Defining an Ethical Framework for Preserving Cory Arcangel’s Super Mario Clouds

Emilie Magnin
The Electronic Media Review, Volume Four: 2015-2016 


The conservation issues raised by complex time-based media installations are many and related not only to the inherent obsolescence of equipment and technologies, but also to the conceptual and aesthetic implications of change, including change resulting from conservation intervention. Therefore, defining an ethical framework in which conservators can operate becomes imperative for sustaining the collection life of such artworks.

This paper, which is based on a Master thesis presented at the University of the Arts in Bern (Switzerland), uses Cory Arcangel’s video game-based installation Super Mario Clouds (2003) as a case study to explore the systematic development of a conservation strategy against the background of current ethical considerations in the emerging field of time-based media conservation.

The research discusses several conservation approaches such as replacing equipment, digitization and emulation, and investigates how they can be applied to Super Mario Clouds over the short to long-term. These approaches are evaluated and compared with established conservation ethics on the one hand, and the artist-proposed strategies on the other. Additionally, it exposes the results of a number of practical tests that have been conducted to assess the feasibility of the different strategies.


Super Mario Clouds (2003) is a video game-based installation by the American artist Cory Arcangel (b. 1978). It consists of the famous 1980s Nintendo game Super Mario Bros. (1985),which the artist has modified. By rewriting the game’s proprietary code, the artist erased every element of the game, leaving only the white clouds passing through a blue-sky background. The new code is written on an EPROM chip (Electrically Programmable Read-Only Memory chip) that is soldered inside the Nintendo game cartridge in place of the original program chip.

When the artwork is exhibited, the modified game cartridge is played back with a Nintendo Entertainment System (NES) console and then projected on the walls and displayed on a monitor via the console’s composite video output. The equipment and cables must remain visible on the floor of the exhibition space. Depending on the size of the room, the number of projections and monitors can vary, but most iterations use two to three projections and one monitor, all placed on the floor (fig. 1).

Fig. 1 Cory Arcangel, Super Mario Clouds v2k3, 2003, dimensions variable. Edition no. 2/5, Whitney Museum of American Art, New York. Installation view, Synthetic, Whitney Museum of American Art, 2009. © Cory Arcangel. Courtesy of the artist and Team Gallery, New York.
Fig. 1 Cory Arcangel, Super Mario Clouds v2k3, 2003, dimensions variable. Edition no. 2/5, Whitney Museum of American Art, New York. Installation view, Synthetic, Whitney Museum of American Art, 2009. © Cory Arcangel. Courtesy of the artist and Team Gallery, New York.


Super Mario Clouds exists in multiple manifestations. Arcangel first released the work as an Internet piece in 2002, when he put his modified code and a tutorial on his personal website (Arcangel 2013a), explaining to the online audience—which at the time of release was mostly composed of hackers and gamers—how to hack their own cartridge. He also provided a gif animation to illustrate the work, later replaced by a video (Rinehart 2010). The artist-modified code, the modification instructions, and the gif constitute the 2002 “Internet version” of the work. The artist’s website also provides a downloadable version of the piece, or ROM file,1 that can be run live on an emulator.  Whether this emulated version was already released in 2002 could not be established with certainty, but the artist mentioned it as early as 2005 in an interview for Electronic Arts Intermix (Arcangel 2005).

The better-known installation, or “gallery version” of the artwork, on which this paper focuses, was created in 2003 for an exhibition at the Team Gallery in New York. On this occasion, the artist decided to use several projectors and create a multi-projection installation that the gallery later sold as an editioned work.

Today, both versions continue to exist in parallel, in different contexts and economies: one version is distributed freely online, while the other has entered the art market in an edition of five, as an installed artwork coming with specific installation instructions. The history of the work’s creation as two parallel versions—the Internet version and the gallery version—impacts the way the work is perceived and must be considered when discussing the notions of authenticity and variability for Super Mario Clouds.

