Immersion: The Believability of Virtual Environments
Kingston University lecture, 1998
Programming Note: After returning to London after my time at The Jan van Eyck Akademie, while I was looking for a full-time design position, I did some one-off teaching at Kingston and Middlesex Universities. This is a presentation I gave at Kingston which was essentially a love letter to all the games I grew up on, and encapsulated a lot of the ideas Iβd been working on during the last few months at the van Eyck.
This work later matured into an article I wrote for Eye Magazine on the history of video games. The day I gave the presentation at Kingston there were all manner of technical problems, as the presentation I had developed was built for a Mac, and wasnβt safely written enough in the 8/3 file format then native to PCs. Thankfully these kinds of problems are a thing of the past, but Iβm still envious that I wasnβt able to use YouTube to build out the presentation with a lot more color, something Iβm glad to finally rectify here 20 years later.
Ever since the birth of computer games, there have been virtual environments. Yet with the increasing development and prevalence of graphics hardware, software, and interface design possibilities, what happens to the role of the VR environment when the rendering of real or imagined spaces becomes immersive?
The answer perhaps ironically lies in regressing to the early years of computer game development, when the spatial techniques of vector graphics, two dimensional scrolling and platform games, explored and expanded upon a vast wealth of creativity amongst not only computer programmers, artists and designers, but the gaming community as a whole.
Now, with far more emphasis upon virtual reality creation in the hands of the user, expanding the longevity of games and giving rise to large online communities of amateur digital architects, how might the principles of the inventive early games scenarios be applied to the current rash of pseudo-immersive digital adventures?
My discussion here will take the form of exploring a number of issues pertaining to these questions, not least with specific reference to the crucial role of the Internet in these matters, which has, within the last couple of years, allowed real-time multiplayer gaming to become a global event, as well as empowering the player to be able to join and mobilize other players from around the world. I intend to initially trace a history of the virtual environment, stopping off at various points of scenic interest. I will, however, be focusing upon specific recent examples of the genre throughout the discussion, including the environments of Cyan Software's Riven, Tomb Raider, the vector graphics of Asteroids and Battlezone, text adventures, and even Pac-Man's haunted house. Armed with this premise of deconstructing a space which doesn't physically exist (although I may end up proving that statement wrong, or at least problematic), we should begin with the notion of what type of space is exactly played back from the screen, and to set ourselves some parameters or boundaries for exploring these questions. Therefore, with the itinerary set for the tour, we should begin.
With almost any discussion of the birth, growth and domination of computer games, an obvious starting point is in 1972 with Pong. Described by American author J.C.Herz as the uber-arcade game, Pong had only two instructions - "Deposit coin" and "Avoid missing ball for high score". A monolithic paddle game, with its simple bat and ball premise, this suitably minimalist game still explores many of the gaming principles at the heart of games development today. The revolutionary aspect of Pong was that it was play without gravity - and this really differentiated it not only from the various forms of play which had preceded it, but also from the applicable physics of the so-called real world. Objects no longer behaved as we understood that they should. The ball here for example doesn't have a height at which it's traveling, nor does it have a real velocity, merely a series of paths between two moving or shifting points on the sides of the screen. Pong, as a digital translation of table tennis, distorts the space actually used, by flattening the entire experience down to the bitmapped two dimensional plane, yet importantly, leaving enough of the original source for the viewer or player to still hold a relevant frame of reference.
Notably, Pong has no real location in time and space - there is no particular geography to the environment, nor is there any indication of an environment or space outside of this arena. Douglas Rushkoff, author of Children of Chaos, expounds on this point by suggesting that whereas before computer games, most play had to do with some form of defiance of gravity, which he equates with the shape of Aristotle's poetics - you build something up to a climax and then fall away - the real revelation with a computer game was that within these virtual spaces, all the known physics of the world can be altered, or at least controlled. A possible example of the technology being able to bend the perceptible parameters of the world.
This started to become developed further by building on the technology of the ground breaking Pong, and the invention of what became known as vector graphics, as demonstrated by the likes of games such as Asteroids or Battlezone. Vector graphics essentially allow the drawing of lines on the screen to appear as if there is perspective, and therefore depth to the environment, placing the player in a kind of first person viewpoint (this was later substantially developed by id software's Doom, which I shall return to later).
