How it's made

How Takras is made

Takras is made with AI, and says so plainly. The tools are not the point. The feeling is. But the work is honest about how it gets there, so here is the whole of it.

The voice

Every Takras song is carried by a single voice: a careful, devoted, slightly weathered male register that stays the same from the first chapter to the last. It is a cloned voice model, kept consistent on purpose so the catalog feels like one person telling one long story rather than a collection of demos. The voice is the instrument the rest of the production is built around.

The songs

The writing and the audio are produced with Suno. Lyrics, melody, and arrangement are shaped and reshaped until a song says the thing it is reaching for and not a word more. The catalog is structured as a long-distance love story told in chapters, so each song has a place in a larger arc rather than standing alone.

The films

The music videos use synthetic characters generated with Higgsfield. The recurring face, Whispers Of London, is the visual identity of the project: a stand-in that lets the songs have a face without putting a real person on screen. The films are made for the songs, not the other way around.

Why say so

Plenty of music is made with these tools quietly. Takras would rather be transparent. Knowing a song was made with AI does not make the longing in it less real, and pretending otherwise would be its own kind of distance. The craft is in the choices: which voice, which words, which restraint. That part is human, and it stays that way.