What the app does with it
- The track is the video: a single part, the length of the track (10 seconds to 4 minutes). The soundtrack is your file, at full volume, from start to finish.
- Nobody speaks: no narrator, no lines. The characters, if there are any, live the song on screen.
- One shot every 2 to 3 seconds, synced to the lyrics. An abstract line becomes a concrete image (a gesture, a place, a metaphor), never a text.
- No text drawn in the images: no lyrics, no title, no readable sign.
- The lyrics are transcribed automatically, for free. Paste them for a more accurate sync.
- For an instrumental, your description guides the images.
What you provide
The form that works
For “Describe your clip”:Before / after
- Weak
- Improved
Common mistakes
- A file that is too short or too long: under 10 s there is no clip, beyond 4 minutes the breakdown no longer holds. Trim the track.
- Unreadable format: MP3, WAV or M4A only.
- Lyrics that differ from the singing: the sync is based on the transcription; paste the real lyrics or leave it empty, not an approximate version.
- Asking for text in the image: lyrics on screen or a drawn title don’t work. Use “Lyrics as subtitles”.
- Describing a dialogue: the characters don’t talk to camera, they act.
- Too many settings: a coherent world gives a clearer clip than ten places.
Settings that go with it
- Visualizer: the wave that follows your track, laid over the clip. Turned off, the clip comes out with only the images and your track.
- “Wave style” and “Color” (custom color by hex code).
- “Gradient” and gradient direction, “Glow”.
- “Position”, “Height” and “Opacity”.
- “Animate the wave with my track” to preview.
- Lyrics as subtitles: shows the lyrics on screen.
- Consistent characters: keeps the face of a singer or a hero the same from one shot to the next.
- Quality and style: the style applies to all the images. Refine it in “Refine the style (optional)”.
- Short test: if your track is long, try a 30 to 60 second excerpt first.