The development of such technology, interestingly originated from the U.S. Defense Organization, who had been using the technology to render virtual battlefields for their troops, pilots, and drivers to train on - this of course, has specific reference to a game such as Atari's Battlezone, where it's a battle for survival amongst a gladiatorial contest of highly sophisticated tanks in a top secret location. Essentially the architecture was still being drawn as a series of lines, yet you could actually see the horizon now for example, and feel as if you were approaching it. The U.S. Military has been using video games to simulate the nuts and bolts of war for quite some time. At first, computers were used merely as scorekeeping devices in conventional war games. Instead of, say, using clipboards and drawing pins to track a field exercise, planners at the Army's National Training Centre in Fort Irwin began using computer screens upon which they could squiggle red and blue lines in a high tech military version of Match of the Day. While real tanks rolled around on the field, computers would track their positions and render an overhead view with a series of brightly colored icons : squares for tanks, diamonds for armored personnel carriers, umbrellas for radar, and stick figures for infantry units.
As computers pervaded every nook and cranny of the armed forces, they replaced the Invaders' bombs. By removing the architecture from the gameplay, the programmers here were able to separate the need for graphics (which were limited at the time anyway) and focus upon the gameplay. This has recently come round full circle again with the peaking development of the 'fighting' genre of games, such as the Tekken series, Virtua Fighter, or Battle Arena Toshiden, whereby the landscapes on which the players fight resemble nothing more than stylized, blurred out and translucent backdrops to the main action, allowing the player to concentrate upon the thread of the game. Therefore, by 'turning the environment down', in this case, the argument could be made that this is a way in which to increase gameplay, possibly at the expense of a realistic rendering of the actual situation which the game supposedly mimics.
Building on the success of PacMan in creating a virtual celebrity with his own haunted house, complete with a deluge of ghosts who just won't go away, developers, now looking to expand into the ever growing home market, used this strategy to try to market games to the players. In this instance, characters were created to add to the idea of the player being able to identify with them, and therefore become more 'immersed' in the experience. What this strategy forgets of course, is that that same degree of immersion, requires the character to be placed 'somewhere', and order to react to a set of varying circumstances. Essentially a platform game based upon the original Donkey Kong arcade, one title which perhaps succeeded at both was Bug-Byte Software's Manic Miner and Jet Set Willy series, which were both two of the most popular titles and commercially successful games of the home computer boom of the early eighties.
In Manic Miner, Miner Willy had to traverse a number of screens rendered as underground caves, containing all sorts of nasties, whilst collecting keys so that he could initially progress to the next level, but also ultimately escape from the caves altogether. The environments which Manic Miner's creator Matthew Smith put together consisted of 20, brightly lit caverns, even including quicksand, so the player could alter the landscape of the level as he played it, and a bizarre, almost Python-esque selection of meanies, including spinning rabbits, mutant telephones, flying eyes and mobile bank safes. Far from reworking his original game, the sequel - Jet Set Willy, was a sprawling affair, set in Miner Willy's house - a huge mansion bought with the profits from his mining expeditions. After a massive party, Willy finds himself with the Herculian task of trying to clear up all the glasses from around the building. In this case though, the player was, for the first time, able to freely roam the collection of screens, without having to complete specific tasks in order to progress from one to another.
As a platform game, Jet Set Willy broke the mould by offering the player the possibility of exploration without real direction. You could either really try to clear up (which I understand was impossible due to bugs in the source code, and could only be achieved through cheat modes), or just enjoy the multitude of inventive solutions which Smith had come up with to digitally render the Miner's home. This ironically was reciprocated back into the Mario Brothers series, with the portly Italian plumber facing strangely similar puzzles to those faced by Willy in his huge mansion. Considering that all of this was done with a sixteen color pallet on the ZX Spectrum, it sheds an interesting new light on games such as Tomb Raider in terms of how far such 'revolutionary' games really are in terms of the progression of the medium. I'll get to Lara Croft a bit later on though.
As game developers therefore began attempting to render more and more believable worlds or arenas where they could place their characters, the rise of three dimensional graphics in computer and console systems started to try to bring the player closer to complete immersion within a virtual world. This could be construed as having been at the expense of the two dimensional games, and indeed, in a recent article in Edge magazine, Jez San, from Argonaut Software, interestingly put the comparison in this way:
βSome games are better in 2D, particularly puzzle games, and also some platform games. Although thereβs a lot more you can do with 3D where youβve got a full world you can roam around, there are some specific character movements - like jumping, and judging where youβre going to land - that are actually easier in 2D than in 3Dβ.
