EMULATION
An emulator is an application that tries to mimic another system in order to run applications the way they were run on their original system. It is usually used by gamers interested in playing older games on newer systems: older console or arcade games on PC, DOS games in a recent version of Windows, etc. It can also be a way to play contemporary games incompatible with a specific system. In the 1990s, many video games were incompatible with Apple computers. Virtual PC, an emulator mimicking a PC on an Apple computer, was therefore an interesting option. Emulators are also used by consoles for backward compatibility: for example, the PlayStation 3 console uses an emulator to load original PlayStation games.
Emulation is convenient for any player or researcher seeking to have a substantial collection of video games without having to preserve a very large range of hardware. Arcade games are the most relevant example: it is uncommon for an individual to keep a video game arcade cabinet in his or her game collection, and each cabinet is often a dedicated-machine. The Multiple Arcade Machine Emulator (MAME; www.mamedev.org/), an arcade games emulator for PC, is therefore a way to have a collection of arcade games without owning the actual games. But emulators can only load an image of the game and do not necessarily preserve accurately other aspects of the game itself: its graphical aspects, sound precision, controllers, and original playing context. It is, nonetheless, often the only way to have access to some games. Usually, emulators add specific additional features to gameplay itself: the possibility to save game states everywhere, to fast-forward or leap backward in the game, to record a game session, etc. These new possibilities can change a game’s difficulty and distort the original length of the gameplay experience.
Emulation is an important tool in building video game history. Even if it was possible to have a perfect emulation of an original game, most analyses focus on play, which cannot be easily preserved. I seek here to elaborate on the virtues of using emulators as an archiving tool and as a means of accessibility, but still want to indicate the limitations of integrating emulation into game studies. Video game history should not only be a project of preservation, but also one of contextualization.
Usages and Consequences of Emulation
A work of digital art can be lost in less than ten years; if not permanently lost, digital artworks are usually at most inaccessible (Winget & Murray, 2008, p. 1). Floppy disks will last a maximum of 30 years (Edwards, 2012, p. 1). CD and DVD lifespans vary a lot depending of the quality of the products: some recent DVD-R could last less than five years, some CD-R are expected to last between 30 to 45 years in controlled condition, while a lot of CD-R should be readable for more than 45 years (Final Report, 2007, p. 16). But any video game support is in some way ephemeral, whether it is a CD-ROM, cartridge, magnetic tape, or hard drive. Emulation seems like a normal response in order to preserve at least a virtual image of a game. In most emulators, a game will be preserved in a computer file, usually referred to as a “game image,” for it is an exact replica of the game’s original software.
But emulation changes the gaming experience in different ways, whether because of the player’s hardware or the emulator’s lack of precision itself. For instance, different screens render graphics in different ways. A 1980s NTSC TV set is not as precise as a 2010s DEL screen. The cathode-ray tubes of home TV sets and arcade cabinets give a blurry quality to the graphics that masks “imperfections” in the final render (Therrien, 2012, p. 15). In terms of preservation, a sharp-edged pixel is not faithful to what a common player experienced in the 1980s. Even pixel aspect ratios are different from one screen type to another. Moreover, the GameCube and Wii emulator Dolphin can upscale the original resolution up to 1080p, which of course is not faithful to the original conception.
Console games’ controllers can be replaced in emulators by a keyboard or any joystick. PC gamepads can to some extent be similar to console controllers, although the shape, the distance between buttons and the directional pad can have a big impact, especially on games that require quick actions such as fighting games. Arkanoid (Taito, 1986) used a spinner control in its arcade version, and even though a replica was built by some MAME developers, the initial accuracy was never really completely achieved. The NES Zapper, which was used in Duck Hunt (Nintendo, 1984), works by light feedback from the screen, but only works with a cathode-ray TV. Some emulators let you replace this kind of accessory with a mouse: it adds an aiming overlay over the game image and your mouse becomes the shooting device. Yoshi’s Safari (Nintendo, 1993) in a ZSnes emulator, for instance, looks like a desynchronized browser-based flash game when using this method.
Emulators tend to be a tool for accessibility, rather than a tool to preserve the intact object and its original context. Most emulators use high-level emulation, which means that instead of reverse-engineering the console itself, they will simulate its functionalities (Fenlon, 2012). They are coded so that popular games run correctly: the emulator recognizes the loaded game and uses specific hacks in order to run the game. It is easier to use high-level emulation, considering that each game has specific needs and would thus require a large amount of time and processing power. At the end of the 1990s, Nesticle could emulate NES capabilities with around 25MHz, at the cost of precision. Today’s most popular NES emulators are Nestopia, which requires 800MHz, and Nintendulator, which requires 1.6GHz (Kuchera, 2011, p. 1) and these have greater precision than their predecessors. These processor requirements were not met by most if not all home computers in the 1990s.
This goal of precision is often central for some emulator projects. For those behind the MAME project, playing the game is considered merely a “nice side effect” (MAME, n.d.) from their objectives. Sure, you have to actually play the games to see if the emulation worked; but making games easily playable (on a modern PC) is not a specific goal they pursue. With a similar philosophy, the bsnes emulator was created by byuu—a pseudonym—with the idea of creating an efficient SNES low-level emulator, which means that accuracy has priority over playability. There can be problems when precision in emulation is not a clear and central objective. High-level emulation is not always faithful to the original, even in terms of gameplay. In Air Strike Patrol (Opus, 1994), for example, an isometric shooting game for the SNES, there is a shadow drawn under the aircraft the player controls. Since it is difficult to render, it will not show up with a common emulator (ZSnes or Snes9X, for example)—and the typical player will not know that it should have been there. With a shadow in place (in the bsnes emulator), you can drop bombs with more precision (Kuchera, 2011, p. 1). If you cannot see this shadow, the game thus is more difficult than it actually was on the original system. Accuracy through emulation is therefore an important goal for historical purposes, though never completely possible or measurable.
Even bugs are important to preserve from a historical perspective. Star Fox (Argonaut Software, 1993) used a specific chip, the Super FX, in order to push the SNES system to its limits in graphics rendition. As a result, console players with the original SNES system experienced slowdown while playing (Kuchera, 2011, p. 1). An accurate emulation should in theory restore these imperfections in the final render, but it is not always the case. You will never know if an obscure game you load within an emulator is correctly rendered, if you don’t have the real functionality of the original hardware. Obscure games—and obscure game platforms—also tend to have less documentation about them that can be used to evaluate a given rendition of a game. As such, common emulators’ developers will make sure popular games are adequately emulated and will use game-specific hacks to do so, but they cannot guarantee that rare games are executed correctly. Byuu’s goal with the bsnes emulator is to reverse-engineer the machine, in order to make sure any game is correctly rendered without having to fix every bug game-by-game (Kuchera, 2011, p. 2).
Of course, preserving a virtual image of a game image is something, but the gaming context is much wider. The place in which a gaming experience occurs gives it a general ambience: arcade cabinets were often placed in arcade, with noisy pinball machines’ mechanics and chimes, loud music, low lighting, etc. Arcade cabinets are designed to give a specific experience that a common computer desktop cannot render. For the purpose of convenience or immersion, game cabinets differed from one game to another: racing games featured a driving wheel and pedals, card games can appear on a “cocktail” table cabinet, etc. Even the simple ambience of a console in a 1980s living room on a small TV set is not the same as a researcher’s office computer screen. Since most emulators are for the PC, it is rare to see a handheld console game still being “handheld” through an emulator.
Multiplayer console games can usually be emulated normally, with a different controller for each player. Some browser-based flash emulators like NESBox let you play online with friends in an Internet browser without any prior installation. But you will not have a similar gaming experience when your opponent is not in the same room. DOS games with modem or LAN multiplayer are harder to emulate, and very dissimilar to the experience of what a phone-line dial-up modem could give. Even when you can emulate a multiplayer game, a vivid online community does not necessarily exist. Emulation can’t revive MUDs (multi-user dimensions) or MMORPGs (massively multiplayer online role-playing games) communities when they disappear.
