Free Pixel

discovering games as expressive media

Site search



Add to Google



[Other RSS Feeds]

  • CoLab
  • DWIG blog
  • [About]

    FreePixel looks at video games as part of the moving image culture. Games are not movies. But games use moving image tradition in their presentation. That is why FreePixel offers a critical look at games and their expressive qualities that grow from the use of the moving image.


    [December 31st, 2009]

    Machinima definition

    Posted by Michael

    This is one of those issues I that can cause only trouble but I am supposed to write a short piece on a related issue and the question I cannot avoid is: How to define machinima?

    A caveat upfront, nobody who is too busy doing it should be bothered. This might be one of those self-perpetuating problems academics like. It just so happens that I am such a creature and that I cannot avoid it.

    With all due respect, Marino’s definition of machinima as “animated filmmaking within a real-time virtual 3D environment” or Hancock/Ingram’s “Machinima is making films with computer games” and a “technique of taking a viewpoint on a virtual world, and recording that, editing it, and showing it to other people as a film” are both mainly utilitarian. The phenomenon needed some description and these statements provided the necessary descriptions by people who were part of this phenomenon. They offer an initial approach and remain immensely helpful as such. But once you shine any overly critical flashlight on them problems arise. So here are some musings about a work-in-progress machinima definition.

    It becomes exceedingly difficult to define machinima based on a technique, which is one reason the term “anymation” has been used as in parallel to machinima by artists such as Tom Jantol. Likewise, the connection to gaming is shrinking, the once dominating term “Quake movies” has become obsolete with increasing diversification of game engines used to create machinima (e.g. across different platforms and away from the dominating PC based systems). Tying it to a game in general has become equally problematic as special machinima creation packages like Moviestorm and iClone launched without any basis in gaming. However, replacing “game” with 3D world does not solve the problem as the definition of a 3D virtual world is equally unclear – is a Maya scene a 3D world? Why does it have to be 3D? And where is “film” in the equation?

    So here is a different take on the same question of defining machinima. I will define machinima based on the element of procedurality and will try to bring in the idea of performance. To get there you have to think back to the beginnings of machinima when “movies” were in fact demo files that were re-processed in real-time in the game engine. To a time before YouTube and rendered single versions of game movies. Sure, it runs against the grain, but there you go …

    I argue that maintaining a kind of performative control is key for the presevation of machinima’s identity. At its core, machinima is part of the same digital procedural media family as video games but it differs from games in the way that this control is weighted. In one word: Machinima is more flexible than games because it tries to do something different. Machinima can switch between modes of what is producing and controlling easier then games can.

    It has been another of those self-inflicted definition squabbles to define what a video game is but we can take it as generally accepted that a game depends on some form of player interaction (yes, once you enter the definition game even the most obvious statements can be dubious). A machinima piece might be game-based but it can completely switch to the machine as performing entity (see the demoscene example or Brody Condon‘s 3 modifications). It can also completely concentrate on the performance of the virtual puppeteer (see the original QuakeDoneQuick demos; Leeroy video; Dance Voldo Dance; or any bragging movie). Finally, a machinima can include customization and adjustment of the piece through the audience (see the biofeedback experiments of Champion; the old News at Seven project; or Kumagames Leaving the Game project – where I mentored during production). A machinima is a form of performance. Thankfully, this has been outlined in other places rather nicely – see Cameron and Carroll. Where I differ from their view is that I would like to extend the performance part beyond the capturing and generation and into the playback.

    The procedurality of a video game is dependent on the player doing stuff that affects the action. Machinima with its much stronger focus on the cinematic presentation, the “telling,” looks at a different use of procedurality. Where games play with the changes of the action, machinima plays with the changes in the cinematic narration.