Indeed, it's hard to dispute the fact that certain games - and arguably, a few genres - are less appealing in 3D. Reduced to their constituent elements, as the early programmers had to make them, most non polygon games retain the creative mainstays inherent to their sprite based forebears. So, with the introduction of an additional dimension, developers acquaint a 2D design brief with a variety of complications. From camera angles to collision detection, it's a complicated process. And certain features just don't survive the transition between the two mediums. The 'update', that enduring cash cow of the gaming industry, perhaps best illustrates the shortcomings of 3D where many simple traditional gaming concepts are concerned. In a game such as Defender for example - could this be rendered into a three dimensional game? This indeed would be problematic, because you've got to be able to see behind you - if you can't you've lost some of the original's appeal.
Essentially, it would be impossible to create a modern, 3D reiteration without altering the original concept. Would Defender's simple, compulsive 'twitch' gameplay survive that process? Probably not. In terms of opening the floodgates for the development of three dimensional games, the milestone by which all efforts are usually measured is id software's Doom. A first person perspective horror rollercoaster of a game, immersing the player within a labyrinth riddled with puzzles and monsters spawned straight from hell itself, Doom's environment could quickly transcended their original use as electronic scorecards and subsumed the game themselves. As the arcade craze boomed in 1978, Atari was building a souped up version of Battlezone for the Defense Department's Advanced Research Project Agency (ARPA for short), who also were significantly involved in the initial development of the Internet.
While consumer flight simulators proliferated as home computer games, Air Force material managers programmed their computers to spin variables (fuel levels, weapons load, radar profile, what the pilot had for breakfast that day etc. etc.) into a supersonic airborne battle with impressive color graphics. The Navy's Enhanced Warfare Gaming System allowed a game director to operate scenarios from SimNorthSeaSkirmish to SimGlobalTorpeclothon while opposing ensigns peered into their separate sets of screens.
Yet to return again to the development of the computer game, the next logical progression of the digital environment in particular was to expand upon the graphics, whilst still keeping the gameplay and environment simple and compelling. A good example of this might be something such as Pac-Man. He, of course, needs little or no introduction, being really what became the first virtual celebrity. There are now huge online archives devoted to the merchandise of PacMan and its sequels.
There were even strange albums made such as Buckner and Garcia's PacMan Fever, and, in 1981, How to win at PacMan was on the best seller list. In creating PacMan, and subsequently a range of characters around him, Namco and Atari were able to chain together an incredibly lucrative string of titles, for an ever (at the time) increasing range of computer platforms. The idea was simple, the architecture even simpler, as this excerpt from the original Midway arcade manual suggests:
βThe player, using a single hand control, guides the pacman about the maze, scoring points by munching up the dots in his path. Four ghost monsters - Inky, Blinky, Pinky and Clyde - chase after the pacman trying to capture and deflate him.
The pacman can counterattack by eating the big power capsule that enables him to overpower the monsters for additional score. After all the dots are gobbled up, the screen is cleared, and pacman continues for another round.
Each rack features a special fruit target in the maze, which, if eaten, earns bonus points. Players start with three pacmen. An additional pacman is awarded for 10,000 points.β
In terms of the architecture of PacMan, the instantly noticeable thing is that there is no variation in the make-up of the screen from level to level. The walls are always in exactly the same place, with the level of difficulty increasing due to different paths the ghosts take to try to ensnare you. The same could be said of PacMan's contemporary Space Invaders. Here, the four defense barriers are always placed in exactly the same way from level to level, despite continual erosion from above.
Yet Doom certainly canβt be deemed three dimensional, despite itβs promise, as unfortunately the sprites were not - still adhering to the flat frames of animation which were the mainstay of character movement at the time. This is particularly noticeable when you kill a monster and subsequently walk around it - the monster will appear to rotate with you, always facing the same way from wherever you are looking at it. This is because the code of the game would tell the computer that 'this monster is dead now - call up the dead monster frame'. This, of course, then doesn't change, causing the 'bug'. id software later fixed this with a completely new 3D game engine in the form of Quake but disappointingly you were still trapped in the gothic labyrinth, rendered again in a delightful earthy brown.
