Automatic Chord Detection with Chronotron

In a previous post, I showed how you can use Chronotron to extract vocals and instruments – stems, actually – from almost any clip. This enables many scenarios, like learning the song chords by listening to the accompaniment in isolation.

Starting from Release 167, Chronotron makes the job a bit easier by detecting chords automatically and displaying them along with the waveform.

Detecting chords isn’t a trivial task, though, and here’s how it works in a nutshell: an algorithm extracts the notes being played and match them against a number of chord patterns, figuring out the one that fits the best.

As you can imagine, many things can go wrong in this process. Hear the wrong note, or the right note at the wrong moment, and you’ll get a mismatch. This is why chord detection isn’t yet 100% accurate, but it does get the ball in the park; think of it as driver assistance rather than autonomous driving.

The feature can be enabled from the Instrument pane, as shown in the screenshot below.

There you can also restrict the chord families that you’re interested in, which not only improves detection accuracy but also helps you dealing with the ambiguity arising from missing notes – or from wrong assumptions about the root note.

Chord detection works best on instrumental material. Like the waveform itself, the detected chords react to the Audio Track Selection and Solo Channel, so you can choose what works best for your clip.

In addition to transcribing/learning, another scenario I wanted to target is improvising with an instrument along with a backing track, in which case knowing the chords as you play is a true gift.

The toggle switch Apply key changes to detected chords comes in handy when your instrument is tuned differently than the track being played. For example, some guitarists prefer to tune their guitar a half step down. In such cases, if you disable this switch, you can transpose the track in Chronotron to match your instrument key, while still showing the detected chords in their original key.

Future app releases will keep refining the chord detection algorithm and the features around it. If there’s anything you would like to see improved, let me know through the usual channels or in the comments section below.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s