The Sound of New York City Evictions
A data sonification and visualization of NYC evictions from 2017 to 2024
The Sound of New York City Evictions is a sonification project I created to make the trend of eviction data audible.
I had these goals in mind:
○ Use Python to map data to audio ✔︎
○ Craft an intentionally designed narrative audio story. ✔︎
○ Share the step-by-step method with readers of this article. ✔︎
I had these goals in mind:
○ Use Python to map data to audio ✔︎
○ Craft an intentionally designed narrative audio story. ✔︎
○ Share the step-by-step method with readers of this article. ✔︎
New York City eviction data shows a distinct trend: in 2017, the city reached a peak level of evictions (over 20 thousand), followed by a pause during the COVID-19 pandemic moratorium. Since the moratorium ended, evictions have returned to concerning levels. This is an ideal data set to sonify, because of the noticeable change in pattern over time.
Step 1: Find the Data
I found New York City eviction data (NYC OpenData) through Data Is Plural, a “weekly newsletter of useful/curious datasets” edited by Jeremy Singer-Vine.
I found New York City eviction data (NYC OpenData) through Data Is Plural, a “weekly newsletter of useful/curious datasets” edited by Jeremy Singer-Vine.
Step 2: Use Python to Convert Data to Audio
This approach is adapted from Matt Russo's sonification tutorial. Using it as a guide, I wrote a Python script to convert the NYC eviction data to a MIDI file (a format which is readable by audio interfaces).
The scripts and data are available in this GitHub repository. I also created a version for Jupyter Notebook for users who are less comfortable using a IDE.
This approach is adapted from Matt Russo's sonification tutorial. Using it as a guide, I wrote a Python script to convert the NYC eviction data to a MIDI file (a format which is readable by audio interfaces).
The scripts and data are available in this GitHub repository. I also created a version for Jupyter Notebook for users who are less comfortable using a IDE.
I chose to do a parameter mapping of eviction numbers to pitch velocity, where larger numbers of evictions correspond to powerful low notes, becoming higher pitched and lighter in feeling with fewer evictions.
Step 3: Customize the Audio with Logic Pro X
I imported the generated MIDI file from Python into Logic Pro X so that I could customize the sonification and add additional layers of sound.
I imported the generated MIDI file from Python into Logic Pro X so that I could customize the sonification and add additional layers of sound.
Customization choices I made:
○ Electric piano: powerful and responsive to velocity changes.
○ News segment audio: audio clips from relevant news updates.
○ Time marker: triangle tone marks the passing of each year.
○ Narration: I recorded my voice to give context and tell a story.
○ Electric piano: powerful and responsive to velocity changes.
○ News segment audio: audio clips from relevant news updates.
○ Time marker: triangle tone marks the passing of each year.
○ Narration: I recorded my voice to give context and tell a story.
Step 4: Add Visuals and Narration
I created a data visualization to accompany the sonification using Flourish. I then animated the visualization with Adobe software, making the visualization components appear in time with the sound.
I created a data visualization to accompany the sonification using Flourish. I then animated the visualization with Adobe software, making the visualization components appear in time with the sound.
I created a video montage, adding an introduction portion and recording my voice to share the context of evictions in New York City.
Published Project on YouTube
Watch the full project here! 👇
Watch the full project here! 👇