Reality and Realities
A Brief Flight Through the Artificial Landscape of the Virtual Worlds

Olev Koop

Back to Table of Contents
In the years to come it is likely that the twentieth century may be designated the golden era of cartography. The field once before (in the sixteenth century) experienced what amounted to a renaissance and reformation from the traditions originally established by the Greeks. The current rapid development may be a second reformation.
Arthur H. Robinson in The Look of Maps, 1952

How real is reality?

Here we are, 1995, in the middle of one of the hottest summers since the beginning of meteorological recording in Europe, sitting in the reality of a hot lecture room in Madrid, Spain. Listening to a guy who is trying to tell us this is not at all real, what we are experiencing, but an illusion of the mind. Reality is in the eye of the beholder. As far as reality is concerned it is as artificial as the reality of the movie screen. What we sense around us is part of the psychological and cognitive process of the human awareness. What a Dutchman considers to be extremely hot is quite normal to the Madrilenean who takes an afternoon siesta for granted. So the perception of our environment determines what is real to us and what is not. Actually man tries to allude its peers by tampering with the psychology of cognition. In architecture and industrial design it is common practice to use optical illusions and clever use of shape and colour to create a specific perception of space. Is he or she experiencing reality? You probably would answer this question negatively, but argue that nature must be an objective reality to the human spectator, since humans did not (in principle) interfere with it. This is not at all true. Even the experience of nature is very subjective. The British historian Simon Schama (1995) recently published a book called Landscape and Memory in which he treats the way mankind has dealt with its perception of its natural environment throughout history. He describes the way that art, literature and political propaganda used, or perhaps abused, nature to create metaphors that induced certain emotions with the viewer, reader or spectator. These illusions may be found back in modern advertisements and TV commercials where the vendor tries to sell us chocolate candy bars by presenting us with images of ever waving palms, white beaches and spectacular surf (obviously ignoring the hardship of the people usually living in such places).
If people walk through an exhibition of 19th century romantic paintings of Arcadian landscapes they immediately know they are fooled, but nevertheless they are drawn into these scenes by the secret forces of escapism. Actually the same motivation drives people who enjoy the outdoors by going into nature for leisure. They go well equipped and pretend that they are in primeval times again. They experience nature from a romantic point of view. When people have to live permanently in this same environment a completely other perception will arise. Nature becomes one's friend and enemy alike. It provides and it takes. One must immerse, become part of it, to survive. Only then you will experience its reality.
Cartographers have always tried to mimic this awareness of one's environment by using the semiology of graphics (bertin, 1967 ) in order to recreate or, perhaps better, approach this sense of reality with the map reader. Of course in cartography the depicted reality is not always tangible. Especially in thematic cartography many or even most map images represent abstract themes, usually of a quantitative nature that lure the reader into a virtual space in which the abstract numbers become part of a spatial realm of symbols. Everyone knows that this realm will never present the undisputed truth but that it provides a subjective view created by the masterful hands of the cartographer. White lies, or discrepancies introduced deliberately into the map to enhance the cognitive properties, are common practice (monmonier, 1991 ). The cartographer may create hundreds of different views of the same data creating ever so many virtual images, each of which forming part of an endless puzzle that will yield 'reality'. This notion of multiperspective mapping has become more relevant since the advent of instant cartography through the availability of geographical information systems and desktop mapping software (muehrcke, 1991 ). In a sense this notion elaborates on the ideas of cartographic communication, like the famous model of Kolacny (1970) in which a perfect state of communication is intended by overlapping the reality of the cartographer with that of the induced image of reality in the map reader's mind. Since map maker and map user have become one in many cases the nature of this communication changed. Involvement of the cartographer has become similar to that of a native guide who actually does the exploring, although the explorer takes all the credit. The task of the cartographer is shifting much more towards (co-)designing user interfaces and introducing intrinsic expert knowledge than designing the actual map (or mapoid) product. The software that handles spatial data is becoming ever more interactive and transparent in its use. The user may explore the digital equivalent of reality in a intuitive fashion not constrained by the technology. Multimedia and animation technology allow us to shape this 'virtual' space according to any prerequisite.

