# How to Automatically Generate Musical Harmonies from Audio

> This 2012 patent describes a system that listens to music and automatically generates harmony notes to accompany a melody, even detecting and ignoring accidental strums on stringed instruments.

- **Patent:** US 8168877
- **Original title:** Musical harmony generation from polyphonic audio signals
- **Owner:** Harman International Industries Canada Ltd
- **Granted:** 2012
- **Status:** Active
- **Times cited:** 24
- **Field:** consumer_electronics, software, telecommunications

## What it does

This patent details an apparatus that can create musical harmonies. It takes a melody signal and an accompaniment signal, analyzes their sound content, and then figures out what harmony notes would sound good. Specifically, it identifies the current melody note and looks at the 'spectral content' (basically, the mix of frequencies) of the accompaniment music. Using this information, it generates at least one harmony note. A clever part is its ability to detect and ignore 'unintentional strums' from stringed instruments, like when a guitarist accidentally brushes strings during a chord change, so it doesn't mess up the generated harmony. It can even output these harmonies as MIDI data or mix them with the original sounds for real-time performance.

## What it does NOT cover

- Generating harmonies without analyzing the spectral content of an accompaniment signal
- Generating harmonies without identifying a current melody note
- