2024
Conclusions + next steps
CONCLUSIONS
Ice Spice’s meteoric success in 2023 and 2024 came through hard work and the support of executives and marketing teams. This is not a criticism—most pop stars today rise to fame through a combination of talent and strategic backing.
An overwhelming majority of artists featured in this project hail from North America. Is this a result of accessibility, language or industry dominance? Consider this an open invitation to share music from artists outside of North America!
All five of my top albums this year fall under the experimental umbrella, reflecting my strong preference for genre-bending music. I’ve noticed that I tend to rate albums higher when they score well on the innovation scale.
Utilizing music information retrieval tools such as chromagrams and spectrograms offers analysts a unique perspective that complements traditional methods of music analysis. Ideally, both approaches—modern and traditional—should be used in tandem. Information such as key signatures, motivic movement, and accurate chord naming cannot yet be reliably detected by algorithms. These nuanced details, often evident to the ear and on paper, risk being overlooked by modern methods alone.
Boy’s a liar pt.2 is in Bb Lydian!
The noisy nature of unmerited suffering, with its extremely wide range of frequencies (as shown on the spectrogram), presents challenges for stem splitters and pitch-detection tools like BasicPitch. These programs often misinterpret the frequency data, detecting pitches where none exist. This issue is similarly observed with MIDI readers in ProTools and Ableton, highlighting the limitations of these tools when analyzing complex audio signals.
NEXT STEPS
Automation, Automation, Automation... I love this project, but I need to learn how to automate the process that currently takes me the better part of a year. The most time-consuming parts are the listening and analysis stages, and I know I can find a better use of my time. My goal is to create an algorithm that automatically inputs the data points I track in Excel using Spotify’s API. This would eliminate the need to open a database and manually track all the data points myself.
Additionally, it would be exciting to develop an algorithm that predicts my album ratings (with quality control, of course). I imagine this could function similarly to Spotify’s Discover Weekly, which recommends songs you’re likely to enjoy based on your listening habits.
I’m still uncertain about Tableau’s role in this project. While I appreciate its ability to create dynamic dashboards, Tableau has been a thorn in my side when it comes to embedding data into my website. This is especially frustrating when graphs must be uploaded as static PDFs instead of retaining the interactive features Tableau offers.
I think that with the addition of an automation algorithm, it would be interesting to continue to uncover applications for spectrogram and chromagrams outside of data visualizations. Are there any applications for these that can help the general population?