Friedrich Kirschner,

Introduction
The development of machinima and its perceived potential has, on many occasions, been linked to its production tools, underlying technology and distribution methods.
This is due to the shape that machinima related modifications usually take - mainly: they serve as a way of distributing machinima for playback, they change the game assets and visual style, or they are themselves tools for machinima creation. While the process of distributing machinima as a game modification has all but vanished, tool and asset modifications were not only responsible for the growth and original inception of machinima, but represent an essential aspect of it to this date.
The modding community was in place before machinima was common practice. (Lowood, 2008) Its ways of reconstructing video games, as I will argue, led to the emergence and shape of machinima.
In creating tools and workflows for creative misuse of provided technology, the modding community not only changed video games itself, but also moved beyond the idea of user-generated content. It effectively created its own form of creative expression that became machinima. I call this development user-generated process. It constitutes a new quality in user involvement that moves beyond modification and optimization of a given product into the birth of new processes and even new media. To examine machinima’s development in relation to the use and implementation of game modifications, let us look at three distinct historical phases: (1) machinima’s initial conception within the “Quake Movie” community, (2) machinima’s emergence in mainstream media and (3) the current state of machinima production and distribution. I will place particular emphasis on the second phase. It constitutes a time in which there was much discussion about machinima’s potential and a strong community with what Marino calls an “Outside In” approach – filmmakers exploring machinima as a new emerging medium, and not primarily interested in the narrative of the video game. (Marino, 2005)
Phase 1: The formative years
The term machinima originated from the “Quake movie” community, a group of people using the games “Quake” (ID software, 1996) and “Quake 2” (ID software, 1997) not for gameplay, but to film movies (Marino, 2004). These early machinima movies relied heavily on user-created programs to change the basic data structures of the games’ demo-files. A demo file contained recordings of events and user-input happening during a game. The way in which these first movies were recorded was very performative in nature. Characters were controlled live by the players, and become the basis of this demo file. Cameras and editing were either set up later in a process known as “re-camming”, or controlled by commands that would switch the camera to pre-defined positions. The emergence of live-performed Machinima, as in the case of the ILL-Clan’s “Larry and Lenny Lumberjack” (ILL Clan, 2000), or the “Bob Block Machinima Show” (Kirschner, Neumann, Scholz 2004) seemed like a natural form emerging from this mode of production. Quake movies made extensive use of user-generated tools to modify the demo files. In the case of one of the most important tools, Keygrip, the creator was well situated within the modding scene:
“These coding exploits provided a foundation for David “CRT” Wright’s influential Keygrip and Keygrip2 programs, which became the most widely-used utilities for editing and other post-production work on machinima movies. Wright was a mod coder known among Quake players for the Rocket Arena series of one-on-one dueling games... “ (Lowood, 2005, p.68)
Another important aspect of these early Machinima works was the way they were distributed and played back. Instead of being packaged as video files, the work was usually distributed as a demo file and played back within a copy of the game itself, on the viewers’ own computer. This required that modifications to environments, characters, scripts and textures had to be installed on the computer of the viewer as well. Machinima was effectively presented as a form of game modification. Machinima’s potential for transforming the way animated movies were created was readily visible to early practitioners (Kirschner, 2011). This potential was closely related to the early tools, distribution and production processes laid out by key Modifications such as the aforementioned “Keygrip”. While the production method was deemed revolutionary and full of potential, the creative choices left to the Machinima producers in terms of characters, textures or animations were often meager. In addition, the games themselves provided no options for creating these types of assets outside of import/export scripts for the content creation tools used in the production of the games themselves.
The community, often lacking access to these programs, used affordable software like Milkshape 3D – a shareware, low-polygon modeling and animation program (Chumbalum Software, 1996) - to modify or create
Figure 1; The Bob Block Machinima Show (Kirschner, Neumann, Scholz, 2004)
animations and models to import into Quake and other games. Milkshape in particular was well situated in the modding community, with users often providing importers and exporters for a number of games. In Machinima’s formative years, the Modding Community went to great lengths to provide tools for user generated content that went above and beyond the infrastructure provided by video-game production companies. For example, the ILL-Clan created software that allowed their Quake 2 characters to have changing textures for the different stages of their mouth movements, allowing real-time lip-syncing of video-game characters. Some of these tools were not initially intended for Machinima purposes, but were nonetheless crucial in its inception. This reliance on community tools and modifications allowed Machinima’s unique production process to be established.
