Data Art: The Sound of Music Charts

Screen Shot 2017-04-30 at 4.28.07 PM

A couple weeks ago, I came across an infographic made by the music blog Pitchfork about the gender breakdown of performers at 2017 music festivals. Male groups comprised 74% of performances, while female groups represented just 14% of acts. The remaining 12% contained had both genders.

I play guitar and write for a music blog, and while I do feel that as a woman, I am in the gender minority in both of those activities, I was still surprised by the extremity of those numbers. I decided to investigate more. Does the 2017 lineup represent a more extreme gender disparity than other musical venues or do men dominate popular music that starkly?

To begin to answer that question, I turned to two different charts: the Billboard Top 100 Artists of the Year, and Pitchfork’s Top 50 Albums of the Year. I chose these two different sources because they are each considered, to a certain degree at least, the authority in their respective musical realms—pop and indie. To start, I collected data from the last five years of charts on each site, making a CSV that included the year, artist name, ranking, and gender.

Screen Shot 2017-04-30 at 8.46.52 PM

I suspected before I started visualizing the data that the Billboard charts would have a greater discrepancy between genders than the Pitchfork charts. Pop music, I thought, would reflect less progressive values than alternative music, plus Billboard uses a more scientific method  in their rankings (streams, album purchases, etc.) than Pitchfork, which, as far as I can tell, is selected by editors who can subjectively consider factors such as diversity.


When I visualized each data set, though, I discovered that the ratio of male to female performers is similar. In the Billboard Charts, the breakdown is:

349 Male; 132 Female; 18 Both

In the Pitchfork chart, it is:

167 Male; 61 Female; 22 Both

In both of them, all-female acts make up about a quarter of performers, and acts with either all females or a mix make up about a third of acts.

I decided to represent each chart year on a separate line, showing each of the three categories with rectangles of different colors (red for female, purple for both, and dark grey—a kind of default negative space—for male). Here are the two chart sources shown together:

Screen Shot 2017-04-28 at 8.32.01 PM


In class last week, I loved the projects that we saw that represented data with sound, and I thought that it might suit this project well. It is about music, after all. Also, seeing the stretches of dark space is one thing, but I thought that hearing a series of notes that always seems to return to a low drone would make the point more effectively.

Using a timer with millis() and the Processing sound library to create a sine oscillator, I made a very boring electronic song with three distinct pitches. Here’s that sequence choreographed (without sound) on the Billboard chart:

Different Sound

Finally, because the sine oscillator doesn’t make the prettiest of sounds and because this project is about female musicians, I decided to play the chart on guitar. I like the idea that, as Jer has talked about throughout the course of the semester, labor—time or physical exertion—can be another expression of the data. In the time that it took to get the notes right, would the finger responsible for strumming the string that represents male groups grow tired? (Yes.) It also brings the semester full circle for me, as it returns to the abstracted, aestheticized data representations that we explored in the first unit.

Here’s the song—not much more exciting than the one with the sine oscillator but with more reverb—that ensued:



Data & Publics: Tree Mode

Screen Shot 2017-04-09 at 10.15.37 PM
For the Data & Publics unit, Rushali and I wanted to create a conceptual project that would take place outside of a screen, on the streets of New York. We started by digging around the NYC Open Data site, looking for interesting data sets, and we found the 2015 Tree Census. It’s an incredible collection of data that contains information about 682,515 trees across the five boroughs, including species names, unique ids, even the status of each tree’s health.

As a jumping-off point, we imagined the city as a giant forest. What if, like in a real forest, the trees were the most pertinent landmarks? We started conceiving of an alternative city map and way-finding system that uses trees instead of street names and house numbers. In this alternative map, trees would become central to our experience of the city.

A Sampling of Inspirations:

    • Rushali sited her grandparents’ relationship to trees, talking about how in their villages in India, they knew the nearby trees intimately—what fruit they gave, when they bloomed, and so on. We wanted to regain some of this knowledge in our city, where people are detached from nature and don’t rely on it in the same way.
    • Trail markers: On many hiking routes, trails are marked by symbols on trees. We looked at these markers—sometimes a particular color of paint or a reflective medallion—as a model for using trees as guides.trail_marker
  • We also thought about accessibility in devising the form of the tree markers. Rushali talked about her friend’s braille magazine and using symbols that could be touched, to use a more reliable classification system then color.

Tree Map: 

For our alternative city map, we thought of creating a digital map such as Google or Apple maps (or a toggle option within) that the user could view in “tree mode.”

In our “tree mode” example map, we imported the tree census data set into a custom Mapbox Studio basemap, stripping the map of the usual markers—streets and street names, buildings, etc.—to highlight the paths created by tree dots. The numbers on our example show the IDs for different maples in the city and begin to show how people could adapt their notions of place in the city, from neighborhoods and streets to clusters of tree IDs. In the process of creating our map, we also found a couple of projects that created similar one, including NYC Parks and Jill Hubley.

Screen Shot 2017-04-09 at 6.24.11 PM


Tree Markers:

For the tree markers, we wanted to create labels that would serve multiple functions: showing the tree’s ID and “address,” educating passersby about the kind of tree, and presenting a fact 1 the tree that would emphasize the role of trees in nature and society. Each tree would have its own marker, which would have a color and identifying cut-out shape for the visually impaired corresponding with the tree type.

data art_tree markers-01

Screen Shot 2017-04-09 at 10.24.53 PM Screen Shot 2017-04-09 at 10.25.08 PM