aimastering.dev / dev log
Post #018 · 2026-03-08

The Data Conclusions — Same Chord Progressions as Enka, 20 Years of Stagnation, and the Foundation Is Japanese

Roland TB-303, TR-808, TR-909. Japan already won. It just does not know it yet.

conclusionschord progressionenkaRoland303808909Japan

1.Methodology: No Cherry-Picking

One rule governed the selection of analysis targets: no exclusions based on personal taste.

Hard trance, epic trance, progressive trance, and euro trance — the prototype of EDM I dislike — were all included alongside psychedelic trance. The things on the opposite side of personal preference are most valuable as data. Bias enters through exclusion.

GenrePeriodTracksPersonal view
Psychedelic Trance2003–2024~3,200Like
Progressive Trance2000–2024~4,100Like
Hard Trance1999–2015~1,800Dislike
Epic Trance2005–2024~2,600Dislike
Euro Trance / EDM prototype1998–2010~1,400Dislike
Techno (Detroit / Berlin)1995–2024~2,900Like

Approximately 16,000 tracks total. All Spotify Audio Features fields retrieved: energy, valence, danceability, tempo, key, mode, acousticness, instrumentalness.

2.Chord Progressions Have Not Changed

This was the first conclusion out of the data.

Finding 01

From 2003 to 2024, the distribution of modal key and chord progressions showed no statistically significant change.

Minor key (Aeolian) concentration: 72–79% across all periods. Top chord progressions (Am–G–F–E, Am–F–C–G, etc.) share: approximately 60% throughout.

When I searched for this progression in a different context, I stopped cold.

Cross-reference result
Psy-trance most frequent
Am–G–F–E (Aeolian + Phrygian cadence)
Enka fundamental progression
Am–G–F–E (Yonanuki scale + same cadence)
Shared structure
Minor descent, ♭VII stability, semitone descent resolution

This is not coincidence. Both encode the emotional pathway of grief–ecstasy–release into the same harmonic structure. Enka arrived there over centuries. Trance arrived at the same place in 20 years — and has been stationary since.

The absence of change is not a criticism. The structure does not change because it works. The problem is that nobody is using this consciously as a design principle.

3.Top Artists — No Earthquakes, No Wind

I cross-referenced the top 20 artists by monthly listeners and play count at five-year intervals: 2005, 2010, 2015, 2020, 2024.

Finding 02

Of the top 20 artists, an average of 3.2 names changed between 2010 and 2024.

Infected Mushroom (debuted 2000), Astrix (early 2000s), Vini Vici (from 2014) — no new "genius" has appeared. Every artist who appears to be a new entrant is either a brand refresh by a veteran of 10+ years or a sub-project of an existing label.

The market expanded — Goa Trance Festival attendance is 3–4× the 2010 figure. But that expansion was driven by existing brands and festival economics, not new creative forces.

A genre with no young geniuses is enterable — if "the person to beat" is fixed and calcified, the adjacent empty coordinate is structurally available.

4.The Foundation Is Japanese

While organizing the data, a question kept nagging at me. What is the core machine behind the sound of this genre?

Roland TB-303
1981 · Bassline Synthesizer

Origin of acid sound. The resonance + cutoff "voice" is the foundation of trance bass

Roland TR-808
1980 · Rhythm Machine

Kick, snare, hi-hat harmonic structure carries analog warmth. Foundation from hip-hop to techno

Roland TR-909
1983 · Rhythm Machine

The reference standard for kick "weight" in techno, house, and trance. MIDI sync changed dance music production

Roland Corporation — Hamamatsu, Japan. A Japanese company.

The sound ringing in clubs worldwide has its foundation in machines built by Japanese engineers. The reference timbre of techno, house, and trance all originates in Japanese circuits. Western artists exported it, made it a global culture — but the source material is Japanese.

Japan does not recognize this victory. The cultural assumption that "electronic music belongs to the West" is blocking entry into a genre built on its own machines.

And in the NDL, the same sensibility that produced those machine sounds — ma, negative space, semitone descent — exists in songs from a thousand years ago. This is not coincidence. It comes from the same people's sensory apparatus.

5.Why Everything Is Public on GitHub

All code and data for this project are public on GitHub. One reason: transparency is the premise.

aimasteringPublic

The TRIVIUM mastering engine. GRAMMATICA / LOGICA / RHETORICA three-agent implementation and DSP pipeline

spotify-trance-analysisPublic

Spotify API analysis code from this post. Scripts for fetching, aggregating, and visualizing Audio Features for all 16,000 tracks

neuro-master-dspPublic

NEURO-MASTER DSP chain prototype. The Pure Python failure is left in the repo, code and all

The failed code stays in. The Pure Python loops of NEURO-MASTER, the half-finished refactoring — none of it is deleted. A failure record becomes the basis for design — the same premise as this blog.

6.What These Conclusions Mean

The data pointed to four things.

01

Chord progressions have not changed in 20 years — resistance to change is structurally low

02

Top artists are calcified — adjacent empty coordinates exist by structure

03

No young geniuses — no active innovator to be measured against

04

The foundation is Japanese — the legitimacy of entry exists in the origin of the materials

When these four aligned, the premise "I have no talent, so it is impossible" collapsed logically. The talent question is a comparison question. A coordinate with no one to compare against is a place where talent as a concept becomes invalid.

Japanese enka sensibility. Sound built from Japanese machines. A thousand-year NDL archive. And the empty coordinate of trance × Japanese.

When four coincidences align, the word for it is structure.

The next question is not "what to build" but "how to make it reproducible." That is what connects to the TRIVIUM mastering engine design — and why the technical posts from #001 onward were written after this conclusion.