Apart from its context of creation, another important aspect that aides in grasping the work’s identity is its exhibition history. The installed version from 2003 has been exhibited more than 17 times since its creation and presents many variations between iterations, depending on the exhibition space or on specific artistic or curatorial choices. Comparing the different iterations is a precious tool for conservators to see the evolution of the work, and to comprehend its variability. What makes this case study particularly interesting is that Arcangel has often been involved when reinstalling his work, and abundantly commented on the iterations that he considered successful or unsuccessful through several interviews (Arcangel 2005, 2013b, 2014; The Variable Media Questionnaire 2008). A comparison of these sources—which almost cover the twelve years that have passed since the creation of the piece—shows that the artist has always been concerned with the way his work was exhibited, and always wanted to be involved in decisions about its re-installation.


The history of the work’s creation, as well as the comparison between the many iterations, allows for better comprehension of the variability of the work and to investigate what comprises its authenticity. These elements also help to establish the significance of the different constituents of the artwork, leading to the development of a case-specific preservation plan.

Establishing the work-defining parameters of an artwork is a process that helps in understanding what defines the authenticity of the work, and how it is allowed to change. Laurenson (2008, 7) recommends thoroughly documenting the “work-defining properties” of an artwork as means of defining a work’s identity. These are “plans and specifications demarcating the parameters of possible change, display equipment, acoustic and aural properties, light levels, the way the public encounters the work and the means by which the time-based media element is played back” (Laurenson 2008, 7). Such information can be provided by the artist in more or less detailed installation instructions. Alternatively, it is collected and completed by conservators, preferably while acquiring and installing the piece.

Several sources provide valuable information to assess Super Mario Clouds’ defining properties. These include the installation instructions put together by the Migros Museum in Zurich (Schori 2006), the author’s personal research on the technique and history of the piece, the author’s interview of the artist (Arcangel 2014), as well as information collected from his aforementioned earlier interviews.

Since an installation is a complex object where many elements interact, the work can be fully experienced only if all its components are interacting in the way they were meant to. Therefore, all these elements are important in order to define the work’s authenticity and their individual significance must be established to better comprehend the artwork as a whole. Particularly significant components of Super Mario Clouds comprise the artist code, the Nintendo hardware, the display equipment and the image characteristics that are typical for early video games.


The first significant element is the artist-written code, which Arcangel himself considers as the “original element” of his work (Arcangel 2005). Multiple values can be attributed to the code: its functional and conceptual value makes it a core element of the piece, considered as the most original component by the artist himself; while being an invisible element in itself, the code also determines the aesthetic properties of the visual content.

Additionally, the research potential offered by the artist-written code is also essential for the comprehension of the work, both on aesthetic and conceptual levels. Studying an artist’s code gives conservators an insight into the process of creation and informs specific decisions regarding design or functionality (Engel and Wharton 2014). Specific operations decided by the artist as well as his comments next to the code lines can be considered as an equivalent of the artist’s handwriting, where his individual style and creative process are revealed. The fact that the artist also published his code in text form enhances its value, while facilitating and encouraging research.

The code defines how the work looks and behaves, making it the core element of the piece. But in order to be functional, the code needs to be interpreted and performed by appropriate elements of software and hardware.


The values attributed to the Nintendo hardware by conservators are also manifold. It is functional because the cartridge works as the image carrier and the game console as the playback device, transmitting the video signal to the monitor and projectors. Second, the fact that the equipment remains visible, placed in the middle of the room gives it an aesthetic value. Furthermore, the deliberate choice of using a NES video game, a technology that was already obsolete at the time of creation, attributes a conceptual value to the hardware. In this sense, the NES console acts not only as a playback source, but also as an indicator for the audience to recognize that the work is indeed a video game that you cannot play, and not a simple video projection. Additionally, the game cartridge is hand-modified by the artist and therefore has a unique character, making it less replaceable than the game console (fig. 2).

Fig. 2 Original artist-modified cartridge. The plastic casing was cut open to accommodate the new EPROM chip. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 2 Original artist-modified cartridge. The plastic casing was cut open to accommodate the new EPROM chip. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.