Phase 2: Machinima Renaissance
In what Paul Marino dubbed a “Machinima Renaissance” in his book “the Art of Machinima” (Marino, 2004), a new generation of movies shed their video-game-style graphics ties and presented themselves as visually independent animated films that gained significant mainstream exposure. This newly found graphical fidelity came alongside a new generation of video-game technology, which in turn provided a new generation of tools and creative options:
“The release of new development tools greatly ushered in new approaches to digital filmmaking” (Marino, 2004).
Some of these new games and tools exposed film-recording functionality. Instead of implementing functionality as laid out by the modding community though, most games were straying away from the early production approach of the Quake movie era, instead introducing a scripted approach suited for cut-scene creation and embedding character and camera control into user-generated levels. Notable example games include “the Sims 2” developed by Maxis (2004) and Unreal Tournament 2004 by Epic Games (2004). Many Machinima movies of this “second phase” could be seen as a direct modification of the game itself. As the games used to create movies stepped away from the demo format and provided tools for procedural scripting of characters and cameras, some of the initial embedded potential lost focus - the idea of live performance, for example, became less and less prevalent. Still, Modifications were employed as a means of distribution, sharing of assets and in the form of tool-sets for producing Machinima. Filmmakers were excited about these new games and their possibilities. Prominent Machinima creator Ken Thain, director of the music video “Rebel vs. Thug” (Hanson 2004)
summarized this notion in an interview for Machinimag:
“There is no one faucet of this medium that is not boiling over with possibilities. The real-time playback aspect, the packaged content aspect, the neo-production possibilities, the high speed of technology advancement, etc, etc. - all that gravy” (Thain, in Kirschner, 2004, p.7)
An indicator of how far Machinima had come in this second phase was the inclusion of Machinima categories in major film and animation-festivals (such as the Bitfilm Festival 2003 or the Ottawa International Animation Festival 2006) and the emergence of contests and competitions with significant incentives for filmmakers to participate, such as the 1,000,000 $ Make Something Unreal Contest.
The Make Something Unreal Contest
The Make Something Unreal Contest (2003) was set up by Nvidia and Epic Games with significant resources to reward excellence in mod making in several categories, including Non-Interactive Real Time Movies (NVIDIA Corporation, Epic Games, 2003).
To be eligible for the non-interactive movie category (the name being an indication that the word “Machinima” was still struggling to gain acceptance in the wider gaming community), the final movie had to be a modification running within the game UT2003 (and later, UT2004). This requirement, as much as Machinima’s actual inclusion as a category in the contest was a testament to its Game Modification roots as much as a sign of its potential as perceived from a games industry perspective. The contest aimed to highlight the functionality and ease of use of the then latest Unreal Tournament engine based on the game UT2003 and then later, UT2004. The toolset provided by the game greatly benefited from the inclusion of Matinee – an interface designed for the scripting of camera edits - and functionality for Character animation. It was hailed as the inclusion of moviemaking functionality in the Unreal Tournament Modding toolset provided by the creators of the game itself.
As a participant in the contest, (I was involved in the contest with two movies) the use of Matinee proved to be essential, but it became clear that its design was lacking flexibility. The overall design philosophy was quite different to that of the Demo-recording days. No live-performance was necessary, but neither was it possible. In addition, camera control and character scripting were separated in the Matinee interface. The only way to test a scene was to “run” the level. This crucial disconnect required a distinct switching from the editing environment to the un-editable game environment in which the camera could be performed, but no positions could be recorded and time could only be slowed down and not otherwise manipulated (no reverse, no fast forward, no jumping to time positions, etc...). While earlier tools like Keygrip were also limited in the way timing could be manipulated for editing, this new way of setting up things via script in an outside tool was even more limited. This made editing decisions and camera placement a lot harder than in “traditional” Machinima environments. On the other hand, the tools included a full Asset importing pipeline for all major 3D content creation tools, a scripting language and plenty of documentation. These provided more options for artistic expression than previous iterations of the engine or other games at the time.
For my contest entry “the Journey” (Kirschner 2004), I created 5 levels that served as scenes for my machinima film, creating custom props, textures, and characters, very similar to creating a game modification. However, my modified landscapes and characters were not meant to be used for play (even though they easily could have been made available as a game level), but as sets in my “real-time non-interactive movie”. During production, I realized that a complex setup like this required custom scripts not just for added effects and animation functionality, but also to add organization and structure to the filmmaking process.
Figure 2; The Journey (Kirschner, 2004)
I had to create two sets of modifications. One being the actual Mod that would be my submission for the contest, containing levels, assets and camera scripting. And another one, invisibly intertwined, that helped me create the first one: a set of simple tools that structured my scripting - sometimes as simple as just displaying a dividing line in the list of scripted actions for my characters. Many other participants were performing similar tasks, defining a workflow that would serve Machinima filmmakers as opposed to game modification.