    At the same time, the performance exceeds the control of mere rendering. Performance of the game, the player, and even the audience can include rules that go beyond the creation of the image itself and instead can affect the action that is displayed. These rules can be based in the engine and include elements such as collision control, physics simulation, and various AI behaviors. Or they can originate in a player who controls an in-game avatar, a light, a vehicle, or a camera in real-time of the final image rendering. Or they can apply audience elements into the event, like biofeedback, customization, and dependency on the local computer settings. All this is very different from the way 3D modeling packages are set up. Unless you modify a scene in Max or Maya to the extent that it becomes almost a game engine, these systems completely concentrate on the image render and not the inner workings of the rendered world. So here goes my first try to brew this down into a machinima definition as a work-in-progress:

    Machinima is digital performance that controls procedurally animated moving images in real time.

    With that in mind, its differences to traditional CG animation becomes clear, as does its difference to games and classic film. The downside: this merely touches on the current practices. Because the image should procedurally animated, the distribution of machinima as rendered movies e.g. on YouTube is seen as a step away from the core of what machinima is. On the other side, the advantages of this kind of machinima can be exceptional. Machinima, here, is not described as a technique but as a media format that should not strife to deliver the next Toy Story but might just have the necessary qualities to become a new media form like television or radio.

    However, this only works if the performance aspect and its flexibility remain supported. This kind of machinima definition depends on live play back, on engines running on the delivery end, on innovative control mechanisms, on new ideas of how we can manipulate moving images. Which is a lot of work for 2010.


    Comment from Phil Rice
    Time: December 31, 2009, 10:17 pm

    Very sound reasoning here, one of the most logical lines of demarcation I’ve seen proposed.

    Comment from Linn
    Time: January 1, 2010, 7:03 am


    I see a lot of similarities to Lowood’s puppeteering concept – although you’ve taken it a step forward into media adulthood.

    I’m very fond of your “controls procedurally animated moving images in real time.”

    I’ve always felt that the core aspect of machinima is mastering of an artificial intelligence that the artist had no role in creating – but rather learned a way of communicating with it in order to create a performance or story. Which, I suppose, has always cornered my own definition into the role of machinima as a media technique, without intentionally doing so.

    So I like the way that you’ve made it a media format – which it should be.

    Machinima is definitely a media format – not just a technique – but how to keep it that way definition wise, I’m not sure.

    Definitely intriguing analysis either way! I look forward to reading your progress!

    Comment from Michael
    Time: January 1, 2010, 11:52 am

    thanks for the remarks, this still very much feels like an ongoing process and the post is more an indication than a final conclusion (if there ever is one – which I highly doubt)
    Linn: yes, it has some parallels to Henry, although he comes from the high performance/ sport-like game players and my background is more a somewhat “traditional” digital puppetry approach.

    Comment from spyvspyaeon
    Time: January 1, 2010, 10:39 pm

    Excellent perspective, I have to agree fully

    Comment from James Spencer
    Time: January 3, 2010, 4:01 pm

    This was interesting to read Michael, thanks for posting.

    I have just recently found myself struggling even more with the machinima definition since we just started looking into iClone. Now when I go to explain machinima I have two vastly different tools for making it and I find myself getting lost in the explanation.

    As interesting as your post was I have to say that I still don’t feel confident in giving a definition to anybody. I imagine using your definition on somebody and the first thing I will get back is:

    “OK, so what does that mean?” or “So how is this different from what Pixar does?”

    Now I could go back and forth and scrape my way through trying to explain this but it really seems there should be a simple way to do this that makes sense to somebody not at all familiar with machinima (and to me). :)

    I plan to read it over again and think about it more and I realize you were not here to solve the problem or give the definitive answer.

    I think it is a very worthy topic to be worked on until there can be a clear line drawn and a clear definition accepted that is easy to understand and that can stand the test of time with evolving technology.

    Also people are more interested in what they can understand so sorting this out could very well boost machinima viewership, creation and general interest.

    Comment from Michael
    Time: January 4, 2010, 10:19 am

    James: very understandable difficulties.
    There is definitel more to machinima than the “media” aspect I put front and center here. For example, one could approach it as a community of practice and a form of emergent play. All true.
    However, I am trying to make a case that shifts machinima away from Pixar (unless they start to produce machinima, that is). The way Pixar, Disney, Dreamworks Animation … work is just not set up to deliver procedural CGI. So if one could say “machinima is when everybody truly gets their own version of an animated movie” that is not a definition really, but might explain what I am trying to say.
    Whether this really helps machinima as “an industry” I don’t know.

    Comment from Henry Lowood
    Time: January 4, 2010, 6:52 pm

    The approach you are taking requires a definition of “real time.” I hate to ask for another definition, but real time is a concept that is often used imprecisely. (I have written about this with respect to RTS games.) I am not sure if you mean the performance, the editing, or the viewing is occurring in real time, or maybe all three, in which case there is also a criterion of simultaneous performance and viewing — I doubt that you mean that, but there are such things (e.g., Ill Clan live performances).

    There is a subtle (or maybe not so subtle) distinction between real time as a performance studies person
    would use it (liveness) and real time as a game person would use it (machine time).

    This is an issue with demo, especially when sometimes people mix up the liveness and machine time in preferring this format. But it’s also an issue for
    capture, unless you just want to deny that some categories that are widely accepted as machinima are in fact not machinima (e.g., Baron Soosdon), because they are not created or performed in “real time.”

    Pingback from 2ndFantasy – SLMachinima issues
    Time: January 5, 2010, 8:01 am

    […] issues After reading and comment on the article Definition of Machinima Posted by Michael, the number of discussions and conversations within Second Life have raised the […]

    Comment from Michael
    Time: January 5, 2010, 9:20 pm

    Henry: I have to think more about it but I would define as real-time the processes between that which performs and the image. That implies, however, a live rendered image.
    So, yes, according to my outline the Machinima elements of Baron Soosdon’s movies are shrinking the more post-production there is because the processes that control the image I see are no longer performed in the same time it takes me to see that image.
    That does not mean that every machinima has to be live performed like the shows done by the ILL Clan, TSL, or Kirschner’s more recent work. But live performed machinima does indeed show most clearly the connection to the image control.

    Comment from Erik
    Time: January 6, 2010, 6:04 am

    “I’ve always felt that the core aspect of machinima is mastering of an artificial intelligence that the artist had no role in creating – but rather learned a way of communicating with it in order to create a performance or story.”

    ooh I like this way of looking at it-I really do-but all these notions of the uncanny can really wobble the grey matter…

    Real-time, yes indeed, it is one of those terms that sounds so definitive, but is it? Another slippery slippery concept.

    Comment from Kate
    Time: January 9, 2010, 5:11 am

    What kind of definition are you hoping to achieve Michael? I mean what function do you want the definition to have?
    Is the search for a definition part of the philosophy of the academic process, or is it a more of a practical need to categorise?
    I find the ‘how is this different from Pixar?’ question hard to answer, especially as I’m coming from an experiential approach. I can’t compare my experience of making machinima with my experience of working at Pixar because that second set of data is missing.
    I have come across creators who have worked at Pixar, and made machinima, but I’m guessing their comparisons would vary depending on their role within a large professional animation team.
    A comparison made by a Pixar film director would be very different from the experience of a polygon welder (made up name!)
    The director might say he deals with people rather than technology…the lowly keyboard tapper might say ‘with machinima you have more control of the whole project’ etc
    Individuals who make ‘big tool’, high quality rendered CGI tend to focus on the render quality, or their perceived lack of render quality in Machinima films..hence the slightly snitty definition of ‘low end rendered CGI’.
    I can’t help thinking that there is something in the Gestalt of making machinima..the whole being greater than the sum of the parts, which makes it so appealing for creators.
    Any creative group is hard to define using words alone, maybe impossible.
    I enjoy following your thoughts on this topic.
    How are we defining the word procedurally?

