Phonk Outlier Analysis: Brazil vs. Russia

About This Project

What is Phonk?

Phonk is a genre of hip-hop and electronic music that started in the early 2010s on SoundCloud. It was originally built on sampled Memphis rap vocals, lo-fi production, and heavy 808 bass. Over time, two very different regional versions of the genre emerged, each taking it in a completely different direction.

🇧🇷 Brazilian Phonk

Brazilian Phonk turned the genre into a high-energy club sound. Producers dropped the lo-fi aesthetic and replaced it with tight, aggressive loops, heavy compression, and tempos locked around 130 BPM. It now dominates Brazilian streaming charts and shows up in workout playlists and DJ sets. The style is very consistent by design.

Key characteristics:

Tempo: ~130 BPM (very consistent)
High energy and loud
Minimal melody
Made for clubs and gyms

🇷🇺 Russian Drift Phonk

Russian Drift Phonk went in a darker, more cinematic direction. It grew out of online car-culture communities and is known for haunting cowbell loops, minor-key melodies, and a much wider tempo range. The genre blew up on TikTok and YouTube through drift car videos.

Key characteristics:

Tempo: 80–180 BPM (very spread out)
Dark, minor-key sound
Strong cowbell and synth melodies
Linked to motorsport and action content

The Corpus

For this project, I submitted two Spotify playlists to the class corpus, each representing one of these subgenres. Together they contain 114 tracks, split 57/57 between Brazilian Phonk and Russian Drift Phonk.

The two playlists were chosen because they represent opposite ends of the Phonk spectrum. This contrast makes them a good subject for computational analysis: if the two styles are truly different, the data should reflect that.

The central question of this portfolio is:

Can computational musicology capture the difference between Brazilian and Russian Phonk, and if so, which audio features are most useful for telling them apart?

Research Overview

This portfolio covers the analysis in five sections:

Corpus — The dataset of tracks, including the Spotify playlists
Chroma Features — Pitch and harmony analysis of one track from each subgenre
Loudness & Rhythm — Tempo and loudness patterns across both subgenres
Classification — A machine learning model trained to distinguish the two styles
Conclusion — Key takeaways and what they mean

Use the navigation bar at the top to move between sections.

--- title: "Phonk Outlier Analysis: Brazil vs. Russia" --- ## About This Project ### What is Phonk? Phonk is a genre of hip-hop and electronic music that started in the early 2010s on SoundCloud. It was originally built on sampled Memphis rap vocals, lo-fi production, and heavy 808 bass. Over time, two very different regional versions of the genre emerged, each taking it in a completely different direction. :::: {.columns} ::: {.column width="50%"} #### 🇧🇷 Brazilian Phonk Brazilian Phonk turned the genre into a high-energy club sound. Producers dropped the lo-fi aesthetic and replaced it with tight, aggressive loops, heavy compression, and tempos locked around 130 BPM. It now dominates Brazilian streaming charts and shows up in workout playlists and DJ sets. The style is very consistent by design. **Key characteristics:** - Tempo: ~130 BPM (very consistent) - High energy and loud - Minimal melody - Made for clubs and gyms ::: ::: {.column width="50%"} #### 🇷🇺 Russian Drift Phonk Russian Drift Phonk went in a darker, more cinematic direction. It grew out of online car-culture communities and is known for haunting cowbell loops, minor-key melodies, and a much wider tempo range. The genre blew up on TikTok and YouTube through drift car videos. **Key characteristics:** - Tempo: 80–180 BPM (very spread out) - Dark, minor-key sound - Strong cowbell and synth melodies - Linked to motorsport and action content ::: :::: --- ## The Corpus For this project, I submitted two Spotify playlists to the class corpus, each representing one of these subgenres. Together they contain **114 tracks**, split 57/57 between Brazilian Phonk and Russian Drift Phonk. The two playlists were chosen because they represent opposite ends of the Phonk spectrum. This contrast makes them a good subject for computational analysis: if the two styles are truly different, the data should reflect that. The central question of this portfolio is: > *Can computational musicology capture the difference between Brazilian and Russian Phonk, and if so, which audio features are most useful for telling them apart?* --- ## Research Overview This portfolio covers the analysis in five sections: 1. **Corpus** — The dataset of tracks, including the Spotify playlists 2. **Chroma Features** — Pitch and harmony analysis of one track from each subgenre 3. **Loudness & Rhythm** — Tempo and loudness patterns across both subgenres 4. **Classification** — A machine learning model trained to distinguish the two styles 5. **Conclusion** — Key takeaways and what they mean Use the navigation bar at the top to move between sections.