Finally, the fact that the equipment is becoming rarer due to obsolescence is adding a historic value to the hardware. Since the Nintendo Company stopped producing NES devices almost two decades ago, their presence in the exhibition space acts as a bond to a specific moment in our recent cultural history.


Even though the artist does not require specific make and models of display devices, specific properties of the equipment are significant for the work, and their obsolescence might render the work more vulnerable. For example, cathode-ray tube (CRT) monitors have a typical cubic form, a 4:3 aspect ratio and a particular image quality, all which influence the aesthetic of the installation. Since CRT technology also calls to mind the period when early Nintendo games were in use, the value attributed to the CRT technology is simultaneously functional, aesthetical, and historical.

Finally, the constellation of equipment inside the exhibition space is also an important parameter of the work. The classical way of showing the piece usually combines one CRT monitor placed on the floor with two or three projections, all displaying the image content from the same source (one channel). But sometimes Arcangel showed the piece as projections only, or on a single monitor. Understanding what motivates his decision, what the effect is that he wants to achieve when posing the display sources in the space, or what the relation is between the number of display sources and the space is capital to re-installing the work in the future.


Factors such as color space, speed, and aspect ratio all determine the specific look and feel of the work’s visual content, which is representative of 8-bit era video games. A modification of these parameters would then not only impact the image quality, but also make the reference to the original game Super Mario Bros. less recognizable.

Thus, determined by the artist-written code, specific characteristics of the image are also highly dependent on the display device and the type of video connector. It is the entire system—code, Nintendo cartridge and console, connector, and monitor or projector—that affects the final image. To preserve the original aesthetic of the work, the specific qualities of the image need to be understood and documented thoroughly in their native environment.

Identifying the significant elements of the work, and how they interact together led to the conclusion that since all the physical components operate together as a system, a change to one component would have consequences on the entire system. For example, the artist code needs to be executed by the hardware and displayed by the monitor and projectors in order to make the image content visible. Also, the reference to 8-bit video games stems all at once from the functionality and physical presence of the Nintendo, but also from the resulting visual characteristics of the displayed image. In the end, the interaction of these elements in the exhibition space and the way they are experienced as a whole are as important as the individual constituents of the installation. Hence, there is a need to analyze the acceptability of different changes linked to these constituents, along with the degree beyond which change would be considered loss for the piece’s authenticity.


As are other time-based media artworks, Super Mario Clouds is subject to a variety of inherent risks that can often be avoided or minimized, but only for a certain amount of time. Material risks such as equipment deterioration and increasing difficulties in finding replacement parts might compromise the work’s functionality and visual appearance in the future. Even if the code itself—the core element of the work—can be preserved without too many difficulties, the work’s integrity remains potentially threatened by the rapid obsolescence of the hardware and software required to access and execute its code, and by the changes in display technologies and resolution.

Furthermore, immaterial risks linked to the variable nature of the piece and to the difficulties of transmitting its historical and cultural context for future generations might jeopardize the integrity of future iterations. As cultural references change, there is a risk of losing contextual knowledge about the piece. For example, a future audience might not recognize the Nintendo hardware or the game aesthetic as we still do today.

These risks are inherent to the variable nature of the work, so they cannot be completely avoided and some change will have to be accepted in the long-term. But the loss can often be minimized or compensated for with appropriate conservation strategies such as replacement, digitization and emulation, and with thorough, varied documentation.


This chapter will evaluate possible strategies to preserve Super Mario Clouds while considering the artist’s intent, the risks that the work is subjected to, and the ethical framework that guides the conservation profession. In order to find the best approach, it will explore not only the established time-based media conservation methods, but also strategies cultivated in relevant adjacent fields, such as software-based art conservation, video game preservation, and gaming communities.

Keeping a time-based media artwork functioning and accessible through time is a dynamic process, change being inherent to it. Often multiple strategies are ethically acceptable, yet eachof them results in specific changes for the artwork, which must be rigorously considered in anticipation. Also, the most conservative approaches, such as replacing equipment by the same make and model are not feasible in the long term. For all these reasons, it is not a unique conservation strategy, but rather a number of strategies ranging from short-term to long-term efficacy that must be implemented. In general, conservation suggests favoring the option that introduces the least change and best respects the artist’s intention, but sometimes alternatives have to be weighed, especially if different stakeholders have different preferences.

In the case of Super Mario Clouds, conservators’ decision-making process can be informed by a rich exhibition history, as well as the artist’s availability to share his thoughts and to be involved in the re-installation process. Since the work has often been exhibited in a relatively early stage of its life, its demonstrated flexibility is well-documented and informs future re-installations. The role of the conservator when choosing a suitable strategy is to assess when the change or loss is so disruptive that it impairs the artwork’s meaning. The decision-making process must weigh the gain (recovery from loss) with the additional change induced by the treatment. As stated by Hummelen and Sillé (2005, 405), “the solution is inevitably a compromise between meaning and condition.” Hence, with variable artworks, it is important to remember when and why one decision has been taken, and to realize that this was probably not the only possibility. Transmitting the understanding of the work’s flexibility and mentioning alternative conservation options should avoid the fixing of the work in a single interpretation.


While replacing broken devices with identical ones is a fully acceptable strategy for conserving time-based media artworks (Laurenson 2007), the exchange of devices for other makes and models or for newer technologies should be approached with caution. In the case of Super Mario Clouds, Nintendo consoles can only be replaced with identical devices, since their technology is dedicated, and there are no existing alternative devices that can play the proprietary game cartridges.

Exchanging or updating projectors is less problematic, but the image displayed by new models should always be compared to older, artist-approved versions in case unforeseen changes in image quality might occur. For instance, current projectors still support 4:3 aspect ratio, but it is possible that at some point in the future it will no longer be the case.

As of today, a bigger issue is the replacement of CRT monitors. Since the technology is obsolete, it will sooner or later be impossible to find functioning CRT monitors. Thus, they will have to be replaced with modern screens, probably LCD. The image quality will therefore be very different, and the original 4:3 aspect ratio will not be sustained. Once the change appears inevitable, the artist should be asked whether he considers these changes acceptable or too disruptive.

The replacement of the game cartridge is also problematic, because it has been hand-modified by the artist and conservators tend to consider it as a unique object. But the artist encourages the replication of a hacked cartridge by conservators following his published instructions, which appears to be an ethically acceptable measure. Indeed, reconstructing the cartridge does not have any influence on the aesthetical qualities of the work, because the cartridge is only the physical support where the code is stored, or the image carrier. Since the artist-written code can be copied onto a new chip without any loss (cloned), any possible copies would be exactly identical. The recreation process can then be compared to the generation of new exhibition copies on DVD for video artworks, except for the fact that the hand-modification that the artist performed on the physical media carrier has to be repeated by the conservator. Therefore, the artist-modified cartridge must be considered as original and kept for documentation and research purposes. But since the cartridge is not visible once inserted in the game console, using a replicated cartridge during exhibition will not impact the artwork in any way.

In the end, the main problem with the strategy of replacing equipment is that the required material is becoming obsolete, and it will get harder to find functioning game consoles and cartridges in the future. For this reason, the maintenance and replacement of equipment is only a short to midterm solution.


Since the Nintendo console outputs a video signal, it can be captured to create a video file from the work. Creating a video from a game is indeed a strategy that is already used to document the aesthetics of video games, in addition to other documentation forms, and gives good results in terms of reproducing the image quality (Guttenbrunner, Becker, and Rauber 2010; MacDonough 2014). Typically, the video game output signal can be digitized for documentation of the original look and feel (color, speed, aspect ratio) and to have a reliable point of comparison to verify an emulator’s accuracy. But since this approach does not maintain the interactivity specific to video games, Guttenbrunner, Becker, and Rauber (2010) recommend using it only as a means of documenting the game’s audio and image characteristics.