    Comment from fiezi
    Time: January 9, 2010, 11:58 pm

    Michael, I would love for machinima to take this route, but it seems you’re talking about something quite ancient in machinima terms that is almost lost nowadays. There are very few “real-time” production uses in the sense of your definition as I understand it that are being produced nowadays.
    Most game engines really only rely on canned animation that is triggered and played back, not modifiable while doing so at all. It seems like 10 years ago your definition would have made sense – with groups like the ill clan and all the recamming stuff in full effect, but I feel that the “making movies with games” definition is hard to wash off. Look at the thousands of movies produced in a very non-procedural way.
    Wouldn’t it make sense to go off the technical definition track and look at Machinima as a social phenomenon?
    I feel the social implications (democratizing animation, repurposing games) are much more prevalent today than the technical possibilities and properties you are describing.
    On the other hand, I see your definition in a lot of things happening to a large extent in productions like “Avatar”, where animated character feedback was being included in mo-cap/greenscreen shots on set (I just read that in a recent Digital Production magazine), or in Massive, where thousands of digital characters are simulated without any animator ever controlling a single bone.
    Wouldn’t this definition put machinima back into Pre-Viz territory and away from a new way of telling stories?

    Comment from Anthony Bailey
    Time: January 10, 2010, 3:46 am

    I read the original post with much interest, and bit my tongue lots, because the conversation it deserves to provoke will be a long one.

    To express my immediate thoughts succinctly seemed like very hard work. Now it’s easy: “+1 re pretty much everything that Kate and Friedrich said”

    I hope to read the rest of the conversation with equal interest and laziness! :-)

    (Disclaimer: I have a fuzzy and mostly historical relationship with this word., plus watching the definition form and shift in the earlier years.)

    Comment from michael
    Time: January 10, 2010, 2:49 pm

    Kate’s and Friedrich’s comments read a bit comparable. I really agree that machinima has all the points of a community of practice. It is rather complicated to define something that is so amorphous. And Friedrich is right: what I am suggesting is really pointing back into the stone age of machinima.
    What I am trying to accomplish might be two things. On the one hand us academics like to know what we are talking about and then engage in endless debates on our different readings of “know.” Like I mentioned in the original post: if you are busy doing machinima you really should not worry about such questions. Second, I became a bit worried about the future of machinima. One is used to the praises of the potential that it has but I also see some dangers. There are more and more advanced production pipelines (like LucasArt’s Zed/ Zeno) and more and more 3D packages allow real-time controls.
    It might just be that machinima’s only defining element is its cinematic reference to games; see Monty Oum or one of my favorites: Jun Falkenstein, who also has a traditional animation background.
    If that is the case and all other defining elements fade away, then machinima might become something like a genre.
    Without disregard to that, I am trying to identify something that allows machinima to provide another level. It still has the power to be a novel media format and this opportunity seems too important to miss.
    Anthony: great to hear the voice of the big bang! Yes, you surely had a front row seat in that discussion.

    Comment from Michael
    Time: January 24, 2010, 9:52 am

    Following the question here, some interesting discussions spread on other blogs: Kate and Michael take a stand mainly in the community practice (here and here) and Armanus looks at it a bit more from a aesthetic point of view. I think especially the idea of machinima as a community of practitioners is useful as another view, so give it a read. Overall, good to see that the discussion can still flare up.

    Comment from Gabe Salgado
    Time: May 29, 2010, 2:08 pm

    This is my favorite nutshell definition: “Machinima is digital performance that controls procedurally animated moving images in real time.”

    It suggests that a live digital performance can still be machinima, which means machinima is a format that transcends the limitations of other formats: film, TV, live TV, etc.

    I have some experience with highly choreographed live virtual world events/performances. While this can’t be considered cinema, it certainly is the essence of machinima.

    Write a comment: