Right now, I’m listening to an album I downloaded from a band called Black Tambourine. Their music is blissfully reverb- and echo-heavy, punctuated with hiss and feedback. There’s a weight to it, a warmth. But there’s a part of me that feels guilty for funneling it through laptop speakers and making it cold.
In the transition from analog to digital, mp3s have often stripped music of its essence, and people have spent time and money attempting to find ways to recover that somewhat intangible substance. But what about the sounds that escape when you compress an mp3?
Ryan Maguire has spent a lot of time with those ghosts.
Maguire is a doctoral student at the University of Virginia’s Center for Computer Music, and his project The Ghost in the mp3 explores those often unheard sounds. The soundscape “moDernisT” was pieced together “by salvaging the sounds lost to mp3 compression from the song ‘Tom’s Diner,’ famously used as one of the main controls in the listening tests to develop the mp3 encoding algorithm. Here we find the form of the song intact, but the details are just remnants of the original. Similarly, the video contains only material which was left behind during mp4 compression.”
Separated from the 1987 Suzanne Vega song, it’s actually a pretty cool piece of experimental music.
On the site, Maguire also deconstructs columns of white, pink, and brown noise, “when compressed to the lowest possible mp3 bit rate.” It’s an interesting look into what exactly is lost in compression, and translation.
We asked him a bit more about these ghosts.
How did you get involved in this project?
A couple of years ago, I was studying the history and development of electronic music with Tara Rodgers, and we were spending a lot of time looking at the different kinds of music that grew out of advances in audio technology. Music history can be traced really closely by following technological history. I noticed that vinyl records had inspired musique concrete and scratching, cassette tapes gave birth to tape music, then CDs inspired glitch music, and I was contemplating what a musical practice based on mp3s would look like. At the same time, we were reading Jonathan Sterne’s amazing book MP3: The Meaning of a Format, and things started to click.
The main difference between mp3s and every previous medium for recording music electronically is that mp3s are explicitly modeled on the limitations of our auditory perception. They are what is called a perceptual codec, which means that, in order to reduce file size, every mp3 utilizes a model of human auditory perception to erase information that, according to this model, most listeners in most situations will not hear. This idea was a really brilliant one by the folks who created the format—and also pretty audacious and, in a way, impossible. It works extremely well; I listen to mp3s nearly every day, but like most things, it’s not perfect. So I became very interested in the sounds that this codec just discards, and the result of that line of inquiry is what lead to the music I’m making now.
Why was “Tom’s Diner” the template?
“Tom’s Diner” in particular was chosen for the first track in this project because of its use as the primary test track during the development of the mp3 format. There were other tracks, too, which I am working on currently, but “Tom’s Diner” alone is considered the “mother of the mp3.” When the audio engineers developing the format had a new working prototype for the mp3 algorithm, they would test it with “Tom’s Diner,” and based on how well that song was encoded, go back to the drawing board. Eventually, they decided that it was encoded well enough by the format and went to market with it. Jonathan Sterne’s book has a really wonderful history of the development of the format and the role these tracks played.
What discoveries did you make that surprised you?
I think I was most surprised by how interesting the sounds were that were thrown out! I was very curious what it would sound like but didn’t really expect it to be anything that someone might want to listen to. So, when I finally finished my first test and put on my headphones to listen to the result, I had this really wonderful eureka moment. Like, “Hey, these sounds are really haunting and beautiful. I can work with this!”
Do you have a personal opinion on what mp3s do to music quality?
I think mp3s are great for listening through earbuds on the subway or in your car, and I do that nearly every day, but you really can’t replace the experience of sitting down and listening to something in a quiet place with good headphones or speakers and a high fidelity audio format. I don’t consider myself an audiophile, just a music lover, and when there is a recording I really love, I don’t hesitate to buy it in the best format I can, because that experience of just listening to it in all its detail can’t be replaced.
Can you elaborate a little on what “format music” is?
The term format music places this practice in the context of the development of electro-acoustic music through the last century or so. We first had people like Pierre Schaefer making very early recordings of trains and using vinyl records to create experimental music, while other people worked more strictly with electronics to develop synthesis techniques and electronic music. When magnetic tape technology came along, we had folks like John Cage experimenting with cutting up tapes and taping them back together to make audio collages and “tape music,” and then with the advent of compact discs, we have folks like Oval and Nicolas Collins using skipping and damaged CDs to create music, exploring the limitations of these ubiquitous technologies used to disseminate and store music. So, I think of format music as following in that tradition of recognizing what the current state of technology is and testing its limitations, turning it around and looking at it from different angles, and seeing what kind of art can come uniquely from our time.
H/T Death and Taxes | Photo via p_a_h/Flickr (CC BY 2.0) | Remix by Fernando Alfonso III
