-dimensional objects and other props in front of their images, turning panoramas into a modern, commercial form of entertainment. Panoramas became a traveling spectacle, supported and transported from city to city by a well-organized industry.



Diagram of a Rotunda
A: Entrance and cash desk; B: Darkened corridor; C: Viewing platform; D: Observer’s field
of view; E: Cylindrical canvas; F: Three-dimensional foreground (Faux Terrain); G: Trompe
l’oeil objects painted on the canvas


Panoramas as a mass-media phenomenon had their heyday in the middle of the 19th century: no other art form or medium was as popular and had such a radical effect on the way the public viewed the world. Film and television did not yet exist, and photography was in its infancy. Very few people were able to indulge in educational trips, and tourism as we know it today was unheard of. The opportunities for the general public to gain an impression of foreign countries, people, cities, landscapes, events, or exotic animals were extremely limited. Panoramas made it possible for the citizens of the world’s big cities to see and experience such things. Not only were the panorama pictures impressive, but the opening ceremonies themselves became important social events, accompanied by music and speeches.

Panoramas in Contemporary Art

The success of Yadegar Assisi’s 2003 Everest panorama in Leipzig provides us today with a hint of the magic of those first 18th and 19th century panorama exhibitions. Assisi exhibited a huge collage consisting of 3D models, detail photos, and panorama photos of the Himalayas in honor of the 50th anniversary of the first scaling of Mount Everest. The 36-meter-high exhibit—along with atmospheric background music—produced the illusion that the viewer was standing at the foot of the tallest mountain in the world. The largest panorama in the world remained on display inside an unused gas storage tank for two years.



Everest 8848
Panorama installation by Yadegar Assisi celebrating the 50th anniversary of the first scaling of
Mount Everest. Exhibited inside a gas storage tank in Leipzig, with multimedia light and sound effects.



Everest 8848
The lower viewing level.


In 2005, the “Gruppe 180” artists exhibited a painted 180° panorama portraying Berlin’s famous Brandenburg Gate at the end of World War II. A viewer standing in the center of the 28 × 5 meter installation had not only a strong sense of space, but also the feeling of being transported back in time. The transitions between the artificial, historical image and the real, immediate environment were all but undetectable, and the vertical and converging lines within the image coincided perfectly with those of the surrounding contemporary architecture. Marcel Backhaus’ panorama effectively combined conflicting elements of past and present, destruction and reconstruction, and war and peace.







Panorama installation at the Brandenburg Gate, Berlin, Germany.


Wide Images and Wide Angles—Photographic Panoramas

If we had to describe what is special about panoramas using one short phrase, we would have to say, “field of view”. This has always been true in the the world of painting, and it is equally valid when we are creating photographic panoramas. The most important characteristic of a panorama is the large field of view. Since the medium’s earliest beginnings, inventive photographers have searched for ways to capture ever wider angles of view.

Unfortunately, the expression “panorama” is not always used consistently in a photographic context. Although the word was originally coined to signify an all-around 360° view, it has come to be used to describe images with wide formats—for example, with aspect ratios of 1 : 2 or 1 : 3. Although most panoramas have wide formats, they are not necessarily shot using wide angles of view. A wide-format panorama can be equally well shot using a telephoto or a wide-angle lens. The advent of digital photography has, however, helped us return to the historical definition of panoramas as all-around images, or images with a large field of view.



Comparison of Fields of View.


Wide angles of view are produced in the digital photographic world by using special cameras or specific shooting and image merging techniques. Panoramic angles of view are typically much wider than the normal (approximately 50° horizontal) angle of view of the human eye, and extend through the 140° secondary angle of view of the (moving) human eye right up to 360° all-around views. In the real world, a viewer can only observe a 360° angle of view by physically turning around his/her own axis.

Just in case this all sounds a little confusing: the viewer can see an entire panorama image at once if the image is small enough and if it is displayed on a flat plane in front of the viewer’s eye, like a photographic print.

Freedom of Choice: Wide Images—Straight or Curvy

Wide-format images set themselves apart through their dimensions, rather than through any particular aesthetic characteristics or unusual production techniques. Such images are often shot using specially constructed cameras, a similar result simply by cropping the top and bottom edges of a conventional photo. Cropping an image taken using a 6×6 Hasselblad produces the same basic result as using the specialized Hasselblad XPan panorama camera. Similarly, you can achieve results equivalent to those from a 6×12 or 6×17 panorama camera by cropping large format images. The digital photographic medium uses electronic image sensors to capture images. Most digital images can be cropped to wide/panorama formats without noticeable loss of image quality, thanks to ever-increasing sensor resolution.