For Super Mario Clouds on the other hand, digitization remains a viable option since the artist deleted all interactivity from the game. However, while it is technically feasible to display the work as a video, this solution would make the original code useless, which goes strongly against the conceptual intent of the artist. For this reason, it is recommended to only use digitization as a documentation strategy, but not as an exhibition format.


Emulation is a conservation strategy consisting in “imitating the original look of the piece by completely different means” (Depocas, Ippolito, and Jones 2003, 125). The use of emulation as a preservation strategy for time-based media art and software-based art in particular has been discussed by Depocas, Ippolito, and Jones (2003) and Wijers (2011), among others. Recently, authors such as Guttenbrunner, Becker, and Rauber (2010) and Romero (2012) have been exploring the possibility of emulating video games within a museum or archive environment. In this context, the term emulation takes the more specific sense of utilizing a software emulator that has been programmed to emulate video games. Such an emulator is “a program that virtually recreates a system that is different to the one on which it is running” (Guttenbrunner, Becker, and Rauber 2010, 77), allowing one to play video games on a different, and otherwise incompatible hardware system. A great advantage of using emulation for video games is that since it is the hardware that is emulated, the game data remains unchanged, which is especially relevant for video games that have been created or modified by artists.

Video Game Emulators

There are a wide variety of existing game emulators that could potentially be used by conservators for preserving obsolete video games or game art. Using already existing emulators for presentation and preservation purposes presents both advantages and disadvantages. The first advantage is of course that institutions will not have to undertake the costs of developing a new emulation tool. Also, most of the available emulators are open-source software, which is considered “better suited to preservation” (Wijers 2011, 87) and allows for case-specific improvements.

On the downside, since the majority of video game emulators are developed freely by online gaming community members, their main goal is not preservation or sustainability, but only the quality of the gaming experience. This bears the risk of introducing unwanted features. Guttenbrunner, Becker, and Rauber (2010, 87) also note that since emulators are not produced commercially, they “tend to receive few updates and are frequently discontinued when the authors become distracted from development. Therefore, hardly any emulators exist in a final version that perfectly emulates all games for a system.” Hence, emulators need to be tested, evaluated, and sometimes modified to fulfill the intent of preservation.

In addition, a major issue for using such emulators remains copyright infringement. Even though emulation can be seen as a way to keep obsolete games alive, the Nintendo Company is strongly opposed to their use and refuses to release its old games’ copyrights (Nintendo of America Inc. 2014). Nevertheless, since Super Mario Clouds has the status of an artwork, the artist owns the copyright for his creation. Therefore, using a video game emulator is a valid option for displaying the work in the future. The Nintendo Company could argue that the creation process of the piece itself constitutes a copyright infringement (in this regard, copyright issues would not concern the use of an emulator, but rather the work itself), but the work has been very mediatized since its creation and it has not led to any legal confrontations so far.

Using an Emulator for Super Mario Clouds

The main advantage of using emulation is to keep the artist code unchanged and functional when the original hardware is no longer available. Criteria for choosing an emulator to play Super Mario Clouds should be the speed and color accuracy, support of the artist-provided ROM file format, and privileging open-source and preferably widespread and regularly maintained software. Other criteria that are considered outside of the conservation field such as the gameplay experience and the sound quality are both irrelevant for this particular piece.


In order to compare the potential conservation strategies described before, three different playback options have been tested: (1) artist-provided Nintendo cartridge and console, (2) digitized video file and media player, and (3) artist-provided ROM file (downloaded from his website) and emulator. The aim of these tests was to compare how image characteristics such as color, aspect ratio, speed, and overall image quality behave when played back from different types of sources.

Test Protocol

The three setups were tested using a Panasonic PTAX200E 3 LCD projector as display device because this is the make and model that the Migros Museum in Zurich is currently using to display it. The original Nintendo system was hooked up to the projector using the composite video output.

The signal was digitized from the artist-provided Nintendo console and cartridge with an A/D converter deck using the composite video output of the Nintendo console. It was captured in Final Cut Pro as an NTSC 10-bit 4:2:2 uncompressed QuickTime movie and then converted to MPEG4 for testing in exhibition conditions. It was played on a WD media player connected to the projector via composite.