Mods replacing tools
Conversations with other filmmakers engaged in the contest led to much discussion on programming and processes. Some of it was about replacing functionality that was provided by the given toolset, but technically too challenging or cumbersome to use. For example, while the Unreal Engine had a built in tool for facial animation, called “Impersonator”, it proved too involved a process to be useful, as described by Eric “Starfury” Bakutis, another entrant in the contest:
“[...] Impersonator is only useful to a Machinima creator if they already have a custom built model that has a fully realized facial bone structure, and phoneme animations to match... a model of this complexity is currently out of reach for the majority of amateur UT2K4 machinima makers, and as such, Impersonator is more promise than actually useful.” (Bakutis, 2004).
Bakutis was no stranger to Machinima filmmaking. He was head of the team creating “Devil’s Covenant”, a major achievement in the Quake movie scene (Bakutis, 1998) and was credited with adding to the official Unreal Development Network documentation (Lin, 2003). His solution was to create his own system for swapping textures on a character model, a way of creating lip-synching animations also used in community toolsets for Quake 2, as discussed earlier. A connection Bakutis acknowledges:
“Also, my focus is definately[sic] to make this accessible to the ‘one guy in his garage’ audience... I still remember the days of Quake2 machinima when just about anybody could make a movie without extensive modeling and coding backup.” (Bakutis, 2004)
In the end, none of the nominees for “Best non-interactive real-time movie”
actually used Impersonator. The Impersonator case shows that the existence of sophisticated tools alone does not necessarily translate into actual use, while at the same time the machinima community constantly creates their own modifications, tools and processes to aid the movie-making process. It illustrates that the conception and implementation of machinima tools requires profound knowledge of the machinima process itself, and that machinima filmmaking is in a constant state of redefining its technology and production processes. And modifications play a crucial role in forming and sharing these processes.
Investigating Machinima’s potential
My follow up work has investigated the tools and processes for machinima filmmaking in addition to the creative process of actually telling a story. At the time of the ”Make Something Unreal Contest”, the general belief was that gaming technology, tools and graphical development would lead to a new dawn in machinima filmmaking (Marino in Kirschner, 2004). That tools would get easier to use and more powerful. Or, as put into words by Henry Lowood:
“It is safe to predict that game developers will soon put robust but easy-to-use machinima tools directly into the hands of an increasing number of players” (Lowood, p.74).
Figure 3; Person 2184 (Kirschner, 2005)
To me, UT2004 was very close to being an ideal machinima production environment. Its shortcomings - described earlier - seemed fixable from a mod-maker’s standpoint. For my work, “Person2184” and “the Photographer”,
I developed a set of scripts that replaced Matinee and combined camera and character controls (Kirschner, 2004). “Script Composer” was a modification adding to UT2004’s editing environment (UnrealEd). It could be used within the UnrealEd environment meaning that it still allowed movies to be distributed as modifications themselves. This “packaged content aspect” as Thain described it, allows the machinima movies to be played back as a level using the game itself – as opposed to a movie file or streaming video. In his text “Machinima as Media”, (Nitsche, 2011) Michael Nitsche illustrates the importance of this in defining machinima as its own medium:
“Game engines provide this real-time rendering as well as other features, including levels of interactivity, that allow dynamic replay of a data file on the local machine of the player audience. 3D modeling programs do not offer such a game-based feature and thus lack the media possibilities associated with it.” (Nitsche, 2011, p.120)
As well as being a defining part of machinima filmmaking, this way of distribution also allows for a number of creative possibilities, such as non-deterministic movies that play out differently every time, and on-the-fly modification, as further illustrated by an American Film Institute project that investigated this specific aspect of machinima filmmaking (Nitsche, 2007). It also allows for the examination of the individual assets, levels and scripts, similar to having access to the project files in a 3D animation project as opposed to just watching a rendered movie. This effectively allows remixing of work, and in the case of “Person2184”, this aspect was used by students at the University of Plymouth to create original work based on assets I included in my movie:
“We also deconstructed Person2184. [...] We’ve managed to export some of the characters but they’re all in DDS format but with the correct plugin, Photoshop can work with them so we can now see how Friedrich created them.” (Saunders, 2006)
According to Nitsche, “Machinima is digital performance that controls procedurally animated moving images.” Nitsche does not just talk about real-time playback capabilities but references live interaction and performance as well (Cameron and Carrol in Nitsche, 2011). Real-time performance was impossible within the scope of UnrealEd, given its strict division of the scripting and playback process. The only way to allow for machinima performance within UT2004 while at the same time keeping scripting possibilities, was to create a total conversion modification that would bring UnrealEd’s functionality into the actual game aspect of UT2004. Moviesandbox (Kirschner, 2006), was an attempt to do just that. It implemented the same scripting possibilities as scriptcomposer beforehand, retained the “packaged content aspect”, but also allowed for live acting and puppeteering of UT2004 characters. I have written about the concepts behind Moviesandbox more extensively (Kirschner, 2011), but want to mention it here as an example of a total conversion modification for machinima filmmaking that tried to follow machinima’s potential as voiced by Thain and others, and shared by myself.