Truly wide-angle panoramic images are characterized less by their wide format than by the large field of view that they cover and their own, very specific curvilinear appearance when displayed as a flat image. These all-around images can only be produced (whether in analog or digital form) using specially constructed cameras or dedicated software—a situation which is not likely to change in the foreseeable future.

The flexibility of modern, digital panorama creation processes means that we (fortunately) no longer need to construct specially designed buildings or oversized canvasses to display our images. These days, we can use the same source material to construct a wide-format, a wide-angle, or an interactive panoramic image for display on a computer monitor.



Wide-Angle
All-around view (cylindrical projection) of the above shot, cropped to the same horizontal
angle of view, and showing curvilinear distortion.



Wide-Format
A conventional architectural shot, where straight lines are reproduced straight (rectilinear projection).


The Challenge of Shooting Effective Panoramas

The major aim, and simultaneously the greatest challenge, for every panorama photographer is capturing a wide field of view. In other words: how can I photograph something that cannot be photographed, and how do I make a useful image from the resulting raw material?

The easiest way to understand how panoramas are made is to use three theoretical models of the various ways to make a conventional photographic image in one’s immediate environment.

Model 1—A Sheet of Glass: Set up a (theoretical) sheet of glass at arm’s length in front of you and paint exactly what you see through the glass. This model represents the taking of a conventional photo, and the resulting image is described as flat (also “planar” or “rectilinear”). If you wish to paint more of your surroundings than the sheet of glass allows (i.e., to increase the angle of view), you need to increase the size of the sheet of glass. Herein lies the first problem: the viewpoints of the painter and the sheet of glass are fixed, and no matter how large your sheet of glass, you can never see objects positioned behind you. Practically speaking, the limits to conventional viewing lie at around 120° (both vertically and horizontally). This limit also generally applies to conventional photographic techniques.



Model 2—A Glass Cylinder: Position yourself in the center of a glass cylinder with a radius of an arm’s length and paint what you see through the glass on the cylinder’s inner surface. You turn on your own axis and paint everything that can be seen in all directions, including objects originally positioned behind you. The angle of view here is 360°. This type of image is described as “cylindrical” and is similar to the first panorama images that were painted and displayed in specially constructed circular buildings. The problem here is that you cannot capture any objects positioned either above or below you, as the cylinder has a limited height. As with our first model, the practical limit of this model lies at a vertical angle of view of approximately 120°. Our next model solves this problem.

Model 3—A Glass Sphere: Position yourself in the center of a glass sphere with a radius of an arm’s length and paint what you see through the glass on the inner surface of the sphere. If you turn around the center point of the sphere, you can not only paint everything around you as you did for the cylinder model, but you can also see (and paint) objects above or below the limits of the cylinder’s height. You can thus cover a horizontal angle of view of 360° and a vertical angle of view of 180°– in other words, everything surrounding you in every direction! This is the ultimate panorama, and the resulting image is described as “spherical”.

What is the significance of all this to panorama photography? The key question is what you want to photograph, or more accurately, which field of view you want and how it can be subsequently displayed. As described, a flat surface cannot display objects that lie behind you, and a cylinder cannot display objects that lie above or below you. A sphere is the only type of surface that can display your entire surroundings. What camera can take such photos?

Conventional cameras produce simple, plane images on a piece of flat film. For all practical purposes, the angle of view is limited to approximately 120°. Revolving cameras (i.e., those that rotate around their own axis during the shot) produce 360° images on film that is loaded into the camera in cylindrical form. In order to shoot an image with a 360° horizontal view and a 180° vertical view, you would need a spherical film (i.e., a light-sensitive ball) with a lens somehow built around it. Such a device simply does not exist.

Contemporary panorama photography has solved this problem digitally, making it possible to take a sequence of images from a single viewpoint that can subsequently be combined into a single, virtual, spherical image using a computer. But how can we transform this “virtual, spherical negative” into a viewable image?

Producing a flat image from a flat negative for display on a wall or in a magazine is simple, and a cylinder can also be cut open easily and rolled flat. It is, however, not so simple to transform a spherical negative into a flat image. You cannot simply cut it open and spread it out. If you had a rubber ball film, you could cut it open vertically and then stretch its corners until you had a flat, rectangular image. This type of transformation is nowadays made mathematically with the help of a computer. We will address the subject of projection types and transformation in detail in the section dealing with image output.

Panorama Cameras Then and Now

The first photographic panoramas were made not long after the invention of photography itself in 1839. They were shot using normal cameras, panning in order to capture sections of the subject that could not be shot at once due to the limited viewing angle of the lenses. The resulting prints were then appropriately cropped and aligned in sequence. Most of us have seen such “cut-and-paste” panoramas in our grandparents’ photo albums. The individual images often show differing brightness and color reproduction, and often have different magnifications. These differences make the joins between the individual images very obvious.

