top of page

HIDDEN TRACK

Audio Feature Analytics of Super Junior Albums using Python and Javascript

Tools

Team Size:

Time Frame: 

Tools:    

My Role: 

1

12 weeks

Python, Tableau, Rawgraphs.io, JavaScript, Google Colab, Spotipy, Numpy, Pandas

Data Analytics, Data Visualization Engineer, Super Junior ever lasting friend (ELF)

Recording Equipment

Media from WIX

Why This Topic

Super Junior (SJ or SuJu) is an award winning South Korean boy band since 2005, but I did not become a big fan of them until recently. However, I became obsessively interested in SuJu and I was excited when I discovered I could know more under-covered through Spotify API. The artists I am comparing are actually both Super Junior, Super Junior D&E (Suju-dne) is a sub unit of Super Junior, with two dancing machines, sorry two dancing members Donghae & Hyukjae, from the father group of 13 members. Here are some of the questions I wanted to ask:

  1. Suju-dne is known for dancing, so does Suju, what is shown from the data?

  2. Is there some rule about what kind of songs would be popular?

  3. Are there typical keys or beats of songs that leads the way to success?

Data Analytics

Python Code to Request Data From Spotify API Steps 1-12

request spotify.gif
  1. Spotipy installed to save me from node.js.

  2. Used Spotify search to request data of specific keywords "Super Junior"

  3. Prepared artist_uri for further research

  4. Requested all albums data, with the artist_uri from above.

  5. Requested all songs from all albums and put them into each album dictionary.

  6. Set up a return information when they are finished one by one.

  7. Put audio_features also in the same dictionary as another key-value pairs.

  8. Set return information to show the process when every 5 albums are finished.

  9. Created a dictionary saved as a DataFrame of audio features of each song.

  10. Presented all data in a DataFrame, plus artist name as the last column.

  11. Refined all lines of data without duplicates.

  12. Wrote the DataFrame to a CSV file in my google drive directory.

CSV Dataset Overview with audio features from the albums of the artist

dataframe .gif

Correlation of Audio Features and Popularity of SJ D&E and SJ Songs

audio features.gif

Steps to Discover Correlation: Show Trend Lines --> Describe Trend Models --> Find the R squared --> Calculate the Square Root --> Get Correlation Coefficient

All steps the Pearson correlation measures the linear relationship between two variables. It turns out the Correlation Coefficient are very small which means there is no strong correlation between popularity and audio features.

Definition of Audio Features from Spotify

The full Audio Features , key and popularity explanations can also be found on Spotify documents. 

Energy - Energy is a measure from 0.0 to 1.0 and represents a perceptual measure of intensity and activity.

Danceability - Danceability describes how suitable a track is for dancing based on a combination of musical elements including tempo, rhythm stability, beat strength, and overall regularity.

Speechiness - Speechiness detects the presence of spoken words in a track.

Acousticness - A confidence measure from 0.0 to 1.0 of whether the track is acoustic.

Duration Ms - The duration of the track in milliseconds.

Valence - A measure from 0.0 to 1.0 describing the musical positiveness conveyed by a track.

Data Visualization

Distribution of Keys of Songs in All Albums of Super Junior and

How to make a Sankey Diagram from rawgraphs.io evolutions in steps

Step 1 Tableau pie chart shows the size of count 

Step 2 Tableau pie chart shows the size of count and which songs belongs to what key

Step 3 export the data for visualization in another tool

Step 4 Procedures in rawgraphs.io for a sankey chart 

Step 5 The top 1 key picked by Super Junior is C#.

Step 6 The top 1 key picked by Super Junior D&E is C#.

Music Industry is sales driven, in future there maybe more indicators 

Record Store

Media from WIX

Key Takeaways

The python library Spotipy installed saved me from node.js.

RawGraphs.io made it much easier to make a Sankey Chart in Tableau.

WIX is frontend friendly to have builtin Javascript interface to build a keyboard.

The top 1 key picked by Super Junior and Super Junior D&E are both C#.

Interactive Music Player  

Play the piano

PLAY| PAUSE the song

Key on piano keyboard

                 Key on staff lines 

Here We Are
Polygraph
SUNRISE
もっとぎゅっと
​ Breaking Up
Lights, Camera, Action!
Kiss Kiss Dynamite
촉이 와 (Can You Feel It?)
君が泣いたら
Hello
C
D
E
C#
D#
F
G
A
B
F#
G#
A#
Motorcycle
MOTORCYCLE
​ Teenage Queen
The Beat Goes On
Sweater & Jeans
Circus
Can I Stay...
Mother
Let It Go
너는 나만큼 (Growing Pains)
IF YOU
떴다 오빠 (Oppa, Oppa)
아직도 난 (Still You)
Oppa, Oppa
Ten years
LOSE IT
1+1=LOVE
Android Syndrome
Hot Babe
Love That I Need (feat. Henry)
Love That I Need
Champagne Girl
I Wanna Dance (사네)
I WANNA DANCE
​ You don't go
Take It Slow
Bari 5 !
C
C#
D
D#
E
F
F#
G
G#
A
A#
B
player

Created by De Han

  • LinkedIn
  • Instagram
  • Spotify
bottom of page