Nitsche further argues that, “Whether it was the release of the various editors by id Software, Epic’s UnrealEd, or Valve’s Source SDK with its Faceposer tool, all helped and empowered the machinima makers, commercial as well as independent. They became the artistic tools for machinima and demo production.” (Nitsche, The Machinima Reader, p. 120) I believe this statement is missing a crucial part of machinima’s development regarding its Mod heritage. The role of the Machinima filmmakers, and their frustration, in bending and extending the toolsets mentioned is crucial to understanding the turn in development that Machinima was taking. Instead of making extensive use of these advanced toolsets to create more sophisticated and individual movies, the vast majority of machinima filmmakers reverted to their own tools and modifications, that were closer to the initial workflow laid out in the Quake movie era. These workflows ultimately define machinima as “user generated process”, a re-framing of given technology into a new medium not just by working within its technical infrastructure, but also by extending it with tools and processes, elevating it to a form of media of its own. This second phase illustrates how the defining factors did not originate from the video-game developers themselves, but were still driven by community effort to shape machinima’s form and function.
Phase 3: The current state of Machinima
Bungie’s “Halo”(2001), a game for the XBOX console, and Blizzard’s “World of Warcraft” (2004) have spawned not just an incredible number of Machinima movies, but also some of the more defining works of the art form. It is safe to say that the potential that others and I saw in it did not reveal itself in the form that Thain or Marino anticipated. The vision of affordable, powerful tools developed by video-game companies that allow for total creative control by expert Machinima filmmakers implemented directly into the games or game tools does not form the basis of Machinima’s practice today. Software specifically designed for Machinima creation, such as Moviestorm (Moviestorm, 2007), iClone (Reallusion, 2003) or my own Moviesandbox (Kirschner, 2006) did not manage to gain the same amount of users, movies, press, or artistic success as the aforementioned games. Machinima.com, the social hub of the game-agnostic machinima online community is not hosting discussion forums anymore. Nitsche clearly agrees that Machinima did not take the route anticipated during the second wave:
“Popular Machinima practice clearly did not follow the strengths of its own media specifics.” (Nitsche, 2011, p.121)
One aspect has not changed over the years since its emergence, however. The perceived deviation from “the strengths of its media specifics” is missing a crucial point of what I see as a fundamental concept in machinima - modding. We know that machinima is tied to the modding scene by its heritage. In fact, Lowood sees machinima as a direct outcome of the Quake modding scene:
“Not only did providing an editor and scripting language stimulate modification and extension of the game, it encouraged the development of tools for unforeseen purposes, such as the editing of demo movies and, eventually, the making of animated movies using real-time techniques of gameplay as performance.” (Lowood, 2005, p.74)
But the connection between modifications and machinima is more than just historical. Games like Lionhead’s “the Movies” (2005), or Maxis’ the Sims 2 provided their own machinima filmmaking features, but were still modded extensively to allow for more creative user choices. The aforementioned “World of Warcraft” spawned a significant amount of modifications and best practices for machinima production. There are many more examples of mods that have spawned machinima movies in games that do not even have any dedicated machinima tools, sometimes by enabling hidden features, sometimes by brute force hacking (Hancock, 2007)
Conclusion
The definition provided by Marino, of machinima as “animated filmmaking within a real-time virtual 3D environment” (Marino, 2004) is a technical definition that does not cover the wider implications of the movement - its commentary and situation within gaming culture. Nitsche’s definition as digital performance controlling procedural imagery works against the realities of recent machinima production. Lowood’s definition of machinima as “producing animated movies with the software that is used to develop and play computer games” (Lowood, 2008) recognizes machinima’s ties to video games, but his notion of machinima as “exploiting found technology” under-represents the role that the modification process plays for machinima as a movement, and the quality of innovation that is generated by the modding and machinima community. None of these definitions emphasize modifications and user generated process as a defining aspect of machinima.
I would like to draw attention to the machinima filmmaker as creator of technology. Mods are a crucial part in the process of machinima to this date, both technically and conceptually. They planted the idea of not only user-generated content, but also user-generated process. At its core, machinima is always going to be a form of game modification. Even when stripped of all of its technological properties and tools, the very idea of using video games for filmmaking is a modification to the rules of most games – and essential to machinima’s inception. Machinima will not cease to switch engines and the machinima community will not stop creating modifications and hacks, because it is its very essence to do so. It is part of machinima’s transformative power, even hailed as democratization of filmmaking (Matlack, 2005). This user-generated process is a novel way of user innovation that extends beyond modifying media content with tools provided by content creators. Instead, it breaks down provided workflows and established production conventions and establishes a new art form in itself: machinima.

References
Lowood, H. (2008). Found Technology: Players as Innovators in the Making of Machinima. (T. McPherson, Ed.), Digital Media, Digital Yo, 165-196. The MIT Press.
Lowood, H., (2005). High-performance play: The making of machinima. (Andy Clarke and Grethe Mitchell Eds.), Videogames and Art: Intersections and Interactions, 59-79. Intellect Books.
Hanson, M. (2004). The End of Celluloid, 2004, RotoVision.
Kirschner, F. (2004). Machinimag One. PDF http://machinimag.com/machinimagOneRe.pdf, retrieved 15.02.2012.
Kirschner, F. (2004). Machinimag Three. PDF http://machinimag.com/machinimagThree.pdf, retrieved 15.02.2012.
Kirschner, F. (2011). Machinima’s Promise. Journal of Visual Culture, 10(1), 19-24.
Kirschner, F. (2011). Towards a Machinima Studio. (M. Nitsche, H. Lowood, Eds.), The Machinima Reader, 54-71. The MIT Press.
Nitsche, M. (2011). Machinima as Media (M. Nitsche, H. Lowood, Eds.), The Machinima Reader, 113-125. The MIT Press.
Hancock, H., Ingram, J. (2007). Machinima for Dummies. For Dummies.
Marino, P. (2004). 3D Game-Based Filmmaking: The Art of Machinima. Paraglyph Press.
ILL Clan “Larry and Lenny Lumberjack (movie) (2000).
Kirschner, F. Neumann, K. Scholz, A. “The Bob Block Machinima Show” (live performance) (2004)

Web References
Marino, P. (2005). Thinking Machinima, http://www.machinima.org/paul_blog/2005/10/machinima-from-inside-out.html, retrieved 15.02.1012.
Saunders, C. (2006). Bit of an Update, http://www.boxel.co.uk/category/undergrad/4d/page/2/, retrieved 15.02.2012.
Bakutis, E. (1998). Devils Covenant (movie), http://www.tebakutis.com/index_independent_dc_release.html, retrieved15.02.2012.
Lin, T. (2003). Sample Matinee Tips, http://udn.epicgames.com/Two/SampleMatineeTips.html, retrieved 15.02.2012.
Bakutis, E. (2004). Forum post, http://ataricommunity.com/forums/showthread.php?p=5397676, retrieved 15.02.2012.
Bakutis, E. (2004). Forum post, http://ataricommunity.com/forums/showthread.php?t=388116, retrieved 15.02.2012
Press Release NVIDIA Corporation. (2003). Epic Games and NVIDIA Kick Off “$1,000,000 NVIDIA Make Something Unreal®” Contest. http://nvidia.com/object/IO_20030606_4169.html, retrieved 15.02.2012.
Kirschner, F. The Journey (movie/game modification). http://journey.machinimag.com, retrieved 15.02.2012 (2004).
Kirschner, F. Moviesandbox (software/game modification) (2006). http://moviesandbox.net, retrieved15.02.2012.
Nitsche, M. (2007). American Film Institute goes Machinima, http://freepixel.org/?p=58, retrieved 15.02.2012.
Matlack, C. (2005). France: Thousands of Young Spielbergs, http://www.businessweek.com/magazine/content/05_51/b3964049.htm, retrieved 15.02.2012.

Software References
Quake, ID software 1996
Milkshape 3D, Chumbalum Soft, 1996-
Quake 2, ID software 1997
Unreal Tournament 2003, Epic Games, 2002
Unreal Tournament 2004, Epic Games, 2004
Faceposer, Valve, 2004
3D Studio Max, Autodesk 1995-
the Sims 2, Maxis, 2004
Halo, Bungie, 2001
the Movies, Lionhead, 2005
World of Warcraft, Blizzard, 2004
iClone, Reallusion 2003-
Moviestorm, Moviestorm 2007-