I'm Done
One night, I really gave it up.
Nobody listens to my music anyway. It will never sell. Building everything from scratch — writing code, generating prompts, producing audio — I was too stupid for this. Fine. I will do whatever I want and end it here. Make it useful for work, run the experiments, let my dream close out there.
Everything I think of = wrong = does not sell = no talent = give up.
But I tell my children never to give up. The self-contradiction made me nauseous.
Japan has genuinely exceptional people. People worthy of deep respect. I am not one of them. That is a fixed fact. I placed it to the side.
The Job — Recreating Momotaro
Then a job came in.
A YouTube ad production. "A project to create the real Japanese folktale" — a faithful recreation of Momotaro from its original sources. Not commercial reinterpretation. Tracing back to the original text. Research began.
Where does the original Momotaro text live? Where can I find kagura scores? What did gagaku sound like in the Edo period? In what form does ancient Japanese performance survive?
Searching, I reached a place.
What I Found in the National Library
The National Diet Library Digital Collections.
An unbelievable volume of content. Free to use.
- —Momotaro original texts (multiple versions)
- —Konjaku Monogatari
- —Kokin Wakashū
- —Man'yōshū
- —Hyakunin Isshu
- —Gagaku scores & commentary
- —Kagura performance records
- —Folk songs & regional music
- —Shamisen & koto technique records
- —Ukiyo-e & woodblock prints (hi-res)
- —Edo period picture scrolls
- —Meiji period photography
- —Craft & dyeing pattern books
- —Folklore fieldwork records
- —Dialect & oral literature
- —Ritual & ceremony documentation
- —Local gazetteer / Fudoki
Public domain (copyright term expired) or CC BY (Creative Commons Attribution). Commercial use permitted.
Multiple versions of the original Momotaro. Kagura records. Kokin Wakashū. High-resolution ukiyo-e downloads. Folklore fieldwork archives.
All free. All commercially usable. Available right now.
What Is There — The List
I remembered the three waka concepts from #011 — the design principles of "my music": ma, yojō, honkadori.
| Design Concept | Matching NDL Material | Application |
|---|---|---|
Ma (間) Silence, negative space, time outside the beat | Gagaku beat records, kagura interval scores, noh breath notation | Reference source for rhythm design |
Yojō (余情) Aesthetics of incompletion, intentional resonance | Kokin Wakashū / Shin Kokin poetic theory, Motoori Norinaga's "mono no aware" essays | Mix harmonic design / reverb philosophy |
Honkadori (本歌取り) Quoting classical works to layer new meaning | Man'yōshū / Hyakunin Isshu original texts, folk song transcription records | Cultural foundation for sampling & citation |
Repetition & Transformation The core of psychedelic techno | Festival hayashi & kagura cyclic structure records, ritual music periodicity analysis | Prototype for trance design |
The design concepts and the source materials were a perfect match. Not coincidence. The mastering engine's design principles were already derived from Japanese acoustic aesthetics — so of course the source texts existed there. I was just late to notice.
Not Coincidence — Structural Inevitability
Breaking down what "This is it" actually meant: three things locked in at once.
What LANDR, eMastered, and iZotope do not have — "an engine whose design principles are grounded in a thousand-year Japanese acoustic aesthetic archive." This is an intersection no LLM engineer, musicologist, or DJ can reproduce alone.
When sampling via honkadori, the original texts are free. When reconstructing Momotaro, multiple versions of the source exist. The constraint of "building everything myself" dissolved here.
When Japan's content industry — anime, games, advertising — demands "authentic Japanese aesthetic," the person who can translate the source archive into music is me alone. This works as a business.
"No talent" may be true. But "standing somewhere no one else stands" is also true. Both can be true at once. This is that place.
Why I Am Here Now
The morning after the night I gave up, I was doing research. The answer came.
Two things started moving in parallel from that point. First: redesign the mastering engine — make Japanese acoustic aesthetics an explicit design principle. Second: fix the direction as a composer — not a DJ, but a maker who uses the NDL archive as source material.
This dev log is the record of that.
Posts #001 through #009 are the "engine records" — what came after "This is it." #010 through #013 document the path from despair to discovery. Everything after that is what the discovery produced.