How the dial works

Upload your own music and lyrics. In the manual editor, set the timing for each line. In the video editor, fill the scenes — one scene per line. The system stitches your images and videos together. It does not generate a note, a word, or a frame.

Choose the full report to see everything you did. That file is what you hand a producer, a director, or a copyright officer.

Use this setting to learn the map, to keep a story entirely in human hands, or to build the version you will later compare against a machine pass. Off is how you see what the other two settings actually add.

Servers trained on technical skills split your recording into stems. From the vocal stem, the system finds the timing of the lyrics. That timing locks each video scene to the song.

These are logical systems. They separate, time, and measure. They do not rewrite the lyric, replace the picture, or invent a melody.

You can also request feedback on your lyrics and metrics about your lyrics. Feedback is advice. Metrics are a mirror. Neither one alters the file.

Use this setting when the work is already yours and the labor is what is in the way.

From your lyrics it can draft music. From your image prompts it can draft stills and video scenes. This is the setting artists feel most threatened by, which is why it is not waiting for you when you open a project.

A clean way to use it: finish a piece by hand in No AI — or with Assistive timing only — then open a second, private collection and ask what the machine would do with the same material. Ideas. A look at what competition might try. Keep that collection private. Do not let it overwrite the piece you made.

Every generated pass should land in the report: what was human, what was drafted, which model, what date. Same handoff file as the other settings — producer, director, copyright officer.