Drop in a song. Chordbrew separates the band, works out the chords bar by bar and writes the lyrics with their timing — on your own computer, ready to play.
Charts: the chords bar by bar, the bar that is playing highlighted, and the magnifier to fix a chord by dragging it.
Three steps, and the app does all three. You pick the song, then fix the little that needs fixing.
Vocals, bass, drums and the rest, each on its own. Chords come out much better when the bass and the accompaniment can be heard alone.
Finds the beat, the meter and the key, and writes each chord where it enters — with the bass as a second opinion to flag the ones worth checking.
Each line with its chords on top, over the syllable where they enter. And it proposes the parts: intro, verse, chorus, bridge.
Five spaces, each with its place. Whatever you fix in one shows in the others.
The chart follows the music: the bar being played stays in the middle of the screen, so you never let go of the instrument to find your place.
The same chords as Charts, over the syllable where they enter. Passages without singing are written bar by bar, so you can see how long each chord lasts.
Listen to a stretch slower, bass only, drums only. Mark where the 1 lands and the grid realigns. And download the separated tracks, aligned, for your DAW.
Build Sunday's list with each song in its key. On stage: full screen, dark, big lyrics with the chords on top, and the next song one arrow away.
The bar to find your place, the beat to fix it. No app gets everything right; what matters is how long it takes you to make it right.
Each bar shows its beats (1 2 3 4). Grab a chord from the middle to move it, from an edge to change how long it lasts; double-click an empty spot to add one. The magnet snaps it to the beat, the eighth or the triplet you choose.
A new song on Thursday, the chart ready the same day and in your singer's key. It was born in one of these teams.
Forty songs of repertoire, each with its structure, its separated tracks to rehearse with, and the show's set list.
The song the student wants to learn, slower, bar by bar, with the bass alone to hear the harmony.
It isn't magic, and we don't say it is. Against a chart a musician corrected by hand, the app gets the root of 91% of the chords and the exact symbol of 72%. What fails most is the quality of the chord — a Csus2 where a C belonged — and the odd chord that enters a beat late.
That's why the app tells you what to check: an orange underline where the bass plays a note the chord doesn't expect, and a verdict per song. Fixing a chord is a click; moving it, a drag; and everything undoes.
And what you fix stays fixed: it's saved with the song, survives a re-analysis and travels with the session to another machine.
One-time payment. With no servers there's no per-user cost, so there's no subscription: you buy it, it's yours.
Prices in USD, launch proposal. The app works offline; the license key is validated once.
Windows 10 or 11, 64-bit. About 2.5 GB: the models ship inside the installer, so it works offline from the first minute. With an NVIDIA card it uses it automatically.
Download Chordbrew for WindowsMac: later. Tell us if you need it: hola@chordbrew.com
Nothing. The app is a program on your computer; the analysis, the tracks and the charts stay in a folder of yours. No account, no server of ours in between. You can unplug the internet and it keeps working.
8 to 10 minutes per song on a normal PC; separating the tracks is most of it. With an NVIDIA card, about a minute. The app tells you before starting, based on your machine.
Measured against a chart a musician corrected by hand: 91% correct roots and 72% exact symbols, before touching anything. What fails is usually the chord quality (sus2, maj7) and the odd chord a beat late; the app flags what's worth checking.
Chords don't depend on the language. Lyrics are transcribed in Spanish, English and Portuguese, or auto-detected.
Yes: mp3, wav, m4a, flac, ogg. Any recording you have on your disk. It doesn't download from YouTube.
It works the same, it just takes longer. We're preparing per-song cloud analysis for those machines.
