What a daunting task Sagan and his team undertook when they set out to chose 90 minutes of music to “memorialize the human species” (Taylor, 2019) as part of the Voyager Golden Record. This assignment was the first time I had heard of this intergalactic record that is currently 23 trillion kilometers away from earth (Nasa, n.d.). Several of the songs, such as the Bach and Beethoven pieces, were selected due to their mathematical qualities (Taylor, 2019). Many were added because they were songs they felt were beautiful, which is a very subjective selection method. Dr. Smith Rumsey (Brown University, 2017) discussed subjectivity as related to the challenges of what materials are preserved digitally and which will be lost. Not all communities may be represented, just as not all languages and cultures were represented within the Golden Record
In deciding on how I would curate the original list of 27 songs to only 10, I ended up using a subjective method myself. I listen to a lot of music on Spotify, where algorithms are used to suggest playlists and artists that I may enjoy based on what I have listened to in the past. There are also playlists available that use mood as a filter. An example of this type of algorithm was created by Bhat, A.S. et al. (2014) who used measurements of intensity, timbre, pitch and rhythm to denote the mood of a musical piece.
“Moods classified in accordance to audio features” (Bhat, A.S. et al., 2014)
Being unable to use the complicated algorithms developed by Bhat, et al. (2014) or music streaming companies such as Pandora’s Music Genome Project, I decided to use the chart from Bhat as a guide to help me classify the songs by mood. I first categorized the songs by continent and whether it was instrumental or included vocals. I listened to the songs again and added a mood to each based on my limited music knowledge of pitch, rhythm, timbre and intensity. I picked one song for each mood and as I did, I checked to be sure that I had a variety of vocal and instrumental pieces as well as one from each continent.
The songs highlighted in blue are the ten that I curated. As you listen to the songs, do any of the classifications match with a mood you would selected?
References
Bhat, A.S., et al. (2014). An Efficient Classification Algorithm for Music Mood Detection in Western and Hindi Music Using Audio Feature Extraction. 2014 Fifth International Conference on Signal and Image Processing, 2014, pp. 359-364.
Brown University. (2017, July 11). Abby Smith Rumsey: “Digital Memory: What Can We Afford to Lose?” [Video]. YouTube. https://www.youtube.com/watch?v=FBrahqg9ZMc
NASA Jet Propulsion Laboratory. (n.d.) Voyager. What’s happening now? Retrieved October 29, 2021 from https://voyager.jpl.nasa.gov/
Taylor, D. (Host). (2019, April 22). Voyager Golden Record. [Audio podcast episode]. In Twenty Thousand Hertz. Defacto Sound. https://www.20k.org/episodes/voyagergoldenrecord
DEREKDOHERTY
November 7, 2021 — 11:46 am
Hi DeeDee, I like how you took quite a scientific approach to this task and tweaked the mood classification algorithm. When starting out on this project I also thought about sorting according to mood or emotions, but I hadn’t thought of using an algorithm like the one you used. When you analyzed the songs based on pitch, rhythm, timbre and intensity did it for the most part conform with your own feelings about the song or were there surprises?
Nick Hall
November 28, 2021 — 11:01 am
This was such an interesting approach to the task, and it’s really interesting too that even though we took very different routes we ended up with about 50% of the same choices. I love the idea of classifying by mood + instrumentation + continent. In a way it seems like a much less subjective approach than “beauty.”