I used three different datasets to complete this assignment:
1. Monthly U.S air passenger miles January 1960 through December 1977
2. Passenger miles flown domestic U.K. July 1962 through May 1972
3. Oil prices in constant 1997 dollars. 1870-1997
I wanted to contrast the passenger miles datasets with the price of oil while mimicking the arrangement of a typical piece of music, so I created two players, one to play the oil data as a lower, bass
part, and another to play the air miles data as higher, melody
parts. I named them BassPlayer.ck and LeadPlayer.ck, respectively. Both were developed from the data player m220a-Player.ck provided in the assignment.
The BassPlayer uses a sinusoidal oscillator and produces tones in the MIDI range of 40-60 (frequencies 82.4-261.6 Hz). It also utilizes a significant amount of reverb and chorus effects. This creates very mysterious, ominous sounds. In addition, the amplitude of the signal is linearly related to the scaled value of the data so that it is audible throughout.
In contrast, the LeadPlayer uses a triangle oscillator and produces MIDI values between 60 and 90 (frequencies 261.6 - 1480.0 Hz) all rounded to the nearest MIDI value, which produces only notes in the chromatic scale. The update time of the player is half that of the BassPlayer (250 vs 500 ms) and the ramp rate of the envelope is reduced to 25 ms, which has the effect of allowing individual tones to be sustained and heard distinctly. In addition, the player utilizes less reverb and no chorus effects, and the amplitude of the signal is determined by a squared function. This all combines to create the effect of producing a more conventional musical sound.
The results were recorded and mixed in audacity. It is important to note that the sonified data files do not match up chronologically, due in part to the differing spans and measurement intervals of the data, and in part to the desire to create a more musical piece. The two air miles datasets were panned hard left and right, and staggered such that their peaks (corresponding to peak flight seasons i.e. summer) alternate. It is interesting to note however, that when the two lead parts end, the corresponding data is in the same decade as the price of oil at the time, the 1970s.
Note: Due to the presence of low frequencies in this piece, I recommend listening to it through a decent pair of headphones or speakers. It wont sound right through typical built-in laptop speakers.
All data was taken from the Time Series Data Library