Doom was a watershed event on par with the release of the Atari 2600 console, because it changed the way video games circulate and reproduce. Distributed initially over the internet, the Doom demo was downloadable from the University of Wisconsin's ftp site, paving the way for 15 million users to download the demo, which they did all over the world. Atari's machine housed video game code - the impalpable string of zeroes and ones, the game itself - in a plastic cartridge. Atari gave video games a very successful kind of body : rugged, mobile, and interchangeable. The Atari 2600 cartridges gave video games a way to proliferate in any home with a TV. Similarly, Doom gave video games a way to proliferate in cyberspace. Doom is poised like some sort of reptile between water and land. It alternated between physical and virtual generations, shifting into shrink-wrapped form just long enough for it to make money. Down the line, you can see a point where video games will be sold in electronic form and jettison their bodies altogether. Doom is a fulcrum.
The scenario was fairly straightforward. Stranded on a moon base near Mars, you (a battle hardened marine), receive a message requesting military support because all hell has literally broken loose on the moon. Armed with just a pistol and seeming to be the sole survivor of the entire development, you open the airlock and begin blasting your way through an avalanche of evil incarnate. Once you have all your guns, Doom is basically an enhanced 3D first person version shoot-em-up. It's a pretty straightforward descendent of Atari's Battlezone. But a decade of technology has made all the difference to the graphic realism, and a decade of serial killers, celebrity murder trials, and special effects in action movies have made all the difference in the standard of stylized graphic violence. It's all so deliciously clear-cut. But beyond the basic tribal shoot-em-up premise and social itches it seems to scratch, there is something more important which sets Doom apart from the hundreds of first person 3D shooters on the market. And the crucial element is fear. Other games have great graphics, but none, due to the nature of the architecture in this example, have triggered the same degree of deep, primal terror. For hundreds of thousands of people, Doom has invoked the kind of terror usually only associated with the movies.
Doomβs direct descendent, Quake, has even resulted in public events at trendy nightclubs.
Quakeadelica, staged at a South London nightclub with an extravaganza of music, glamour and button-pounding action was a head-to-head battle between the top eight British players of computer game Quake 2.
Hosted by model Jo Guest, who seemed a little out of place (but perfectly suited to the testosterone drenched atmosphere of the boys and their games), the evening included a series of heats, and a 'winner takes all' challenge between the best British contenders and the US champion.
Sponsored by online games network Wireplay, the evening started with brief introductions before they assumed places and virtual carnage ensued on a large video projection screen.
The assembled crowd witnessed the 'chasecam' view of the ten-minute battles, cutting between views of the players in order to catch the best of the action. A special 'expert' adjudicator also played to eliminate foul play.
Moving at alarming speed, the game was awash with the usual virtual blood and guts, the deception and dodging of opponents, and when a particularly gruesome kill was achieved, there was always the simultaneous raucous cheer of satisfaction from the gaming community crowd. The British winner is now set to go on an all expenses paid trip to New York to play in the Professional Gamers League final.
Yet the star of the evening was undoubtedly 'Thresh', the US champion, and top professional gamer, who had been flown over to compete against what Britain had to offer. Aged 21, and sponsored to the tune of $100,000 a year for his gaming endeavors, 'Thresh', who recently won a Ferrari for playing, considers himself to be one of a new breed of sportsmen, and trains for up to 4 hours a day and 6-7 hours a day in the run-up to a tournament. He was quoted as saying that he would like to see computer gaming recognized as an Olympic sport, and hopes that it will make him a millionaire by the age of 25.
The overall winner though, of course, was the game itself, the evening truly being a celebration of one of the most addictive, immersive, and compelling games around today, and confirmed not only by the huge attendance of the event, but by the realization that companies are also willing to fly the best players around the world to compete against each other in the name of publicity - both for their software and online gaming services. It certainly seems as if Thresh's dream of becoming recognized as one of the first sportsmen of the modern era would not seem too far away.
In complete contrast to this genre of computer gaming, there has always been a focus or strand of games which have seemed far more contemplative, concerned with puzzle solving and closer perhaps to adventure gaming, which I would argue, probably arose out of the Dungeons and Dragons trend of the late seventies.
Initially starting out as just text adventures, with a few outcomes per line of text, usually something like - "You are in a forest clearing, exits are west and north", these have developed into all encompassing graphic adventures, the most notable of these being Cyan Software's Myst, and it's sequel Riven. In one sense, a form of 'interactive storytelling', it bears the mark of the 'free exploration' game, which I discussed in reference to Jet Set Willy. Here, you interact with a series of puzzles and a slowly unfolding story, unlocking a world in which the story took place.
This strategy, says J.C.Herz, is instantly recognizable to any murder mystery fan (and why detective fiction lends itself so well to computer games). It's a whodunnit, which is to say two stories superimposed upon one another. One is the sequence of events that happened in the past, which you can't change but is a very good story. The other is the sequence of events that happens in the present (for example, you wondering around trying to solve puzzles), which is usually a terrible story, but nonetheless a highly interactive one. As you hurdle the obstacles in the present, you unfold the larger picture of the past. Piecing together bits of history can feel like an interactive story. This is essentially what happens in Myst.
Yet the people who bought Myst didn't just by it in order to confront the puzzles or unlock the stories. They bought it for the scenery. Myst was gorgeous - and not just in the sense of being realistically rendered or artistic. At the time, there were dozens of games with amazing graphics. But they all used very technically impressive artwork to create some kind of hell filled world with bloodsucking zombies and chainsaw wielding lunatics. Myst, on the other hand, put you in a world you might actually want to visit, if only you had the money and time. Its appeal had little to do with problem solving and some tedious family drama. It was that you got to take a holiday on this great island where you could go hiking and climb trees and be alone as the water lapped against a dock. It was an escape destination. In essence, Myst was a highly successful exercise in virtual tourism. A slew of derivative computer getaway adventures predictably followed, relentlessly advertising their scenic attributes in order to out-Myst the competition. A recent publication stated :
βIf virtual world ever become sufficiently lucrative, electronic publishers will probably try to get famous architects to design them. So youβll be able to play in a world designed by I.M. Pei or Frank Gehry or Philip Johnson, even if you canβt afford to visit, much less commission, one of their buildings.
At some point, Delia Smith or Richard and Judy will put their images onto a CD-ROM set in some insufferably gracious virtual farmhouse or high-beamed seaside bungalow, where at any point you can take a break from the game to shop - that gnarled wooden bookcase may provide important clues, or look great in your study. Please specify oak or cherry. Click here to order.β
By comparison, Riven, the sequel to Myst, far outstripped its predecessor in terms of the graphic embellishment afforded to its screens. Even incorporating animation for the beetles climbing up and down the tree trunks, and a vastly improved soundtrack, and the appearance of people you could interact with, Riven became like this gloriously contemplative and sumptuous interactive Greenaway film, such was the meticulous attention to detail.
The latest trend, particularly over the last couple of years, is to try and combine all of these elements together into a gargantuan juggernaut of interactive horror that you only experience as a small child when the lights go out and the monsters under the bed come to life. Ultimately, you know you're safe - you can always race for the door, and the sheets always feel like their knife-proof. But for that moment, you're exquisitely frightened. It's the kind of fear that makes us feel alive and sends us on skydiving expeditions and rollercoaster rides. Doom gave you a way to get that same thrill from your very own home. Doom took you into its own world.
This was doubly true, because id not only created a compelling fictional world for Doom players, but gave them a stake in it by releasing the actual source code of the game itself, allowing Doomers with the smallest of programming skill to add to the game, essentially picking up where the game's authors left off. They could create custom soundscapes, tweak the game's configurations, or even create new levels, entire episodes of the game. Releasing the source code gave players all the keys to the castle - but after all the shortcuts and secrets were laid bare, the challenge then became to compose better variations on the Doom theme than anyone else. The Doom source code itself became a kind of world. And the online Doom constituency transcended the mere sharing of cheat codes to become a community based on shared, self-made chunks of the Doom Universe called .wad files. Players became a part of Doom's world not just because they played the game but also because they constructed bits of it. They competed not just as players but also as creators.
The culture of creating additions to games, particularly those of the first person perspective variety (such as Doom, or Quake), has grown and grown with the Internet, companies such as Netscape, Microsoft, and even BT actively encouraging the development of such material into online multiplayer gaming. Here the two go hand in hand. In creating an addition to the original game code (usually referred to as a patch), the player is able to become a digital architect and enhance the features of the game which he enjoys and remove those which he finds unsuitable, after which he can immerse himself within the scenario he has created, and roam around the world which he has built. A virtual way to play God on a minor scale perhaps. He can then swap or download other patches from other users to see what other people have come up with. This might just be a new way to make a door for example, or a set of files which turns your game into the wild west, or underwater or whatever. The point is that the creativity in finding a solution and ever more devious ways to play the original game is what is being cultivated here.
This of course, is great for the original creators of the game, because it certainly secures the longevity of the game, and ensures an ever growing community of gamers working on developing the game further. The notion of playing these levels and arenas with other players simultaneously, exploring the environments at the same time, either hunting in groups or fighting against each other, comes together in the form of the network game. The network game returns the uneasy remoteness to the computer gaming idea, and gives your opponents far more brains than the computer could generate. With the network possibility, particularly over the Internet, the computer game can now become a global event, with the virtual environment of the game itself, being played out within the wider virtual structure of the world wide web.
The idea of having a four player game of Doom with your friends (which you of course hope to obliterate) in Australia, America and Japan is no longer a technological fantasy. Indeed, one of the fastest growing areas within the Internet deals with precisely the organization and assimilation of interested players waiting to inflict virtual carnage upon each other. The network games present a simulated version of playing against the computer, yet with all the imperfections and unpredictability of a human player now thrown in. You might not even know the player, so their presence within the game bears close similarity to what the computer would generate anyway. The idea is, that the human opposition would move around in a completely different and unpredictable way, one which the computer would never be instructed to achieve, and therefore the learning of movement patterns becomes completely obsolete. This technique ensures that no two games are ever the same, let alone similar, and almost removes the notion of strategic play (except perhaps for various degrees of hiding).
More recent games have doubled down on digital storytelling, comprising of cinematic excerpts, sumptuous scenery, highly interactive characters, and perhaps most importantly of all - a decent plot to keep you interested. SuperMario 64 is a good example of this, with its Disney-esque characters and cute maneuvers, especially when you can choose to tackle the puzzles in the order you wish, and even walk around the puzzles as they are so huge, yet one title which has perhaps come particularly close is the Tomb Raider series, with the instantly recognizable Lara Croft.
In Tomb Raider Two for example, you embark upon a world tour (which continues in the third installment of the series), visiting vast renderings of Tibet, Venice, an abandoned Monastery, even an underwater wreck of what appears to be the Titanic. The Tomb Raider series is an interesting example of digital architectural creation however, because, however free you feel to explore the beautifully rendered scenery, you are in fact being inadvertently led through the levels, resulting in a fairly linear experience overall. There are some fascinating embellishments to the plot, including the opportunity to try out some alternative forms of transport in order to continue the tour, including a speedboat, a snow bike, and intricate rope swings. Contrary to the publicity surrounding the Tomb Raider series, I would argue that it's really the environments in which Lara exists which account for the immersive experience of the game, plus, with the over the shoulder interface, you can actually see your player (in this case Lara), actually in situ within the level, a compromise between the 2D platform game and the 3D first person perspective of games like Doom.
This essentially brings us up to the present day, so perhaps by way of conclusion, if it is at all possible to conclude on such a subject barely in its infancy,
I would like to offer up a few questions and concerns surrounding the nature of this genre of digital production. For example, I've chosen in this discussion to break the game idea down into its constituent parts so far within the boundaries of how the game has developed over the last twenty or so years. Could new constituent parts be added, if so, what would they be? Would it at all be possible to construct a navigational structure not based upon one which mimicked reality in any great way, or to traverse a 'landscape' for want of a better word, which bore no relation to anything we could have a frame of reference to? And what role does the Internet have in all this? We're already seeing its massive influence with the multiplayer game genre, but how might this develop?
On these questions, Iβll leave the last words to Matthew Smith, programmer of Manic Miner, who, back in 1984 was reported as saying:
βthings get hairy when we get machines which are more intelligent than us ... what I donβt like about certain games is theyβre not a simulation of any real kind of problem. Iβm not into simulated violence. Itβs really not that much fun. If I had to be shut in a room with one spectrum tape it would have to be Atic Atac.β
I recently found my original slides from this 1998 presentation, tucked away on an old archived CD-ROM in a dusty corner of my studio. Here they are for the first time online:
