A generator you can actually mix — while it plays
Most music generators hand you a finished file and a reroll button. This one gives you the board, and the board is live: press Play live and every control keeps working while the music runs. Pull the melody down under a voice, open up the room, brighten it, change key — the music never stops to re-render.
Energy sets how busy the melody is, Motion makes long tracks slowly breathe, Space is the room around the sound, and Brightness tilts it darker or shinier. Low cut and High cut set the frequency range of the whole loop. The channel strips give each layer its voice and level; the Equalizer adds five bands of ±12 dB with live per-band meters. Every control changes the actual synthesis — nothing here is a placebo knob.
The controls respond at different speeds because the sound works differently, and we'd rather tell you than pretend: levels, Space, Width and EQ are instant; Brightness and the band limits land within a fraction of a second; notes and voices — Energy, Motion, key, mode, tempo, instruments — arrive at the next bar with a crossfade, so a key change is a modulation, not a glitch.
Every render is a studio-quality stereo WAV (44.1 kHz, 24-bit) made on your device — nothing is uploaded. Free for personal projects; commercial use is a one-time $19 license, not a subscription.
Loops that actually loop
Background audio has to run for hours, so the file must loop without a click or a gap. The generator renders the track's natural continuation past the end and folds it back over the beginning — the reverb wash flows across the seam, so the loop point is inaudible. Preview seam plays across the exact wrap point before you ship it.
Downloads are WAV because MP3/AAC encoders pad every file with silence, which breaks gapless looping. A WAV loops sample-exactly in game engines, video editors, and apps.
Stems and theme packs
Stems exports the four layers as separate, sample-aligned WAVs — fade the melody under dialogue or rebalance in your editor. Deterministic synthesis means they always line up exactly. Theme pack renders your track's opening as a stand-alone intro and a short stinger — one sonic identity in three lengths.
Synthesized from a seed — and exactly reproducible
The audio is synthesized in your browser from your seed and board settings. The same seed and settings produce the bit-identical file, every time, on any machine.
FAQ
Is it really free?
Yes — every feature on this page, including stems and long renders, for personal use. No account, no watermarks, no ads. Commercial use needs the one-time $19 license.
Does the live preview sound like the file I download?
It is the file — same engine, same synthesis, computed in small pieces as it plays. Away from the loop point the two agree to within one 24-bit LSB. The single exception is the two-second loop crossfade; press Preview seam to hear the file's actual loop point.
What's a "seed"?
Any text you like — the fingerprint of your track. Same seed + same board = the identical audio, forever. Copy link encodes the whole board into a URL.
Can YouTube's Content ID flag this?
No system can prevent a platform or third party from making a mistaken claim. Because the audio is synthesized in your browser from a seed, the same seed and settings reproduce the identical file — a record you can use to explain how the audio was produced if you dispute a misidentification. Results remain controlled by the claimant, platform, and applicable law.
Do I need to credit anyone?
No attribution required, free or licensed.
Why does the same seed always make the same track?
The whole engine is deterministic — that's the feature. It's what makes links shareable, and a loop you shipped in a product reproducible years later.