Who Am I?
In mythology, the Minotaur is misunderstood, solitary, trapped inside a labyrinth.
In many ways, that is me.
Schizophrenia makes communication difficult for me.
So I learned to speak differently: through words, sounds, and images.
Self-taught. Guitarist, composer, poet, graphic designer — never in order, always with determination.
Behind every song, there is an intention, an emotion, a true story.
Trèfle is my wife and my muse.
For 25 years, she has transformed my labyrinth into something worth crossing.
25 years together, homeless three times, once self-employed, countless invisible wounds…
and yet, still here, still creating, still loving each other.
Blues · Country · Reggaeton · Dancefloor · French Variety · Socially Engaged Songs
Every song is an intention.
Every song is a choice.
🩸
About AI Music — Let’s Be Honest
No, I do not give four words to an AI and create a song in two minutes.
Here is how it really works:
✍️ Step 1 — Writing & Musical Analysis
Everything begins with the text.
The lyrics, the story, the emotions I want to transmit.
But writing does not stop at words.
From the text itself, I analyze what it inspires musically:
- The style — does this text call for blues, rock, French variety, reggaeton?
- The chords — which harmonic progression serves the emotion?
- The scales — pentatonic for raw rock or bleeding blues, arpeggios for melodies that breathe
- The riffs — where to place them, how to build them to serve the lyrics
- The techniques — bass pull-offs, hammer-ons, slides, the playing style that gives the song its character
All of this is defined before AI touches anything.
It is my inner score.
AI does not invent it — it performs it.
🎛️ Step 2 — The Mureka Prompt: Directing the Performer
I work with two separate prompts: one for the musical style, one for the lyrics.
I refine them, modify them, and rebuild them until the result gets as close as possible to what I hear in my head.
I say “gets close” because I am honest: I do not always achieve exactly what I imagined.
And that is where something interesting happens.
Did anyone ever tell Johnny Hallyday exactly how to interpret his songs? No.
The composer writes, the singer interprets — and in that space between the two, something unique is born.
Mureka works the same way.
I direct, it interprets.
With its own limitations.
With its own colors.
It is teamwork.
Me for the vision, the emotion, the artistic direction.
AI for the musical execution.
The final result belongs to both.
But the creation, the intention, the soul of the song: that is mine.
🎨 Step 3 — The Visual: Creating the Image That Tells the Song
A song without imagery is an emotion only half expressed.
Every title needs visuals that reflect its atmosphere, style, and lyrics.
First, I define the visual concept: the colors, the atmosphere, the character, the scene that belongs to the song.
Then I build a precise generation prompt to achieve exactly the result I have in mind.
But generating the image is only the beginning — not the end.
The real work starts afterward:
cropping, retouching, compositing, effects, visual consistency between every element.
That is where Photopea and Adobe Illustrator come into play.
This is not accidental.
I seriously studied graphic design — Photoshop, Illustrator, and even professional retraining in digital graphics.
And as a side note:
I learned web development with Notepad, coding HTML, CSS, and JavaScript manually, line by line.
In other words:
when I create an image for a song, I do not simply click “generate” and accept whatever appears.
I direct.
I rework.
I build.
🎬 Step 4 — CapCut: Video Editing
Beautiful music and beautiful images are not enough — they must be merged together.
Editing is the visual rhythm of the song:
when an image appears,
how it moves,
what changes during the chorus,
which effects reinforce the emotion.
Inside CapCut, I build the timeline frame by frame:
music synchronization, transitions, motion effects, subtitles, animated text.
Every clip is staged to serve the song — not just random images scrolling over music.
It follows the same logic as a music video director:
the sound already exists, now the visuals must tell the same story.
🎼 Step 5 — Scores & Tabs: Studio Tabs
This is the most technical step — and the most honest one to explain.
I separate audio tracks, create MIDI files, then encode them into Tux Guitar to generate sheet music and tablatures.
It is long and meticulous work.
But let’s be transparent:
I am self-taught.
I am not a formally trained musician.
My true strength lies elsewhere — in writing, emotion, and building images and sounds.
The scores are a service I offer to guitarists and pianists who want to learn my songs.
And Tux Guitar, like every tool, is not 100% reliable.
Some passages may be approximate.
I do my best to correct them, but I prefer honesty over pretending perfection.
What you will find in every FULL PACK:
piano, guitar, and bass scores + tablatures — meant as serious working material, not conservatory-level notation.
📐 Step 6 — YouTube Layout & Publishing
Thumbnail, clickable title, optimized description, chapters, links, hashtags — every video uploaded to YouTube requires careful formatting to survive within the algorithm.
It is a job of its own, just as important as the creation itself.
📝 Step 7 — SRT Subtitles
Honestly, this is probably the most tedious part.
Creating an SRT file means synchronizing every line to the millisecond.
And when lyrics mix multiple languages, as some of my songs do, it becomes a real puzzle.
Perfect synchronization is difficult to achieve.
I do my best, and I continue improving.
🔍 Step 8 — SEO & Sharing
Correctly referencing your work on YouTube, on this website, then sharing it on TikTok, Instagram, and X — that alone is a full-time profession.
For now, I focus mainly on creation.
The rest follows as best it can.
What All of This Means
From writing the lyrics to the final sharing, a single song easily represents between 12 and 48 hours of work minimum — sometimes far more.
Some songs remain blocked for months because I still cannot find the right solution.
So no —
it is not “four words and done.”