It wasn’t long after the invention of photography that the first specialized panorama cameras were constructed, with the aim of capturing increased angles of view in a single image. This was initially accomplished simply by exposing a particularly long strip of film. An Austrian, Josef Puchberger, patented a camera with a rotating lens moved by a hand crank in 1843. This camera covered an angle of view of 150°, and panorama photography has been in constant development ever since.

The difference between wide formats and wide angles described above divides cameras used for shooting panoramas into two basic groups. With regard to light incidence, perspective, and handling, cameras with fixed lenses are fundamentally different from rotating cameras or cameras with rotating lenses.

Panorama cameras with fixed lenses produce wide images with characteristics similar to human visual perception. Thanks to interchangeable lens technology, angles of view of up to 100° can be covered easily. The appeal of this type of camera lies less in the maximum field of view and more in the wideformat images produced.



Digital Panorama Image
Multiple source images are joined together to produce an all-around view.


While conventional cameras with fixed lenses produce familiar-looking, flat images in which straight lines remain straight, rotating cameras (or cameras with rotating lenses) produce cylindrical images in which straight lines appear curved in the final printed image. This gives such images their characteristic distorted, curvilinear look.

What Do I Need?
Necessary equipment for shooting digital panoramas

  • Camera
  • Wide-angle or fisheye lens
  • Tripod
  • VR panoramic tripod head
  • Spirit level
  • Computer
  • Stitching software
  • Image processing software

Rotating panorama camera lenses pan laterally during the shot and typically cover an angle of view of about 140°. This represents about the same view that the human eye can cover if you hold your head still and move your eyes from side to side, resulting in a natural- looking field of view, but with unnatural distortion of straight lines.

Rotating cameras rotate around their own axis and are capable of producing 360° cylindrical views. The general look of the images they produce is similar to those produced by cameras with rotating lenses, with some distortion of straight lines but with a more unusual 360° field of view.

The recent digitization of many photographic processes has led to especially rapid development in the field of 360° panoramas. Special rotating digital cameras capture images row by row using a process similar to a conventional scanner. In addition, computers can be used to simulate the cylindrical, slightly distorted images produced by rotating cameras or cameras with rotating lenses using source material produced by a conventional digital or analog camera. Specialized software “stitches” sequences of individual images together to form panoramas. Ideally, stitched panoramas are not identifiable as such and appear to have been taken in one shot. Used creatively, a computer can help you produce stitched images that are technically superior to conventionally produced images, and which have an aesthetically enhanced appearance. Stitching makes it possible to shoot not only vertically limited cylindrical images, but also 360° × 180° spherical images. This was impossible to achieve using analog photographic techniques, as even the most extreme fisheye lens always includes the tripod in its field of view.

Even though the original aim of panorama photography was to produce images with large angles of view, stitching techniques are also used to produce conventional-looking images using mulitiple source images. Landscape photographers often use this technique to produce high-resolution, wide-format photos that would otherwise require the use of a special panorama camera or a large-format camera. Scientists also use similar techniques to produce ultra-high resolution documentary images, such as the 16-gigapixel (172,181 × 93,611 pixel) image of Leonardo da Vinci’s “Last Supper”, constructed from 1,677 separate images (viewable at haltadefinizione.com).

All-Around Digital—Shooting Panoramas Using a Conventional Camera and a VR Panoramic Tripod Head

Panorama photos and digital technology complement each other very well and provide us with previously unimaginable creative flexibility. The creation of some types of panoramas is only possible since the advent of digital image processing. While rotating digital cameras produce the necessary shot (or, more accurately, image file) in one go, they are nevertheless highly specialized, inflexible, and very expensive. Using conventional cameras with specialized accessories to produce similar results is becoming increasingly simple. A digital compact camera or DSLR used together with a Virtual Reality tripod head (“VR head”) and appropriate software can produce results comparable with those from expensive, rotating cameras. The only disadvantage is that this approach involves a more complex workflow with many more individual production steps. This technique is very versatile and relatively inexpensive (although there is no limit to the amount you can spend on equipment, should you want to), and is currently the most popular way to produce high-quality digital panoramas.



Digital Panorama Image
Overview of camera types and the images they produce.



Don‘t miss the next Insight article on Graphics.com. Get the free Graphics.com newsletter in your mailbox each week. Click here to subscribe.


Printed with permission from Rocky Nook. Copyright 2009. Mastering Digital Panoramic Photography by Harald Woeste.