Graffiti Analysis is an extensive ongoing study into the motion of graffiti. Custom software designed for graffiti writers creates visualizations of the often unseen motion involved in the creation of a tag. Motion data is recorded, analyzed and archived in a free and open database,, where writers can share analytical representations of their hand styles. Influential graffiti artist such as SEEN, TWIST, AMAZE, KETONE, JONONE, and KATSU have had their tags motion captured using the Graffiti Analysis software. All tags created in Graffiti Analysis are saved as Graffiti Markup Language (GML) files, a new digital standard used by other popular graffiti applications such as Laser Tag and EyeWriter. Graffiti Analysis 3.0 is an open source project that is available online for free in OSX, Windows and Linux. Graffiti writers are invited to capture and share their own tags, and computer programmers are invited to create new applications and visualizations of the resulting data. What Martha Cooper did for archiving graffiti on film, and Chalfant/Silver did for archiving graffiti in video, Graffiti Analysis intends to do for archiving graffiti in code. The project aims to build the worlds largest archive of graffiti motion, and bring together two seemingly disparate communities that share an interest hacking systems, whether found in code or in the city.

Graffiti Analysis is a project by Evan Roth. All of the software changes in GAv2.0 and GAv3.0 were written by Mzz Chris Sugrue, with support from the Fondation Cartier and Les Grandes Traversees. GA1.0 was created at Parsons Art Media & Technology with input and advising from Zach Lieberman. Graffiti Analysis was built in Open Framewerkz. Other code contributions included in GAv3.0 from: Theo Watson (laser input integration), Kyle McDonald (audio analysis) and ps / TPOLM (.gml RSS system).

The original version of Graffiti Analysis, developed in 2004 as part of my thesis research at Parsons in New York City, was created in collaboration with four NYC based graffiti writers: HELL, AVONE, JESUS SAVES and KATSU, and can be viewed here. A Daily News article from June 2005 about Graffiti Analysis 1.0 can be viewed here, and links to the early phases of its development can be seen below.

In 2005, based in part on my work with Graffiti Analysis, I was awarded a fellowship at Eyebeam, a non-proffit art and technology organization in Manhattan. It was there that I teamed up with James Powderly and started the Graffiti Research Lab, as well as collaborated with Theo Watson to create Laser Tag.

Graffiti Analysis 2.0, initially supported by the Fondation Cartier's Born In The Streets - Graffiti exhibition, marks the first update to the project since its initial public release in 2005. This release includes major updates to the tracking, playback, controls and graphics, as well as previously unreleased source code and downloads to Windows, Mac and Linux versions of the playback and capture applications. GA2.0 uses the new Graffiti Markup Language (GML) file format and is connected to the open database, which includes data from a growing number of influential graffiti writers including SEEN, 2ESAE, TWIST, AMAZE, KETONE and JON ONE.

Graffiti Analysis 3.0 was supported by Les Grandes Traversees and released publicly in September of 2010. This release includes additions such as audio input, architectural awareness and laser input.

- Instructional caligraphy research (2004)
- Initial project research (2004)
- Original thesis proposal (2004)
- Early print prototype (2004)
- First projection test (2005)
- Initial project page (2005)
- Graffiti Analysis thesis document (2005)
- Graffiti Analysis 1.0 project page (2005)
- Graffiti Analysis 1.0 video (2005)
- Projection Bombing tutorial (2006)
- Graffiti Analysis 2.0 project page (2009)
- Graffiti Analysis 3.0 project page (2010)

How do I capture my own tag to GML?
GML tags can currently be captured using Graffiti Analysis in two ways. The first is by using the Graffiti Analysis Recorder application, which can be downloaded from the DOWNLOADS page. Information on this process can be found on the HOW TO page. The GAv3.0 Capture application uses a modified marker as the interface and records all pen movement in a single line (including non-drawing movement). The second method for capturing GML tags using Graffiti Analysis is by using the GA:DustTag application for the iPhone. This method does not capture non-drawing movement and supports multiple lines. Other various methods for capturing GML data can be viewed at

What was used to create the GAv3.0 software?
Graffiti Analysis was written in C++ using Open Frameworks, a "C++ library designed to assist the creative process by providing a simple and intuitive framework for experimentation."

Can I export a video file from GA?
It is possible to export a sequence of frames from GA2.0, however, better results can be achieved by simply screen recording the app while it's running using 3rd party desktop recorder software (Snapz Pro in OSX, or RecordMyDesktop in Linux).

What are the particles showing?
The representation of the tag is a data visualization of the hand motion. The line thickness is based on speed; the faster the movement the thinner the line. The particles are thrown off based on changes in speed and direction, and are intended to visually emphasize the movement. Time acts as the third dimension and moves linearly away from the drawing point.

What is the relationship between Graffiti Analysis and Laser Tag?
Graffiti Analysis and Laser Tag are both graffiti software applications written in Open Frameworks, the main difference being that GA is primarily a tool for archiving and analysis, where as LT is a tool built for use in public space. Both applications use computer vision and motion tracking. GA is set up with the capture camera directly facing the light source (i.e., light pointed at the camera), and LT is set up to track a reflected light source (i.e., laser light reflected off of a surface). I wrote the original version of Graffiti Analysis in Open Frameworks in 2005, and in 2009 GA2.0 was updated heavily by Chris Sugrue. Laser Tag was produced collaboratively under the banner of the Graffiti Research Lab. I wrote the prototype version of Laser Tag in Processing in 2007 and Theo Watson wrote the full version of LT1.0 and LT2.0 in Open Frameworks (Zach Lieberman also contributed code and alternative drawing modes to GAv2.0). In the summer of 2010, Theo, Chris and myself worked together on implementing laser input into Graffiti Anlaysis 3.0

Can I use Graffiti Analysis in my advertising campaign?
No. All Graffiti Analysis media is released under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

What is the music you used in that video?
The music in the GAv3.0 video is from Kid Cudi's Demo Tape. The music in the GAv2.0. video on the about page is by Nosaj Thing. The music in the video on the iPhone page is by The Foreign Exchange. And the music in the video on the GML page is by Ratatat.

Do you have plans to release a version for Android?
No. However, a version is currently in deveopment by Turbo Connard.

Graffiti Analysis 1.0 was created while studying at the Parsons Art Media & Design department in New York City. Graffiti Analysis 2.0 was created with support from the Fondation Cartier in Paris. Graffiti Analysis 3.0 was supported by Les Grandes Traversees.

Copyright 2010 Evan Roth,
This software is licensed under GNU GPL

Creative Commons License
All Graffiti Analysis related media is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

Contact: info[at]graffitianalysis[dot]com

Follow: @graffanalysis

Web Ring: