About
I’m Stevza. I write, produce and mix music in the UK, and I have probably spent far too much of my life staring at a DAW instead of doing something sensible. I am still bitter about one particular egg and spoon race from about two decades ago, so clearly I know how to let things go.
The songs are mine, the singers are fictional. Each track is written for one of five characters, with their own personality, attitude and way of speaking. Their vocals are generated using AI, but the song itself still starts the old fashioned way, with me writing it, building the instrumental, arranging everything and deciding how the final track should sound.
I am open about using AI vocals because there is no point pretending otherwise. For me, the interesting part is using technology to give fictional characters their own voices. They are not meant to replace real singers, they are part of the world I am building around the music.
A lot of the songs are about normal life, money, work, boredom, politics, mistakes, wanting something better, or just being fed up. The characters let me write those ideas from different points of view without every track needing to sound like it came from the same person.
How a track gets made
- Pick the character. The song has to sound like something they would actually say, otherwise there is no point using them.
- Write the lyrics, normally reading them out loud several times until they stop sounding awkward.
- Build the instrumental around the track, rather than trying to force the vocal into something that does not suit it.
- Generate and direct the vocal parts, taking different versions where needed, then edit, tune, arrange and mix them into the track.
- Finish the mix, make the artwork or video, then put the whole thing out into the world and hope somebody actually listens to it.
Why fictional characters?
Because writing everything as myself would get boring pretty quickly. Maisie Belle can say things differently to Nina Noise, Aliska Star has her own sound, and the others each bring something different to a track.
They give me room to change genres, moods and writing styles without trying to make every release fit one single voice. Some songs are more serious, some are angry, some are stupid on purpose, and some are just there because I thought the idea would be funny.
At the end of it all, I make music because I enjoy making music. The characters, the visuals and the AI voices are just different parts of the same project.