Questions, answered.

Play a song and it paints. Resonink listens to the music's mood, melody, harmony, and dynamics, and lets that feeling set pigment down on a canvas until the last chord: sprayed color, an ink line for the melody, a watercolor wash for the sustained passages. There are no prompts and no generated images: the music is the artist, and Resonink just brings the picture it already holds to the surface.

Yes. Painting is free, and it stays free. A paid Studio tier is planned for the things that cost real money to run: a gallery that follows you across devices, and prints of your paintings shipped to your door. The painting engine itself will never move behind the paywall.

Never. The whole engine (listening, analysis, painting) runs in your browser. Your audio never leaves your device, and your paintings save to a gallery stored locally in your browser. The only thing that leaves is anonymous usage counts (which buttons get used), and you can switch those off in the app's Monitor panel.

Three ways in: upload a track (MP3, WAV, M4A and friends), capture this device (the sound of a browser tab or your whole desktop, so anything you can play, Resonink can paint), or open the mic and paint a live room. Device capture works best on Chrome and Edge for desktop; uploads and the mic work broadly everywhere.

When you choose capture this device, your browser asks what to share. For the cleanest result, and so Resonink can read the track's name, pick the tab that's playing music, choose the Chrome Tab option (not a window or your whole screen), and turn on “Share tab audio” in the picker. Without that toggle you share the picture but no sound, and there's nothing to paint.

One thing to avoid: don't mute the shared tab. Muting it silences the audio Resonink hears, so the painting stops even while the song looks like it's still going. Turn your own speakers down instead. Sharing your whole screen works too (tick “Share system audio”), but a single tab reads the title best. Device capture is smoothest on Chrome and Edge for desktop.

Because that's the point. The painting is deterministic: same song, same settings, same picture, mark for mark. It's a fingerprint of the recording, not a slot machine. That's what makes a finished sheet worth keeping, and printing.

Color comes from the music's emotional signals (key, mode, tempo, loudness, consonance), mapped through decades of published research on how sound becomes feeling becomes color. In the default Mood mode, color is a direct readout of the emotion. The Blend and Notes modes let the notes themselves tint the wheel (more iridescent, less literal), and the app says so plainly while you're in them.

No image generator is involved. Resonink uses AI only to listen more closely (to pick the melody out of a mix, or tease the instruments apart), never to paint. Every mark on the canvas traces back to something that happened in the sound. That's why the same song always paints the same picture; generated art can't promise that.

It's optional stem separation: the song is split into drums, bass, vocals, and the rest (entirely on your machine), so each instrument can paint as its own voice. It's the heaviest thing the app does: it needs a desktop browser with WebGPU (recent Chrome or Edge) and a one-time model download of about 172 MB. Everything else in Resonink works without it.

In the gallery, stored locally in your browser: favorites, full-resolution sheets, all of it. Nothing is on our servers. Export any sheet whenever you like; if you clear your browser data, exported files are the copies that survive.

Print it, absolutely. That's half the point. Exports come print-ready: shape presets, 300 DPI, PNG or TIFF, with the resolution stamped so a print shop reads it right. For selling: a painting made from your own music is wholly yours. A painting made from someone else's recording is your artwork of their song; for commercial use, clear it with the rights holder first. We can't give legal advice, just an honest map.

Chrome and Edge on desktop give the full experience, including device capture and stem separation. Firefox and Safari handle uploads, the mic, painting, and exports well; their device-capture support is limited by the browsers themselves. On phones, upload and the mic are the honest inputs; capturing another app's audio isn't something phone operating systems allow.