Creative Cartography Final Project

First Published 2018 December 3

Draft 2

One of the classes I’m enrolled in while abroad is Creative Cartography.¹ The goal of the class is to explore the different ways that maps can express information and authorship. For a final project, we were assigned the task of creating our own unique map of London. So, being a musical child like I am, I decided to make my end product something auditory.

While discussing with my professor, we had many potential ideas. But, what ended up seeming most fruitful was a one hour walk, where I would follow a series of strangers, logging myself with a stopwatch. Every time that I made a turn, I would start a new lap on the watch. By the end of the hour, I’d accumulated 426 times, so 425 turns. Along with that, I had GPS data tracking my movements and a heart rate monitor that was linked to that GPS.

To create the music, I then needed to find a way to convert the raw data into something musical. I decided to create a melody that could be played solely on the white notes of the piano, rather than other scale options, for a variety of reasons. Had I chosen a pentatonic scale, the data would be harder to express. Had I chosen a chromatic scale, the final piece would have sounded incredibly atonal.

So, I assigned the time in seconds that I had lasted before turning as the basis for the melody. If I walked for less than a second, I would make the melody note an A, for 1-1.99, B, and so on through the scale. Of course, I knew that some of my turns would last seven or more seconds, so I would then repeat. That is, If it took 7-7.99 seconds, I would still have an A. But, in doing so, I would obscure a lot of data.

To combat this, I decided that I would use multiple voices in the final piece. I chose ninth chords, because they have an interesting lack of resolution to them that I thought might allow my music to feel more natural, especially as they tend to have forward movement that is needed when imagining a walk. If the turn would last for less than seven seconds, it would be the root of the chord, seven to 13.99, the third, and so on through the ninth. I assumed that I would then be able to uniquely identify each number of seconds. But, I ended up having ten turns that lasted more than 35 seconds. I decided that I would simply deal with that problem by ignoring it, looping at 35 seconds.

I then had to choose the way that each ninth chord would be constructed. I decided that I didn’t want diminished fifths from the bass, because I didn’t feel like it. My watch was accurate to the .01 seconds, so I kept that in mind.

To decide whether the third would be major or minor, I looked at the fractional time. If it was less than .5, I would make it minor. Otherwise, the third would be major.

The fifth was perfect.

The seventh was major if within .25 seconds 0. That is, from 0-.24 fractional seconds, it would be major. .25-.74 was minor, and .75 and up was again major.

The ninth was determined by the final digit. If the fractional second was even (.00,.02,.04...), it would be major. Otherwise, it would be minor.

By doing this, I constructed a lot of chords that don’t exist in the diatonic world. Of course, then I needed to find a way to automate the calculation of the data, because manually computing 426 data points sounded like a horrible idea, especially since I wouldn’t have the time. So, I automated the production of each note of the chord. I then plotted the melody and bass line, deciding that each chord would be root position. What was left was four voices. Using the principle of parsimonious voice leading,² I crafted each of the lines by hand. I then uploaded the midi data to a synthesizer, where I mixed the levels a bit until I was satisfied. I did nothing with dynamics or fading because that seemed to detract from the music itself. For tempo, I chose 120 as that is the typical person’s cadence. For voicing, I made sure that the bass was always the lowest and that the soprano was always the highest. The other voices, because of how I wrote them, progressed steadily higher throughout the piece. It creates an interesting effect of growing movement. Happily, it resolves on a CMaj7,9 chord, which fits the C diatonic scale.

If you want to listen, it’s currently hosted on Soundcloud.

Draft 1

One of the classes I’m enrolled in while abroad is Creative Cartography.³ The goal of the class is to explore the different ways that maps can express information and authorship. For a final project, we were assigned the task of creating our own unique map of London. So, being a musical child like I am, I decided to make my end product something auditory.

While discussing with my professor, we had many potential ideas. But, what ended up seeming most fruitful was a one hour walk, where I would follow a series of strangers, logging myself with a stopwatch. Every time that I made a turn, I would start a new lap on the watch. By the end of the hour, I’d accumulated 426 times, so 425 turns. Along with that, I had GPS data tracking my movements and a heart rate monitor that was linked to that GPS.

To create the music, I then needed to find a way to convert the raw data into something musical. Although I’d initially thought of using either a chromatic or pentatonic scale, as those each have their own benefits, I ended up deciding to use the C Ionian⁴ mode as my basis for melody. There has to be a better way to say that. Maybe: I decided to use the diatonic scale with no accidentals. I decided to create a melody that could be played solely on the white notes of the piano. That’s better.

So, I assigned the time in seconds that I had lasted before turning as the basis for the melody. If I walked for less than a second, I would make the melody note an A, for 1-1.99, B, and so on through the scale. Of course, I knew that some of my turns would last seven or more seconds, so I would then repeat. That is, If it took 7-7.99 seconds, I would still have an A. But, in doing so, I would have obscured a lot of the data.

So, I decided that I would have multiple voices. I chose ninth chords, because they have an interesting lack of resolution to them that I thought might allow my music to feel more natural. So, if the turn would last for less than seven seconds, it would be the root of the chord, seven to 13.99, the third, and so on through the ninth. I assumed that I would then be able to uniquely identify each number of seconds. But, I ended up having ten turns that lasted more than 35 seconds. I decided that I would simply deal with that by ignoring it, looping at 35.

Then, I had to choose the way that each ninth chord would be constructed. I decided that I didn’t want diminished fifths from the bass, because I didn’t feel like it. My watch was accurate to the .01 seconds, so I kept that in mind.

To decide whether the third would be major or minor, I looked at the fractional time. If it was less than .5, I would make it minor. Otherwise, the third would be major.

The fifth was perfect.

The seventh was major if within .25 seconds 0. That is, from 0-.24 fractional seconds, it would be major. .25-.74 was minor, and .75 and up was again major.

The ninth was determined by the final digit. If the fractional second was even (.00,.02,.04...), it would be major. Otherwise, it would be minor.

By doing this, I constructed a lot of chords that don’t exist in the diatonic world. Of course, then I needed to find a way to automate the calculation of the data, because manually computing 426 data points sounded like a horrible idea, especially since I wouldn’t have the time. So, I automated the production of each note of the chord. I then plotted the melody and bass line, deciding that each chord would be root position. What was left was four voices. Using the principle of parsimonious voice leading,⁵ I crafted each of the lines by hand. I then uploaded the midi data to a synthesizer, where I mixed the levels a bit until I was satisfied. I did nothing with dynamics or fading because that seemed to detract from the music itself.

If you want to listen, it’s currently hosted on Soundcloud.