Accessibility
Emulators are a means for accessibility for game studies, for research as well as for teaching. If a teacher wants to analyze a specific SEGA Master System game sequence in the classroom that should be played beforehand by their students, it is easier to make a saved game available for them online with an emulator rather than to have a cartridge shared and to risk your saved game to be overwritten. Researchers can exchange game states over the Internet in the same way.
A lot of old games can be obtained legally by emulation. The Good Old Games website (www.gog.com/) is a good example of how emulation can put old games back into the commercial circuit. Amiga Forever, World of Spectrum, and MAME have gained the rights from many games’ copyright owners to put free downloadable games on their websites.
In other cases, the copyright problem is unclear. Some games fall into the “abandonware” category. A video game company can go bankrupt or be sold to another one, leaving some games’ copyright ownership in a gray area. Technically, though, “abandonware” are still illegal to copy (Edwards, 2012, p. 1). But to some extent, using emulation to show games in a classroom could be considered “fair use,” as Clara Fernández-Vara suggests (Cifaldi, 2012). The ZX Spectrum, for example, is a computer that was never released in North America and emulation is the most convenient way to give an impression of that machine’s possibilities outside of its original distribution regions.
Minimal accessibility is a condition for some game systems research. As Bernard Perron and Mark J. P. Wolf remark in The Video Game Theory Reader 2, game scholars usually work on games that are the most accessible to them (2008, p. 9): contemporary games rather than earlier ones, recent console games rather than arcade, handheld console, or Amiga games. In fact, the technological gap that needs to be bridged in order to run older games can be quite great, even for games from the 1990s.
Say you want to emulate DOS games through DOSBox, one of the most famous emulators. First, of course, you need to know how DOS works in order to run any game with it. But you also need to know the DOSBox-specific code. With textual inputs, DOSBox lets you simulate specific partitions, CD-ROMs, and floppy drives. Through a configuration text file, you can also configure screen resolution, windowed modes, speed, sound cards, etc. While there is a tutorial to explain basic functions, it is not always easy to find out how specific games can work, unless you download a game already configured for DOSBox.
Such is the case for a lot of games sold through the Good Old Games website. If you buy any game running on DOS, Good Old Games will provide you with a preconfigured DOSBox application with already defined parameters in order to make it compatible with recent operating systems. However, you may still have to configure some parameters manually if you want to respect the aspect ratio and resolution of the game’s screen or if you want the game’s speed to be playable. With DOS games, there is not necessarily a “correct” speed, since there is no standard in hardware. Even if a game seems to work, it is difficult to tell if the game’s speed, graphics, and sound are rendered in a possible way for a computer of the game’s period.
Other games running with older versions of Windows are more complicated to emulate. VirtualBox lets you configure in details a computer with a specific operating system, provided you have a copy available to install. But running a game on an older operating system is not as simple: you will usually need to install drivers for mouse, keyboard, graphics, etc. Having a stable emulated system is not an easy and quick thing to do for those without the necessary expertise.
A New Layer of Contextualization
To some extent, the context in which emulators appear is, from a researcher’s perspective, similar to what the appearance of VHS was for film scholars in 1976. Michel Chion (2012, p. 13) reminds us that in this period, movies were watched either in theatres or on television, which means that they were seen and heard in a contiguous and limited time. The viewer in this context can’t stop the movie to see an image in detail. Any film analysis of this period is only meaningful if its original projection context is taken into account: researchers who can watch a sequence as much as they want have to be aware of this technological translation and not overanalyze something that couldn’t be perceived by the typical viewer of the time (Chion, 2012, p. 14). Such an attitude must be adopted with video games: emulation is not a preservation of every aspect of an apparatus, but it is still a convenient and accessible way to have an experience analogous to the one on the original system. This “second-hand” experience requires scholars to be aware of the technical and cultural differences between the two systems.
Even though new possibilities offered by emulation can be seen by some players as great improvements over the original version of the game, a researcher must see these modifications as obstacles for the understanding of the initial experience. The Ogre Battle: March of the Black Queen (Quest, 1993) example will help us understand how the emulated experience can be misleading.
Ogre Battle was a fairly rare game. Only 25,000 SNES cartridges were released in North America and none in Europe. It is difficult for today’s collectors to gain an original cartridge. It is through emulation that most of today’s gamers experienced the game and, therefore, that the game gained its actual fame. As Ogre Battle shows, a game can be preserved and highly accessible by emulation, but as such, it can also gain notoriety amongst contemporary players. Instead of only preserving traces of the original context of the game, it also needs a new layer of necessary contextualization. Estimating its eventual legacy should not elude this additional (and later) notoriety.
Most emulators offer a “save state” function, which allows the player to save a game anytime during gameplay. As such, they can disrupt the original difficulty experienced when playing a game. In Ogre Battle, the player controls a rebel army that seeks to overthrow an evil empire. Each time a city is taken by the player, the player can draw a tarot card, which will give the player some benefit or disadvantage depending on the card. With an emulator, however, the player can simply save a state before the card in drawn and, depending on the result, keep the card or reload the previous state to reject it instead. Moreover, each mission can last an hour or so, without any possibility to save the player’s progress in the original game until a mission is over. These difficulties are completely lacking for those using an emulator.
Many emulators also give the option to fast forward a game, which completely changes the gaming experience, though it can be seen as an interesting feature. On the tactical map of Ogre Battle, units are moving very slowly and the game does not feature any accelerating function like the Creative Assembly’s Total War series (2000–present) would offer later. When two squads meet, the resulting skirmish occurs automatically: the player can only choose a general tactic for its units, use a limited number of special skills, or retreat. The game was quite long with a lot of skirmishes occurring in each of the game’s 25 main levels, accelerating the pace of the action at some points, making it more interesting. In many RPG games, say Final Fantasy VI (Square, 1994), this acceleration feature let players accumulate experience points easily, reducing the laborious and necessary “grinding” time to complete the game. The contemporary notoriety of some of these games could partially be increased by these new gameplay possibilities. The players can thus renegotiate what they see as a “fair” game, new possibilities being optional.
Since emulated games are now usually downloaded illegally, games do not have the same value as those that are purchased and players won’t bother to investigate a game in-depth when the first attempts at playing it are not judged as satisfying. In earlier times, a player would have to invest a sometimes substantial amount of money for a game, and would spend a lot more time to comprehend its game mechanics instead of simply abandoning it.
Understanding what Ogre Battle’s role is in video game history necessitates, first, to look at what the game was in 1993 on a SNES console and, second, to look at how emulation offered new convenient gameplay aspects and increased its fame. Along this line of thought, some genres have been more suitable for emulation than others, especially when emulators were less precise, and have gained an additional fame for retrogamers. For example, Japanese role-playing games’ turn-based mechanics do not need to be precise in real-time rendering and the keyboard could easily be used as a replacement of the console’s controllers to implement the player’s decisions without changing too much the original experience. Another example of the new layer of contextualization that emulation brings is the possibility for fan communities to translate games to other languages and offer them a parallel distribution. By hacking the original ROM (read-only memory) files, fans will translate, usually, Japanese games into English. Some games are only accessible for an English-speaking player through fan translation: such is the case for Mother 3 (Nintendo, 1996), even though the second game from this series was released in North America under the name of Earthbound (Nintendo, [1994] 1995) (Pelletier-Gagnon, 2012, p. 76). Final Fantasy V (Square, 1992) was never released for SNES outside Japan. Before its release within Final Fantasy Anthology for PlayStation in 1999, English translations of the original ROM file were already distributed online by amateur translators. It is therefore difficult, for example, to estimate to which extent these games contributed to the legacy of their genre or series for North American players since there is no clear trace of this underground circulation.
