References in Suno v6: turning an image or video into a track
One of v6’s headline features: a reference can be not only text but an image or a video. Here is what the model takes from a visual, how to pick a good reference, and why you should still pair it with a text prompt.
A reference visual conveys mood and atmosphere, not notes: colour temperature, the energy of the frame, era, “season”. A warm film cover → a soft vintage vibe; a neon night frame → synthwave energy. The specifics — genre, instruments, vocal, tempo — are still set by text.
How to use it — step by step
1. Pick a visual with a clear mood — a cover, a frame, a short clip.
2. Attach it as a reference in v6.
3. Add text: genre, vocal type, key instruments, production and BPM.
4. Generate and refine with words — e.g. “make the chorus bigger”.
Reference + text = control
A visual alone gives “vibe but off-genre”. The best recipe is image for mood + text for specifics. Copy the text-part skeleton and pair it with your visual:
It is a sample the model builds from. In v6 a reference can be not only text but also an image or a video — Suno “reads” the mood and style and turns them into music.
Does an image set the melody?
No. An image/video conveys mood, atmosphere and overall vibe (colour, energy, era), not specific notes. Genre, instruments and tempo are still driven by the text description.
What kind of visual works best?
One with a clear mood: an album cover, a frame with a distinct atmosphere, a short clip with pronounced energy. Blurry or “neutral” images give a vague result.
Do I still need a text prompt?
Yes — the best result is reference + text: the image sets the vibe, and words pin down genre, vocal, instruments and BPM. A ready text base is easy to grab from the catalog.