Generating Art from a Computer Game. An Interview with Alison Mealey

Posted by Thomas Petersen | Wed Dec 7th 2005 1:46 p.m.

Generating Art from a Computer Game. An Interview with Alison Mealey

Original article at Artificial:
http://www.artificial.dk/articles/alison.htm

Many artists use various types of processes, events, social patterns
etc. as controlling or contributing factors in the creation of artworks.
Alison Mealey has chosen to base her art on the computer game Unreal
Tournament. More precisely, she lets a number of virtual players play
the game for approximately 30 minutes at a time and uses the data from
the games to produce complex drawings. These drawings are also based on
photographic portraits. Thomas Petersen asked Alison some questions
about her art and the processes behind it. Check out Unrealart at:
http://www.unrealart.co.uk <http://www.unrealart.co.uk/> and Alison
Mealey's blog: http://alison.organised.info
<http://alison.organised.info/> .

Q: Tell me a bit about the basic ideas and the artistic motivation
behind the Unrealart project. What is it all about?
A: Well, it began as my final project for my MA in 3D Digital Design at
Huddersfield University. I had previously been exploring using games for
artistic means, looking at the interactions between players and
attempting to take the 'game' aspect away from the gameplay, creating
something that couldn't be considered a game but was fun to play with
nonetheless.
But I really wanted to use the game as a way of creating art, more
specifically my art, something that I could have great sway over and
control to a certain degree. I didn't want to be left out of the process
while the game produced randomly generated art. I wanted the game to
work for me.
I have always produced portraiture and have enjoyed some success in the
past as an illustrator. I wanted to combine all of my interests and have
the game create portraits using a very simple form of representation
akin to my illustrative style. My main aim was to create something
beautiful (from something far attached from beauty). Unreal Tournament
was used because I found it the most approachable game to modify, with a
well laid out editor, and plenty of information of how to use it on the
internet.

Q: It seems that you have a fair amount of control over the rendering of
the image. What is the artistic point of using autonomous/generative
strategies in your work?
A: I do have a certain amount of control, but I like to think of it more
as persuasion. I'm trying to persuade the bots that these paths are a
good direction in which to walk. It's up to them and the decisions they
make based on the game's stimulus whether they stick to my preplanned
routes or not. The images themselves would be very boring if the bots
didn't deviate at all from the paths, this is the reason I only use
large numbers of godlike bots to create the images. The godlike bots are
more likely to change their path in order to give themselves an
advantage during game play, novice bots barely deviate at all. At the
end of the day it's the game and the activity that's taken place within
it that draws the image, over this I have no real control, only the
power of suggestion.

Q: You use photos as a structuring element. What's the story behind the
photos, and how are they used in the process? A: Photographs of a
subject are an important part of the process. Just as any artist might
use a photograph to base a painting on, I use a photograph of a subject
as a starting point. The photographs are used in two important ways.
Firstly I base the AI pathing in UnrealEd around the features of the
face in the photo, creating an in essence a simple illustration of what
I feel are the most important areas of the face in the photograph that
will need to be defined.
The second important (though not always used) influence the photograph
has is on the colour of the final image. Though a fair amount of
randomness of colour is produced in the final artwork, the colour's base
value is taken from the photograph. Not all the artworks I create rely
on this stage, some of them are specifically tailored toward a certain
colour, some are random and some use a death as a marker for changing
colour.

Q: I'd like to learn a bit more about the relations between the data
from the game and the visual output. How are the characteristics of the
lines and circles determined from the game data?
A: Only two types of data are taken from the game. The position of every
player (taken every second), and the acknowledgement of a death. As the
data from the game is coming in 1 second chunks, Processing takes every
seconds chunk and produces a drawing from it, these drawings are built
up over time to produce the final images.
The circles represent the positions of the players. The X and Y
positions taken from the players are drawn as-is, producing a top down
2D view of 'the field of play'/canvas. The Y values from the players
alter the size of the circles, if a bot is mid jump they are therefore
closer to the camera and the circles will be bigger, if they are
crouching the circles will be smaller. The Y values have been greatly
exaggerated in some of the more recent works, to produce (in my opinion)
more beautiful images. The lines simply connect every second's points to
produce separate drawings from every second.
The death data is used in some of the images as a big black circle,
indicating where a death took place. In some of the other works the
death data forces a global colour change.

Q: You mention the visualization of deaths in the final images. In which
other ways are the thematic characteristics of the game are visible in
the final images?
A: In a way no other characteristics of the game are present other than
the death positions. No other 'statistics' of the game are logged.
However in another way the entire game itself is visualised in the final
images. The map played is shown in its entirety, as is the movements of
every player on that map over the total time span the map was played
for.
When watching the game and visuals take place live its possible to
follow a single bot's track in the drawing, and watch the drawing build
up over time and see the image slowly reveal itself.
I suppose it's the maps themselves that are the main game characteristic
present in the final images.

Q: Tell me about how you prefer to exhibit the piece, and if you have
any plans of exhibiting it in the near future.
A: I would prefer to exhibit the piece in two sections. One area where
the finished pictures can be viewed printed out quite large. And one
area where the live version would be running, I would prefer to have
both UT and the drawing as its being created to be projected side by
side. It's important that people understand the correlation between the
two. At the moment there are no definite plans to exhibit the work;
there are a few sketchy ideas, but nothing definite yet. Though I would
love to exhibit this work on a larger scale!

Q: Do you have any links to other game art works that you'd like to
share?
A: Tom Betts and his work (http://www.nullpointer.co.uk/-/home.htm) has
been an influence to my work, specifically QQQ. I saw QQQ quite a while
ago but still think it's a very nice work. Tom has messed with the Quake
code enough to force it to produce very beautiful and quite abstract
graphics. Tom was excellent at providing advice for me during the whole
process.

Spring_Alpha has also influenced me http://www.spring-alpha.org/. Here a
game is being created from a series of drawings produced by artist Chad
McCail. I have strong views about the game and its merits (that are best
unspoken), despite this however it's hard not to see basic similarities
between this work and my own. I have created a series of drawings from a
game, whereas they have created a game from a series of drawings.

The other work that isn't game related but has been important to me is
Surface Patterns: Walking Tours
(http://www.centrifugalforces.co.uk/surfacepatterns/pages/editable/tours
.html). I took part in a walking tour with Jen Southern in which I lead
her on a walk around Huddersfield and spoke about my memories and
feelings towards the places we were going (Audio Tour 7). This GPS
related artwork strongly influenced my work, not only because of its use
of positional data but also because of something I said which Jen kindly
documented during our tour, Audio Tour 7 and if you open the PDF you can
see something I said a over a year ago relating to paths, choosing paths
and creating paths. This is one of the most important elements of Unreal
Art, choosing a path then marking that path (along with forced pathing).

Original article at Artificial:
http://www.artificial.dk/articles/alison.htm

--------------------------
Thomas Petersen
+45 2048 2585
www.crossover.dk
www.artificial.dk
Your Reply