The video game emulator chosen was the v1.4.1 version of Nestopia2 for Mac OS X. This emulator is often reviewed as the best emulator for NES, with good speed accuracy. It was running on a MacBook Pro and connected to the projector via Thunderbolt to VGA3.

Test Results

The video file showed few differences with the original Nintendo setup except for a smoother, flatter pixel structure (especially noticeable around the clouds) and a more tightened line structure in the border area. But the overall image quality seemed representative enough of the original effect to be used as documentation.

The comparison between the original Nintendo setup and the emulated version showed more prominent discrepancies. When using the default settings of the Nestopia v1.4.1 emulator, the image resolution is 256 x 240, which is squarer than a normal 4:3 aspect ratio. To obtain a regular 4:3 aspect ratio with the emulator, it is necessary to enable the aspect ratio correction setting under the Preferences menu. However, the emulated clouds still appear slightly flattened. Besides, the overall appearance of the image looks different: the original Nintendo version shows a visible pixel structure in the blue background, and the movement of the clouds is stuttery (fig. 3).

Fig. 3 Photograph of the projected image as played out from NES. The blue background shows a light structure, the black contour is thicker. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 3 Photograph of the projected image as played out from NES. The blue background shows a light structure, the black contour is thicker. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.

The image quality of 8-bit video games is recognizable in these details. The emulated image is flatter, smoother, the pixel structure is not visible and the stuttering less noticeable. It gives a much more ‘digital’ impression (fig. 4).

Fig. 4 Photograph of the projected image from Nestopia emulator. The blue background is perfectly smooth, the black contour is thinner, and the cloud appears flatter. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 4 Photograph of the projected image from Nestopia emulator. The blue background is perfectly smooth, the black contour is thinner, and the cloud appears flatter. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.

Some emulators, like Nestopia, offer additional filters that are supposed to imitate the original Nintendo artifacts. Enabling the “bilinear filtering” or the “simulate NTSC artifacts” options of the Nestopia emulator does make the image less sharp and digital looking, but still different from the original. In theory, if emulation were to be chosen as a strategy such filters could potentially be improved by programmers to reproduce particular image characteristics. Additionally, one major difference remains: on the image displayed by the original Nintendo system the clouds end at the blue bar before the edge (fig. 5), while with emulation the clouds go all the way to the edge of the image (fig. 6).

Fig. 5 Photograph of the projected image as played out from NES. There are blue bars on both sides of the image. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 5 Photograph of the projected image as played out from NES. There are blue bars on both sides of the image. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 6 Photograph of the projected image from Nestopia emulator. The clouds go all the way to the edges of the image. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.
Fig. 6 Photograph of the projected image from Nestopia emulator. The clouds go all the way to the edges of the image. Photo: Emilie Magnin. Courtesy of the Migros Museum für Gegenwartskunst.


In August 2015 as the work was exhibited in the Migros Museum in Zurich, further tests were conducted in order to compare the proposed strategies in exhibition conditions, and especially to test different emulators on the same equipment. The emulators tested comprised Nestopia and OpenEmu played on a Mac Mini, and the emulation system RetroPie installed on a Raspberry Pi 2, all connected to the projector via HDMI. These tests showed that different emulators produce different images, with discrepancies observed in color, aspect ratio and speed (fig. 7).

Fig. 7 Comparison of the projected image as played out from NES and from RetroPie, Nestopia and OpenEmu emulators showing discrepancies in color, aspect ratio and shape of the clouds. Photo: Emilie Magnin. Editing: Agathe Jarczyk.
Fig. 7 Comparison of the projected image as played out from NES and from RetroPie, Nestopia and OpenEmu emulators showing discrepancies in color, aspect ratio and shape of the clouds. Photo: Emilie Magnin. Editing: Agathe Jarczyk.


These practical investigations allowed the author to assess the feasibility of the different strategies and to compare their outcomes in exhibition-like conditions. First, the tests showed that the digitization process is not as straightforward as it was assumed. It appeared that digitizing only functioned when using an older generation of A/D converters (such as AJA Io). Tests made with a more recent device (Blackmagic Decklink capture card) showed that the video signal was not detected as a standard NTSC signal, and required using a standard converter. This circumstance might compromise digitization in the future.

Second, comparing the image quality of the various playback sources highlighted the discrepancies between the different playback options that had been envisioned as potential conservation strategies and stressed the importance of documenting the original image characteristics while the Nintendo equipment still is functioning. It especially revealed that emulation induces quite an important change in the image quality and can differ from emulator to emulator.


Regardless of the strategy chosen, many aspects of the re-installation process still need to be addressed in order to produce installations that will respect the work’s integrity and that will also convey the artist’s intention to future audiences. Since such decisions have a big impact on the work’s appearance, having the artist involved in the decision-making process is an advantage for conservators. Next to expressing his views on re-installing his work through different interviews, Arcangel is also experimenting with different ways of displaying his old works in other forms and formats, or by incorporating them into newer works as part of his own artistic practice. The strategies that have been envisaged by the artist for future displays of Super Mario Clouds are coherent with current practices in time-based media conservation, but they might differ in their degree of application, as conservators often have different duties and agendas than the artist. It is in the artist’s best interest to update his earlier work, in order to make it look more contemporary and corresponding to his evolving ideas on how the work should look (Sommermeyer 2011). On the other hand, a museum’s task involves “contextualizing an artwork within an artist’s career” and within its specific time and conditions of creation (Phillips 2012, 141). Museums will then tend to present the work with respect for its historical context, while the artist might wish to offer a more contemporary display.

However, in order to keep a media artwork functioning, updating it to new technologies can be a necessary part of the conservation strategy when maintenance and repair is no longer possible. In such cases, the artist’s desire to keep his work current is not incompatible with conservation’s goals.


Several important points emerged from this research. First, it became clear that the Nintendo technology has a capital value for Cory Arcangel’s Super Mario Clouds, not only on a functional but also on an aesthetic and a conceptual level. For this reason, this technology should be maintained in use for as long as possible.

What also became clear was the fact that the artist considers his code to be the original core element of the work. It is crucial that the integrity and functionality of this code should be preserved. Keeping the code functional (and along with it, the artwork’s reference to video games) should then be the prevailing priority when choosing a conservation strategy. In this sense, emulation appears as a better strategy than digitization since it imitates the old gaming system and thereby keeps the original artist’s code functioning. It is also the strategy that the artist prefers. Digitization, however, proved to be well suited for archival purposes, for it can document the image content, and may be used as an element of comparison for assessing the image quality produced by different emulators.

Finally, the research showed that in order to maintain Super Mario Clouds fully functioning on the long-term, any adopted strategy would have to be reviewed regularly. For this purpose, practical tests and image characteristic comparisons can be used as guidelines to help the decision-making process when a new strategy needs to be adopted. Additionally, conservators can benefit from the artist’s personal experimentations to see how the work reacts against newer technologies.

In the end, let’s not forget that each strategy will induce some new change to the work, and that all conservation decisions will reflect some subjectivity. The appreciation of what can change and to which extent is a key question to decide which strategy is best suited for Super Mario Clouds at every stage of its lifetime, and such decisions should always be taken in cooperation with all involved parties.


Many thanks to conservators Agathe Jarczyk (University of the Arts, Bern) and Joanna Phillips (Solomon R. Guggenheim Museum, New York) for their long-lasting support and commitment throughout my Master thesis and after.


1. A ROM (Read Only Memory) file contains an exact copy of the game that was contained on an original gaming cartridge (though in this case, the original game content was modified by the artist) and can be played with a video game emulator, allowing to play the game directly on a computer, without the original game system or game cartridge itself.

2. Nestopia is an open-source NES emulator written in C++. The Nestopia v1.4.1 version for Mac OS X was downloaded from: (accessed 08/20/2015).