The third reformation

Under the influence of the technology push of the second world war cartography in general and map production in particular underwent a revolutionary change. These innovations formed the fertile soil in which new scientific notions could sprout. In the first paragraph of "The Look of Maps" with which this paper was started, Robinson (1952) calls this the second reformation of cartography. Currently we have reached a point where we could start referring to a third reformation in cartography. Where the second reformation revolutionised map production, keeping the cartographic medium (paper) the same, the third brings the virtual map that primarily serves the use on-screen and is supported by relatively complex user interaction allowing the user to take an active part in the map making process. As mentioned before, although the interaction is technically complex, the user should be guided by intuition through these digital products. Since the user is promoted to map producer this third wave of innovation is part of a process that may be identified throughout society, being the democratisation of information dissemination. Although very profitable to the users of cartographic materials these development pose the cartographers with serious problems. With the increase in technological complexity of cartographic products the need for more profound technological knowledge is becoming more and more apparent. One needs knowledge of technical fields that were previously the realm of graphics artist, animators, sound experts, user interface designers and so on. There is not only the matter of technological advancement, but also a theoretical problem since the communicative nature of the new media products differs substantially from the most successful product so far, the paper map. New research is required to acquire the sort of knowledge cartography typically needs: the knowledge of cartographic communication.
This seminar deals with the issue of animation in cartography. Although Thrower (1959) already recognised the power of the animated maps the cartographic animations remained part of television productions. Only recently since the advent of multi media technology the animated map reached the desktop of regular cartographic editors. Since 1959 only a few cartographers have studied animations in cartography. In 1966 Cornwell and Robinson were the first to experiment with computer animation, by photographing images of a cathode ray tube and sequencing them. Tobler (1970) investigated the possibilities of animating 3D thematic maps. Moellering (1984) attributed substantially to the theory of cartographic computer animation. Examples of recent articles on this subject are from Mackaness and Buttenfield (1992) about the dynamic visualisation of statistical data within a GIS; Dorling (1992) who researches the visualisation of geographic patterns through animation; Fairbairn and Deely (1991) describe the suitability of an animation package for cartographic illustration; Koussoulakou and Kraak (1992) discuss the communicative aspects of spatiotemporal maps, in other words, of animations.
So far most research in cartographic animation focused on the visualisation of spatiotemporal themes, usually in a two dimensional format. Topographical oriented animations are much rarer then one would think. Strange enough the computer game industry must be thanked for providing us the option of animating space itself. Flight simulators are the best known games that apply this type of spatial interaction. The most intriguing sort of computer animated environment is brought to us by the technology of what is fanciful called "virtual reality".

Canned Reality

In the introductory section of this paper we saw that reality as such does not exist but that several realities exist. To capture reality we may use maps, photographs and films. One aspect of experiencing reality is the notion of three dimensional space and time. Moving through space takes time and therefor these two aspects are tightly linked. The best medium to capture these aspects is the film or motion picture, possibly enhanced with some sort of 3D (stereo)imaging system. A major drawback of this approach is the orchestrated nature of the movement. The viewer is not able to change the path to be followed through the captured or, perhaps even better, canned space. Computer technology may improve on this limitation of the motion picture. A digital motion picture may be branched by user interaction, since the image frames do not necessarily have to be played in pure sequence. In Hollywood terms, the spectator may choose between a happy, sad or perhaps violent end. Movies that contain graphically explicit material could be down rated for a younger audience by electronically re-sequencing the movie. In the world of cartography one of the first examples of such a presentation of Space was found in the "Aspen Movie Map" which was developed in the late 70's by the US Ministry of Defence to enable the military to prepare and exercise movement through a town in a simulated environment. In the application the streets of Aspen, Colorado, were recorded on video. The frames were stored on a laser disk as image frames. The required video sequences were retrieved by a computer after the user pushed some buttons (left, right, ahead and stop). This sort of navigation through virtual space may be found currently in video games, where one may move through a fantasy landscape in a similar way as the Movie map (MYST) or by real time processing of a virtual environment (MARATHON). Virtual reality however is generally concerned to be more than sitting behind a screen moving a mouse. It requires total isolation of your environment and your user interface. Immersive virtual reality requires a complete absorption of the senses into the virtual environment. By using a so-called 'head mounted display' that feeds visual and audio stimuli to the user. A walking pad allows one to move one's whole body in the virtual environment. All interaction may be presented to the user in this way. Although not a scientific source the famous novelist Michael Chrighton (1994 ) describes in his novel "Disclosure" a fictional user interface to a corporate computer database in which the various (distributed) users walk through the virtual archives of the company. Files are stored in virtual file cabinets and may be retrieved by using a virtual hand to open the virtual cabinet drawer, picking out a folder, taking a document and read it. Although this story is purely fictional it is all technically reproducible with current technology. The technology is still very costly however and one could wonder if the practical implications of this type of interfacing would constrain further development.
Cartographers are interested in the way the technology of creating fictional landscapes and recreating existing ones could help in designing new products. It could be very advantageous to have a look at the topography from another angle than the orthogonal topographic map. In many cases the landscape users would like to look at does not exist yet, so an artificial oblique view may help them to get an impression of how a particular area may look like in the future. Animation, or to be more specific, flying or driving through the landscape improves the way we experience the landscape. In the next section some products will be described that may bring options to create such landscapes relatively easily.

