Sonilo V1.1 Video to Video: The Essentials

Last updated: July 20, 2026

Sonilo V1.1 Video to Video generates an original, perfectly synced music soundtrack for any video. It reads the footage, its pacing, mood, cuts, and visual rhythm, then composes music that lands on the action and muxes it back into the clip, either replacing the original soundtrack entirely or blending under preserved speech.

What It Does

  • Scores to the picture. Beats, swells, and drops land on cuts and key moments, so the music follows the action instead of just playing over it.

  • Replaces the soundtrack with original music. By default, the whole original audio track is swapped for a fresh, commercially licensed score.

  • Keeps the voice, adds the music. With Keep Speech on, it isolates the original speech or vocals and blends the new music underneath with automatic ducking.

  • Preserves human voice, dialogue, singing, and crowds. Cheering, laughter, narration, and chanting can all stay in the mix.

  • Follows your direction or works on its own. Leave the prompt empty to let Sonilo read the video and pick a style, or steer it with genre, instruments, mood, and energy.

  • Handles full clips or segments. Score the whole video, or just a chosen segment using Start Offset and Duration.

Parameters

Sonilo V1.1 Video to Video

  • Video. The clip to score. Sonilo analyzes its pacing, mood, and visual rhythm, then muxes the new music back in. Cost scales with the scored duration, up to 360 seconds.

  • Prompt. Optional style guidance such as genre, instruments, mood, and energy. If left empty, Sonilo generates a fitting style from the video itself.

  • Keep Speech. Off by default. When on, the original speech and vocals are isolated and layered over the generated music with automatic ducking. Turn it on whenever the video has dialogue, narration, singing, or a crowd you want to keep hearing.

  • Samples. How many scored variations to generate from the same source, from 1 to 3. More samples give you options to pick from and add to the cost.

  • Start Offset. Optional. The point in the video, in seconds, where scoring begins.

  • Duration. Optional. The length of the segment to score, starting at Start Offset. Defaults to the rest of the video.

Examples

Every clip below is a real Sonilo generation. Turn the sound on to hear the score.

A neon-lit alleyway scored with a moody, pulsing synth track. Keep Speech off, so the new music fully replaces the original soundtrack. · Open on Scenario

A lone knight approaching a castle, scored with cinematic orchestral tension that builds with the walk. · Open on Scenario

A barreling wave at sunset, scored to rise and break with the swell. · Open on Scenario

A short action clip (a boxer in motion) scored end to end by Sonilo: the model reads the footage and composes a soundtrack that follows the movement, then muxes it back into the video. · Open on Scenario

Two friends laughing at a cafe with Keep Speech on: their voices and laughter stay intact over a warm backing track. · Open on Scenario

Tips for Better Results

  1. Feed it clear visual beats. Cuts, freezes, launches, and impacts give Sonilo something to hit, so the music syncs more tightly to the action.

  2. Leave the prompt empty when you want Sonilo to read the mood for you, or set a specific genre, instruments, and energy when you have a direction in mind.

  3. Turn Keep Speech on for any video with dialogue, narration, singing, or a crowd. It preserves the voice and ducks the music automatically.

  4. Keep Keep Speech off when you want a clean, fully scored result with no original audio, ideal for cinematics, trailers, and B-roll.

  5. Generate a few Samples when the fit matters, then pick the variation that lands best against your edit.

  6. Use Start Offset and Duration to score just the section you need instead of the whole clip, which also keeps the cost down.

  7. Mix any essential non-voice sound effects into your source outside Sonilo, since only human voice and vocals are preserved.

Known Limitations

  • Only human voice and vocals are preserved. Non-voice sound effects such as engine roar, footsteps, fireworks, hooves, and pure nature ambience are replaced by the music.

  • With Keep Speech off, the entire original soundtrack is replaced, including any sounds you may have wanted to keep.

  • Sonilo composes music only. If you need specific sound effects in the result, keep them in your source and mix them outside the model.

  • Cost scales with the scored duration, so long clips are more expensive to score in full.

Use Cases

  • Games. Score cinematics, trailers, and gameplay captures with music that tracks the on-screen action.

  • Film and animation. Lay down a temp or final score for shorts, sequences, and B-roll where beats need to land on cuts.

  • Marketing and social. Add original, licensable music to ads and social clips, keeping the voiceover or talking-head audio with Keep Speech.

  • Product and UGC. Turn raw vlogs, interviews, and demos into polished pieces with a soundtrack that sits under the speech.