How does the AI pick the best part of a song?
Before you choose anything, the page has already listened to the whole song, in your browser.
- It measures the energy. The song is scanned for loudness over time. A chorus is usually the loudest part that lasts, after the intro and the first verse.
- It skips the start. The scan ignores the first 20 % of the song, where intros and first verses sit.
- It places the window. A 15-second window (or 30 or 60) lands on the loudest part that lasts. In most songs that is the chorus.
- You decide. Drag the window anywhere on the waveform and press play. If your best line is in the second verse, use the second verse.
It works the same for a 2-minute single and a 10-minute track. The window is what becomes the video.
How do your lyrics shape the video?
When you tap Create, speech recognition listens to the part you picked and reads the sung words. An AI writer then plans the shots from them: one picture line per shot, in your style, following the story and the feeling of the lyrics.
So a line about driving at night can become a night road in Neon city, and a line about home can become a warm kitchen in Dreamy. The AI follows the meaning, not each word, and nobody in the video sings your lines.
An instrumental has no words. Then the shots follow the sound and the style you picked. If you add a photo, the same person is in every shot, whatever the song is about.
Pick the frame for the place where you will post. The length is a starting point.
| Where you post | Frame | A good length to start |
|---|
| TikTok | 9:16 | 15 seconds of the hook |
| Instagram Reels | 9:16 | 15 or 30 seconds |
| YouTube Shorts | 9:16 | 30 or 60 seconds |
| YouTube video | 16:9 Wide | 60 seconds or the full song |
| Your website or a press kit | 16:9 | 30 seconds |
Every video is a 720p MP4 with your song as the sound. Make the vertical and the wide one from the same part if you post in both places. Both cost the same.
The look is your choice too: Cinematic, Anime, Lo-fi, 3D cartoon, Neon city, Dreamy, Synthwave, Pixel art, Claymation, Storybook, Dark fantasy, Nature, Concert stage or Retro VHS. A lo-fi track fits the Lo-fi look: a cozy room, rain on the window and a slow camera. If none fits, describe your own style in up to 200 characters; it is saved for your next song.
Your own songs, covers and AI songs
Your own song. You hold the rights, so post it where you like. This is what song to video is for.
A cover. You recorded it, but somebody else wrote the song. The platform where you post may have its own rules for covers, so check them first.
A song you made with an AI music tool. Check the terms of that tool: some let you use the songs anywhere, and some do not. MakeViral does not give you rights to the music.
Someone else's record. Only with permission. The video keeps the song as its sound, so posting it is posting their music.
A simple plan to release a song with video
One song can give you a week of posts. Each one is a new video from the same file.
- Before release: the hook. 15 seconds of the chorus, 9:16, with your photo so people learn your face.
- Release day: the chorus in full. 30 seconds, in the look that fits the song best.
- The day after: a second look. The same part in a different style, to see which one people watch longer.
- Later that week: the verse. Drag the window to your best verse line, 30 seconds.
- For YouTube: the wide one. 60 seconds in 16:9, on the channel or under the song.
Each video is a new create, with its own price on the button.