This is a concept for inspired by Spotify's "audio aura". It creates a color using Spotify Web API data. Similar to how Spotify might use the same data in their annual Spotify Wrapped.
This was originally course work from an Intro to Computer Science course. Uploaded here for future reference & archival.
python rgb.py
You can change the songs.db file to your own sqlite3 data inside rgb.py.
If songs.db contains the top 100 songs of one listener from 2018, how would you characterize their audio aura?
- I would take the average calues of danceability, energy & valence and associate them with the colors of red, green & blue.
- I would use the following query to get the values:
SELECT AVG(danceability) AS r, AVG(energy) AS g, AVG(valence) AS b FROM songs; - I would then convert each decimal value to an equivalent rgb value by multiplying it by 255 and rounding it to the nearest integer.
- I would then have a color I could associate with the audio aura. I could then save that output to a file or template and display it to the user.
- For example using the songs.db user data I would get back the following values:
- Danceability Avg: 0.71646
- Energy Avg: 0.6590600000000001
- Valence Avg: 0.4844429999999997 -Converting them as mentioned above and associating them with a color would give me the RGB values:
- Red: 183
- Green: 168
- Blue: 124
- This would give an RGB color for the Aura of
RGB(183, 168, 124).
Hypothesize about why the way you’ve calculated this aura might not be very representative of the listener.
- If a user only listened to 1 song that had the same values as the average of a user that listened to multiple songs; they would get the same aura color.
- This is also true if a user only listened to 3 songs.
- Let's take for example an unrealistic scenario where a user listened to a song where all the values were 0.75.
- Then another user listened to that same song but also 2 other songs with values that were +/- 0.25 respectively.
- So you'd have 1.00, 0.50 & 0.75 as your values. If you then averaged those out you'd still end up with 0.75.
- That would give you the same audio aura color as the first user or as if you had only just listened to the 0.75 value song.
- This isn't a very accurate representation of the differences in values.
- It could be more accurate to use a slightly modified approach. Such as summing the total from each column.
- Then rounding to the nearest whole integer and using that as the RGB values.
- Using the following query or similar instead of the above averages:
SELECT SUM(danceability) AS r, SUM(energy) AS g, SUM(valence) AS b FROM songs; - Should a value exceed 255. Then we could divide the whole rounded sum by 255 to get our new RGB values.
- This would allow a users auras to change based on how many songs they listened to & songs with higher values would add up quicker.
- If we kept it to only consider the top 255 songs; no user should go over the 255 mark.
- For example using the songs.db user data I would get back the following values:
- Danceability Avg: 0.71646
- Energy Avg: 0.6590600000000001
- Valence Avg: 0.4844429999999997 -Converting them as mentioned above and associating them with a color would give me the RGB values:
- Red: 72
- Green: 66
- Blue: 48
- This would give an RGB color for the Aura of
RGB(72, 66, 48).
- If we limited results from only the top 100. We'd likely just have to adjust the math.
- Maybe times it by 2 and some change to match the 0.00-100.00 values to equivalent 0-255 values.