Creating an artificial landscape

It is not the intention to discuss the theoretical and technical issues surrounding the creation of three dimensional models of the landscape, also known as Digital Terrain Models. For an elaborate description of these topics the reader is refered to existing literature, such as the monograph of Kumler (1994) or the dissertation of Kraak (1988) . Here, focus will be on the design issues and ease of use of the packages. A list of wished-for future features of the software that could benefit cartographic design will be added. The following packages will be discussed:
Landscape rendering packages:
Although all four packages were run and demonstrated during the seminar on an Apple Power Macintosh, they are also available for the PC/Windows platform. Running them on the fastest machine available is recommendable since processing times of movies are very long (hours).

Non-immersive virtual reality

This type of virtual reality is the type where one may walk through virtual space with the aid of the usual computer interface items: mouse, keyboard and screen. No special attire like helmets or gloves is needed. The idea behind these programmes is that it enables the users to build their own space with means of design software. Within VR this may be done in a design window. Rooms may be constructed and furniture may be placed. The objects are picked from a 3D gallery of objects. Any time one may switch to the walk through window, to actually look around in virtual space. To get smooth movement through space one needs definitely Power PC speed, otherwise movement is very slow. The objects, as may be seen in the picture, are very primitive polygon based. For serious design this technology is not yet capable of building more complex environments, but it promises much for the future.

Landscape rendering packages

For this contribution to the seminar most effort was put into the packages with which one can recreate landscapes, fictional as well as real. Besides building still pictures two of the three evaluated package are also capable of creating movies that simulate an actual "fly-through" or "drive-through" of a landscape. Before proceeding some principles of these packages must be clarified.

The DEM forms the basis of the virtual landscape. It consists of a regular grid of cells. The values are translated into grayscale values, which are interpreted by all packages as differences in height. The actual height value in this respect is irrelevant. VistaPro and Scenery Animator ask for the lowest and highest altitude on import of the DEM. The fact that grayscales will form a 3D landscape implies that every grayscale image may be used as source (eg. of KPT Bryce), even any photograph. This gives weird effects however. The relief may be exaggerated or smoothed afterwards. Bryce has a wide array of operators to alter the appearance of the relief (left in the picture).
In the United States USGS DEM files are distributed in the public domain. One may download the files from an FTP-site (edcftp.cr.usgs.gov) or order them on a magnetic medium. With VistaPro many more interesting DEMs are included on a CD ROM. Unfortunately one may only display one tile of the DEM at the time. In most other countries these data have to be bought from the national land surveys. In all three packages the user can generate a fictitious fractal landscape when no DEMs are at hand

Rendering still images

In order to render an oblique image of a landscape a location in the virtual space has to be picked, where we would like to place our camera. Furthermore the target at which to aim the camera has to be defined. As with a real camera, the virtual camera has a field of view, the camera angle. These three parameters determine the orientation content of the picture. Before proceeding, various other parameters that will determine the look of the picture have to be defined. First size/resolution will determine the size of the image on screen. Quality and texture determine the amount of "realism" put in the image. Other switches are also meant for image improvement. The better the required image, the slower the rendering. The dialog from VistaPro below shows that this package also enables stereoscopic images. Now the picture can be rendered.

The picture can be made even more realistic by putting clouds and trees in it. In principle these items are placed at random although the density and size may be varied. Only VistaPro allows placing trees (and blocks that should represent buildings) on particular places. This design process is very primitive and coarse however. An important factor is the light source, the sun, of which angle and azimuth may be set.
Rendering takes between 30 seconds and 30 minutes, depending on the package used and the required quality and image content. Trees are very computationally intensive, because no tree is the same. These packages use fractal enhancement techniques extensively.

