Datamoshing: A Glitch in the Ritual, by Nick Felaris
Imagine turning an MP3 file into a video. Imagine the opposite: turning a video into an MP3 file.
What if you were to add audio effects (echo, pitch or reverb) to the MP3 file before you turned it into a video? What would echoes even look like in video form?
In an age where AI-generated art questions how our fleshy, human creativity might coexist with machinery, datamoshing adds a human element to making art with computers. With a few rudimentary tools (and a little technical knowledge), videos can meld pixels and grids together. If you’re brave enough, datamoshing can be a ridiculously simple exercise in toying with digital media.
So, how does it work, and how do you do it?
Datamoshing is an experimental “glitch art” practice in which digital files are manipulated, often creating messy or “moshed” results. According to New Atlas, datamoshing was discovered in the early 2000s by artists such as Owi Mahn and Laura Baginski. Mahn and Baginski’s 2003 video, “Pastell Kompressor,” glitched found footage to create surreal experiments out of corrupted media. Multimedia artist Takeshi Murata’s 2005 short film, Monster Movie, looped datamoshed GIFs and MP4s. Datamoshing, when edited together, often “blends” two videos together: a feature that is often utilized by high-budget music videos. In A$AP Mob’s “Yamborghini High,” for example, cars and figures blend seemingly out of nowhere.
However, datamoshing’s beauty stems in part from its independent community. Audacity, a free audio editing program created in 2000, featured a forum community for users to help others edit low-stakes MP3 projects such as podcasts or song downloads. Daring Audacity users began importing raw files, adding an echo or two, and then force-exporting the manipulated file as an MP4 video. As editor Jamie Boulton demonstrates on the questionsomething blog, putting an echo on an image file can “duplicate” the image on top of itself. Amplifying or turning up the image file’s volume can add extra contrast. Like an abstract expressionist painting or a William Burroughs cut-up poem, datamoshing is about the spontaneous joy of deconstructing media using accessible mediums.
What mediums? The fruit of all 21st century knowledge: YouTube tutorials.
The easiest way to get into datamoshing is to watch just about any YouTube tutorial you fancy. John Rollercoaster Jr. demonstrates how to use Adobe Premiere to corrupt “reference frames” in order to glitch-transition from one clip to the other. Ultimate-Grand-Tutorial, on the other hand, takes a far more complicated approach to datamoshing photography. Using a program called HxD, Ultimate-Grand-Tutorial copies and pastes strings of text from different resources, corrupting each photograph individually instead of blending them into one another. What results from this “Hex Editing” method is a sequence of photographs that flicker and glitch out of each other.
If this all sounds like computer nonsense, that’s the point.
If you want to datamosh music, FrankJavCee’s hilarious “HOW TO AMBIENT DRONE MUSIC IN AUDACITY” is an excellent guide for datamoshing computer files into audio noise. FrankJavCee uses Audacity to turn an app such as MS Paint into a “raw file,” which, surprise, sounds like a random glitch nightmare. However, FrankJavCee filters the noise through Audacity’s Paulstretch feature. The result? An unexpectedly interesting and spacious drone, completely different from before.
A more complicated tool for datamoshing audio, Max MSP, is a coding program where users can create computer programs to generate sounds. UK electronic duo Autechre, for example, designs their own Max MSP patches. On their 2001 album, Confield, Autechre doesn’t so much play instruments as mess with computers. Confield’s datamoshes create an experience that sounds entirely alien: a 62-minute masterpiece that shows just how far computer wizards go in order to achieve glitchy perfection.
Other complicated datamoshing software, such as ppooll, takes an audio sample, loops it and deconstructs it through a network of mind-boggling effect chains. Originally called “lloopp,” ppooll was designed by Max MSP programmer Klaus Filip. The program is an open-source tool that sound designers manipulate sound with stunning results. Ambient artist Tim Hecker, for example, has expressed his love of ppooll, which he has utilized ever since 2001’s stunning Haunt Me, Haunt Me Do It Again. There are plenty of online resources to create experimental datamosh masterpieces. It all depends on how complicated you’re willing to make it.
One track on Haunt Me…, “The Work of Art in the Age of Cultural Overproduction,” references Walter Benjamin’s clairvoyant 1935 essay, “The Work of Art in the Age of Mechanical Reproduction.” Decades before digital software and AI art, Benjamin believed that accessible, newly-emerging technology, such as the camera and the printing press, made art lose its “aura” through mass distribution. At the same time, Benjamin recognized that changes in art history result from shifts away from historical tradition, or “ritual:”
“An analysis of art in the age of mechanical reproduction might do justice to these relationships, for they lead us to an all-important insight: for the first time in history, mechanical reproduction emancipates the work of art from its parasitical dependence on ritual.”
Maybe our generation’s algorithmic, AI-generated content sewage is Hecker’s “Age of Cultural Overproduction.” Benjamin’s “Age of Mechanical Reproduction” might argue that editing software is our shift away from the ritual: a tool that drains videos and MP3s of their aura. After all, is datamoshing creating something new if it’s drawing on preexisting digital files?
Uh … yeah.
From iMovie to the 4k film camera, from MS Paint to Adobe Photoshop, from Audacity to Max MSP, anyone can use simple editing tools for videos, images or music. By experimenting with readily-available editing software, datamoshers deconstruct digital media for its unusual quirks and glitches. Software may be draining art of its “aura,” but datamoshers break away from traditional artmaking practices by using computer software to manipulate digital media. Beyond technology, datamosh communities represent the Digital Age’s lust for knowledge that, in an unexpectedly positive way, calms Benjamin’s fear of mechanically-reproduced art.
What does the future of art look like? It looks like this: