https://blogs.ubc.ca/ddperrott/2021/10/30/task-8/
I have selected DeeDee’s Golden Record Curation assignment to review.
Like me, DeeDee had not heard about NASA’s Golden Record project until this task. We both understood that the selection of songs for the record was subjective since some songs were chosen for their melodies and some of their mathematical qualities, as well as that not all countries and cultures could be represented. As such, we employed subjective methods for curating a smaller playlist of 10 songs out of the original 27.
My method was entirely too subjective – I simply listened to the original playlist several times to narrow down songs I liked and re-listened a couple of times to reach a final selection. On the other hand, DeeDee had a method of classification based on whether the songs were instrumental or vocal, the continent they represented and her mood in accordance with the song’s audio features.
I particularly found it interesting and insightful that she connected this curation task with Spotify’s algorithm where you can use mood as filter. She details that one type of such 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” (D. Perrott, 2021).
Mood was possibly a factor in my curation, but DeeDee’s take was empirically grounded and certainly sounds more appealing as a mood classification algorithm. In fact, her classification made me rethink how my playlist could potentially change if I were to see mood in terms of audio characteristics.
