Dawn of VR

In the beginning was the Image, and the Image was with Apple, and Apple made a File Format and called it QuickTime. And Apple saw that it was good. But the Image was flat and sometimes did not fit on the Screen. So Apple drew out of the side of QuickTime a new Track, and called it Virtual Reality. Thus was born QTVR.

Panoramic photography is almost as old as photography itself, and really took off with the invention of flexible film around 1890. By 1920 many professional photographers owned rotating slit cameras that could take an image 150, 180 or even 360 degrees wide on a single piece of film. Those images had a cylindrical projection geometry: true angles horizontally, tangent of the angle vertically; they rendered vertical straight lines as straight, but horizontal ones as sweeping curves. Nevertheless, panoramic photos of groups and landscapes were very popular before WWII.

The idea of panoramic photography took hold in Apple’s Advanced Technology Group around 1991, and by 1995 Apple had officially released software for storing, authoring and displaying panoramas. This ushered panoramic photography into the digital mainstream, and gave it a grand new name: “Virtual Reality”. A community of digital panoramic photographers quickly evolved, many of them loosely organized as the International QTVR Society (later International VR Photography Association).

The most notable feature of the new software was that it could construct seamless panoramic images by combining multiple digitized photographs, a process called “stitching”. At first, this involved scanning film, but as digital cameras improved it became possible to make panoramas purely digitally. Even though the resolution of early digital cameras was low, it was possible to make high resolution panoramas by combining many photos.

Helmut Dersch’s open source PanoTools, first published in 1999, was the real foundation of modern VR photography. Apple’s software was costly and limited, but with PanoTools anyone with a digital camera, a PC, and command-line skills could create high quality panoramic images. Dersch’s C code was both efficient and mathematically correct, and quickly became the basis for new stitching and display applications. Two notable graphical front-ends were Hugin, an open source project initiated by Pablo D’Angelo, and PTGui, a proprietary development by Joost Nieuwenhuijse. Both began around 2003 and are still active. They have evolved far beyond the original PanoTools code base but still owe a great deal to Dersch’s example. PTGui is the stitching tool of choice for most professional VR photographers; Hugin is a favorite with amateurs and experimenters.

VR photographs are meant to be viewed interactively with the aid of a computer; and the best medium for distributing them is the World Wide Web. Already by 2003, panographers were publishing their work on websites such as The World Wide Panorama and Hans Nyberg’s panoramas.dk. These served QTVR files that were viewed by means of web browser extensions called ‘panorama players’. Many of the early players were based on the MacroMedia Flash technology that was then a dominant tool for putting images and graphics on a web page. Two notable packages for publishing VR photos on the web were (and are) Thomas Rauscher’s Pano2VR and Klaus Reinfeld’s krpano. Both began as Flash players for QTVR files and have evolved to use modern web technologies like HTML5 and WebGL to play images stored as many small tiles, which allows highly responsive viewing of even multi-gigapixel images. Apple dropped QuickTime in 2008, and from 2010 did not support Flash in iOS; by 2020 Adobe had officially killed Flash (for users outside China). But Pano2VR and krpano continue as the dominant means of distributing VR content on the Web.

The big idea behind the name ‘Virtual Reality’ is an immersive viewing experience, that makes the user feel present in the scene. 1990s Apple Macs certainly did not provide such an experience. Viewing a stereoscopic spherical panorama on a big modern 3DTV screen is far more impressive, but still falls well short of true immersion. For that you need a head mounted display. HMDs have been built since the 1950s in various forms, mainly for military purposes, and there were sporadic efforts to popularize them in the 1990s. But it was not until 2010 that Palmer Lucky was able to create a viable consumer HMD, the Oculus Rift, based on cell phone components. It cost a lot, it had to be connected to a very expensive PC, the resolution was low, the response a bit slow. But the Rift delivered a real sense of presence. And it created a market that continues to grow.

The price and performance of VR headsets have improved steadily since. Resolution is still below 4K TV standards, but cinema-quality VR headsets are definitely possible. The main factor in when we will have them is whether a large public is willing to pay for an immersive but solitary movie viewing experience. Meanwhile gaming will continue to be the backbone of the VR industry. And maybe games are the new cinema.


Leave a Reply

Your email address will not be published.