The major drawback in creating these images is the limited or lacking design options. Except for the DEM (you may alter DEM's in image editing packages like Adobe Photoshop) and some primitive editing in VistaPro the user can not alter the landscape furniture in a specific way. Although Bryce allows us to drape a raster image on top of the DEM, this package does not allow fly-throughs and the draped images are always flat. Our example shows the drape of a satellite image on top of a DEM of the Luberon area in southern France. It would be nice if the user could import objects from 3D programmes, like buildings and cars.

Building Fly-Through Movies

In the beginning of this paper the multidimensional nature of cartography was discussed. When building a fly-through this may be practiced in a very literary sense. A path can be created through virtual space along which the camera will move. Only VistaPro and Scenery Animator have this feature. The packages work with key frames that define a particular state of orientation. The frames in-between change orientation gradually. So going from a wide-angle towards a telephoto field of view goes smoothly. This is one of the nicest features of these programmes and makes this flight design very easy.
Still images are controlled by camera location, target and field of view. In a flight or drive there are some more parameters:
Pitch
This is the view angle in relation to the horizon. This is usually also an angle to the flight path. It gives the opportunity to look up or down during the flight.
Bank
This is an aeronautical term meaning the tilting of the airplane during a turn. In a right turn the plane banks right, in a left turn left. This results in a tilt of the horizon in the opposite direction.
Altitude
During a flight one may alter the altitude of the camera. Both packages use a profile line for this in which the key frames can be moved up or down. Subsequently the in-between frames are gradually changing altitude, creating a smooth ride.

Rendering of a flight
This is the most elaborate part of the design process. First of the number of frames to be include in the picture must be determined. This number depends on the complexity of the path and the relative speed of moving. Both these parameters determine the length of the movie in seconds. On the current machines 15 frames per second guarantee a smooth running movie. It is not difficult to calculate that a 30 second move requires 450 frames. So when each frame requires a two minute calculation one ends up with 15 hours of number crunching (and we're talking 80Mhz PowerMacintosh here, don't even think of trying it on a 68K Mac). VistaPro renders faster than Scenery Animator.
The quality of the flight is determined by the options to modify the key frames. Looking at key frames only, Scenery Animator does a better job. The advantage of VistaPro is the pre-set flight specifications, that also include drive specs (pre-sets are Jet, Cruise Missile, Glider, Dune Buggy and Motorcycle).
Both packages support playback of the movies through the Apple Quicktime system. These so called Quicktime movies have a universal format that may be read by all Quicktime compatible software, such as Macromedia Director and SuperCard. It is very unfortunate that this printed medium does not allow us to include a Quicktime movie as an illustration. There will be an example to download in the electronic version of these proceedings on the Internet (http: //nvkserver.frw.ruu.nl/nvk/ica/madridproc.html).

Conclusions

Flythroughs are about reality, realities, virtual reality and illusions. What's the differenc?. When creating a fly-through and putting trees in at random where heavy urbanisation takes place, a fictitious landscape is created, although the relief does actually exist. In order to make these packages useful some more functions to enhance the image of reality the user of the fly-through has, wil be needed. The way Bryce allows one to drape a satellite image or a scanned topographical map on top of the DEM is a first but valuable step towards a more useful product. It would be a good idea for the producers of this software to acquire some input of cartographers to improve these packages. They are actually needing them in multimedia products. Another improvement would be real-time rendering in the fly-through packages. This is of course a matter of computing power. Something like 8 frames per second is needed as a minimum to make such an application feasible. CPU innovation and perhaps multiprocessing may solve this problem. Until then one must take advantage of the Quicktime technology.
A last comment on the interface issue: This has everything to do with the way people like to manipulate reality. Lindholm and Sarjakoski (1994) state that the graphical user interface that people are used to, based on the desktop metaphor, is not suitable for GIS and cartographic purposes. They suggest other metaphors, like the workshop or kitchen metaphor to cover the extensive set of tools we tend to use. The KPT Bryce package is a nice example how straying from the familiar path of the desktop metaphor may improve the interaction of the user with the subject, in this case virtual space. That's the only relationship this package has with reality. For the larger part it produces surreal images of instant Dali quality.

References:

Bertin, 1967, Semiologie Graphique. Paris/Den Haag: Mouton, 1967 or Semiology of Graphics (English translation) University of Wisconsin Press, Madison WI, USA
Cornwell, B and A. Robinson, 1966, Possibilities for Computer Animated Films in Cartography. In: Cartographic Journal, vol 3, no. 2, pp79-82
Crichton, M., 1994, Disclosure. Arrow/Century Random House UK Ltd., London
Dorling, D., 1992, Stretching Space and Splicing Time:from Cartographic Animation to Interactive Visualisation. In: Cartography and Geographic Information Systems, vol. 19, no.4, pp.201-214
Fairbairn, D. and N. Deely,1991, Animations in the Cartographic Drawing Office. In: SUC Bulletin, Vol. 25, no. 1, pp 1-7
Kolacny, A., 1970, Kartographische Informationen - ein Grundbegriff und Grundterminus der modernen Kartographie. In: International Yearbook of Cartography, X.1970, pp.186-193
Koussoulakou A. and M.J. Kraak, 1992, Spatio Temporal Maps and Cartographic Communication. In: The Cartographic Journal , Vol 29, no 2, pp 101-108
Kraak, M.J., 1989, Computer-assisted cartographic three-dimensional imaging techniques. Delft: Delft University Press, 1988.
Kumler M.P., 1994, An Intensive Comparison of Triangulated Irregular Networks (TINs) and Digital Elevation Models (DEMs). Monograph 45, Cartographica Vol. 31, Number 2, Summer 1994, 99p.
Lindholm, M. and T. Sarjakoski, 1994, Designing a Visualisation User Interface. In:Visualization in Modern Cartography,Ed. A.M. MacEachren and D.R.F. Taylor, Pergamon/Elsevier Science Ltd. Oxford, United Kingdom, pp.167-184
Mackaness W.A. and B.P. Buttenfield, 1992, Incorporating Time into Geographic Process; A framework for Analysing Process in GIS. Proceedings of the 15th ICA Conference, Bournemouth, vol 2 pp. 565-574
Muercke P., 1991, Cartography and Geographic Information Systems. In: Cartography and Geographic Information Systems,
Moellering, H., 1984, Real Maps, Virtual Maps and Intercative Cartography. In: Spatial Statistics and Models, Ed. G.L. Gaile and C.J. Wilmott. D. Riedel Publishing Company, Dordrecht, Holland, pp. 109-132
Monmonier M., 1991, How to Lie with Maps. Chicago:.The University of Chicago Press, 176p
Robinson, A. H. , 1952, The Look of Maps
The University of Wisconsin Press, Maddison, 105p
Schama S., 1995, Landscape and Memory. London: Harper Collins, 652p
Thrower, N.J.W., 1959, Animated Cartography. In: The Professional Cartographer, Vol. 11, no. 6, pp.9-12
Tobler , W.R., 1970, A Computer Movie Simulating Urban Growth in the Detroit Region. In: Economic Geography , Vol. 46, no.2, pp. 235-240
Wooley, B., 1992, Virtual Worlds
London: Penguin Books.

Software:

Virtus VR 1.0
Virtus Corporation
117 Edinburg South, Suite 204
Cary, NC 27511, USA
tel. 1 919.467.9700
FAX 1 919 460 4530
e-mail: virtustech@aol.com
KPT Bryce
HSC Software Corp.
6303 Carpinteria Ave.
Carpinteria, CA 93013
tel 803 566 6200
FAX 803 566 6385
e-mail kptsupport@aol.com
Compuserve: go kptsupport
AOL: KPTsupport
VistaPro 1.0
Virtual Reality Laboratories, Inc.
2341 Ganador Court
San Luis Obispo, California 93401
tel. 805-545-8515
FAX 805-781-2259
e-mail vrli@aol.com
Compuserve GO VRLI
AOL: VRLI
Scenery Animator 1.2
Natural Graphics
PO BOX 1963
Rocklin, CA 95677
tel. 1 916 624 1436
FAX 1 916 624 1406
Obtaining USGS DEMs
American Digital Cartography Inc.
3003 W.College Ave.
Appleton, WI 54914-2910
tel. 1 414 733 6678
FAX 1 414 734 3375
($10/7,5 minute DEM)
Information about USGS DEMs
US Geological Survey
507 National Center
Reston, VA 22092
tel. 1 703 648 6045
FTP download of USGS DEMs
(30 arcsecond resolution)
edcftp.cr.usgs.gov (EROS Data Center)