3. This last setup was not optimal, as it does not reflect current exhibition practices. Further tests were conducted in August 2015 using Nestopia v1.4.1 running on a Mac Mini and wired to the same projector via HDMI. They showed comparable results in terms of image quality.


Arcangel, C. 2005. Interview with Lori Zippay. Electronic Arts Intermix (EAI). August 18. (accessed 08/20/2015).

Arcangel, C. 2013a. Cory Arcangel’s Official Portfolio Website and Portal. (accessed 08/20/2015).

Arcangel, C. 2013b. Interview with Crystal Sanchez and Claire Eckert. Smithsonian Institution Time-Based and Digital Art Working Group. July 7. (accessed 08/20/2015).

Arcangel, C. 2014. Interview by author. Transcript. Artist’s Studio, Brooklyn, NY, February 06.

Depocas, A., J. Ippolito, and C. Jones, eds. 2003. Permanence Through Change, The Variable Media Approach. New York: The Solomon R. Guggenheim Foundation and Montreal: The Daniel Langlois Foundation for Art, Science and Technology. Available at (accessed 08/20/2015).

Engel, D. and Wharton, G. 2014. Reading Between the Lines: Source Code Documentation as a Conservation Strategy for Software-based Art. Studies in Conservation 59(6): 404–415.

Guttenbrunner, M., C. Becker, and A. Rauber. 2010. Keeping the Game Alive: Evaluating Strategies for the Preservation of Console Video Games. The International Journal of Digital Curation 1(5): 64–90.

Hummelen, I. and D. Sillé, eds. 2005. Modern Art – Who Cares? : An interdisciplinary research project and an international symposium on the conservation of modern and contemporary art. London: Archetype.

Laurenson, P. 2007. Research on preservation strategies, part 1: Risk Assessment. detail.php?r_id=83&ct=preservation (accessed 07/15/2014).

Laurenson, P. 2008. Authenticity, Change and Loss in the Conservation of Time-based Media Installations. In (Im) permanence: cultures in/out of time, eds. J. Schachter and S. Brockmann, Pittsburgh: Center for the Arts in Society, Carnegie Mellon University. 150–164. Available at tatepapers/06autumn/laurenson.htm (accessed 08/20/2015).

MacDonough, K. 2014. Conserving Portal: Defining and Documenting the Object. History of Games International Conference Proceedings. Montreal, Canada. (accessed 08/20/2015).

Nintendo of America Inc. 2014. Legal Information (Copyrights, Emulators, ROMs, etc.), Redmond. (accessed 08/20/2015).

Phillips, J. 2012. Shifting Equipment Significance in Time- Based Media Art. The Electronic Media Review 1: 139–154.

Rinehart, R. 2010. In Defense of Blue Sky Art. defense_blue_sky.pdf (accessed 08/20/2015).

Romero, C. 2012. Long March Restart: the Acquisition, Display and Conservation of Video Games. Unpublished Master Thesis, Graduate School of Arts and Science, New York University.

Schori, M. 2006. Technisches Manual für Cory Arcangel Super Landscape 1. Unpublished manual, Sammlung Migros Museum für Gegenwartskunst, Zürich.

Sommermeyer, B. 2011. Who’s Right – the Artist or the Conservator? In Inside Installations – Theory and Practice in the Care of Complex Artworks, eds T. Scholte and G. Wharton, Amsterdam : University Press. 143–151.

The Variable Media Questionnaire. 2008. Super Mario Clouds, 2002. (login required, accessed 08/20/2015).

Wijers, G. 2011. To Emulate or Not – Conservation Case Studies from the Netherlands. In Inside Installations – Theory and Practice in the Care of Complex Artworks, eds. T. Scholte and G. Wharton, Amsterdam: University Press. 81–89.


AIC. 1994. Code of Ethics. documents/code-of-ethics#. (accessed 08/20/2015).

Emilie Magnin
Contemporary Art conservator MA
Conservation and Restoration of Contemporary Art
3001 Bern, Switzerland