Hacking ROM files also leads to game creation; through emulators, game designers can program games as if they were running on older systems. As Byuu underlines, the problem is that some newly-created games relied on specific emulator imperfections and were unplayable with other ones (Kuchera, 2011, p. 1).
Conclusion
Using emulation as an archiving tool and for the study of video games necessitates different ways to contextualize games, their production process, and the way they are played. James Newman and Iain Simons from the National Videogames Archive in United Kingdom suggested that, instead of trying to preserve everything related to a game, the priority should be to interpret any available source. Preserving games or documents of game production is not meaningful for them if something explaining the context of games and documents is not added (Newman & Simons, 2009, p. 5). It is also the attitude suggested by Henry Lowood. As part of a history of video games where the “relationship between hardware, code, use and context for use” (Lowood, 2004, p. 4) is central, emulation is crucial but not the sole tool to make sure we still have traces of older gaming experiences.
Chion, M. (2012). Entendre une langue, en lire une autre au cinéma. Kinephanos, 3(1), 6–19. Retrieved August 15, 2012, from www.kinephanos.ca/2012/entendre/.
Cifaldi, F. (2012). GDC 2012: The unique challenge of making students play old video games. Gamasutra.com. Retrieved August 15, 2012, from www.gamasutra.com/view/news/164690/GDC_2012_The_unique_challenge_of_making_students_play_old_video_games.php.
Edwards, B. (2012). Why history needs software piracy. Technologizer.com. Retrieved August 15, 2012, from http://technologizer.com/2012/01/23/why-history-needs-software-piracy/.
Fenlon, W. (2012). 16-bit time capsule: SNES emulator makes a case for software preservation. Tested.com. Retrieved August 15, 2012, from www.tested.com/news/44376-16_bit-time-capsule-how-emulator-bsnes-makes-a-case-for-software-preservation/.
Final Report: NIST/Library of Congress (LC) Optical Disc Longevity Study. (2007). Retrieved September 10, 2012, from www.loc.gov/preservation/scientists/projects/cd-r_dvd-r_rw_longevity.html.
Kuchera, B. (2011, August 9). Accuracy takes power: One man’s 3GHz quest to build a perfect SNES emulator. Ars Technica. Retrieved August 15, 2012, from http://arstechnica.com/gaming/2011/08/accuracy-takes-power-one-mans-3ghz-quest-to-build-a-perfect-snes-emulator/.
Lowood, H. (2004). Playing history with games: Steps towards historical archives of computer gaming. Retrieved August 15, 2012, from http://aic.stanford.edu/sg/emg/library/pdf/lowood/Lowood-EMG2004.pdf.
MAME (n.d.). About MAME. Retrieved August 15, 2012, from http://mamedev.org/about.html.
Newman, J. & Simons, I. (2009). Make videogames history: Game preservation and the National Videogame Archive. Digra.org. Retrieved August 15, 2012, from www.digra.org/dl/display_html?chid=09287.32127.pdf.
Pelletier-Gagnon, Jérémie (2012), La langue comme marqueur générique. Réaffectation de la langue japonaise dans la traduction amateur du visual novel. Kinephanos, 3(1), 70–91. Retrieved August 15, 2012, from www.kinephanos.ca/2012/visual-novel/.
Perron, B. & Wolf, M. J. P. (2008). Introduction. In B. Perron & M. J. P. Wolf (Eds.), The Video Game Theory Reader 2 (pp. 1–21). New York: Routledge.
Therrien, C. (2012). Video games caught up in history: Accessibility, teleological distortion, and other methodological issues. In M. J. P. Wolf (Ed.), Before the Crash: Early Video Game History (pp. 9–29). Detroit: Wayne State University Press.
Winget, M. A. & Murray, C. (2008). Collecting and preserving videogames and their related materials: A review of current practice, game-related archives and research projects. Arxiv.org. Retrieved August 15, 2012, from http://arxiv.org/abs/0811.3137.