Hannah Choi's Homework 1

Trulli
Listen to Kilauea!

For this assignment, I chose Hawaii’s Kilauea Volcano Eruption of 2018 to be the star of the show. My dataset is based on Sulfur Dioxide (SO2) emissions over the course of 4 months, which skyrocket in the explosive May 3rd eruption. Over a span of 125 days, we can hear unusual increases and decreases in volcanic activity, but nothing quite like what we hear on the last couple of days!

My piece involves a constant accompanying line containing G in the bass, and F# in the tenor voice. The dissonance created with the semitone difference, spaced wide apart, creates constant sizzles and bubbles, mimicking the lava’s activity. In this case, the soprano voice, which is the actual dataset, conveys the rising and falling intensities of volcanic activity. Finally, the alto voice screams an interrupting A# as Kilauea erupts, emulating Kilauea’s destructive explosion and a chilling Emergency Alarm.

(1) Why are the values for gain squared?

The values for gain are squared to better accommodate our perception of loudness, which is logarithmic. By squaring gain changes, we are able to hear smooth and less drastic changes when data values are low, but more prominent changes when data values are high.

(2) Why use MIDI keynums for the freq values? What happens if they're mapped linearly, instead?

Using MIDI keynums for the freq values allows for a greater variety of discernible pitch. Since we perceive pitch logarithmically, and not linearly, MIDI keynums does us the service of converting frequencies into pitches we are able to comfortably detect. If freq values were mapped linearly, it would be difficult to attend to changes in pitch.

(3) What happens if the 100 ms update rate is increased or decreased significantly?

If the 100 ms update rate is increased significantly, perhaps to 400 ms, the sound will be generated at a much slower rate, resulting in an increase in audio length. If the 100 ms update is decreased significantly, perhaps to 10 ms, the sound will be generated at a much faster rate, resulting in a decrease in audio length.

Here is my code and dataset:

Dataset - Kilauea 2018 SO2 Emissions
DataReader
My Code - Duo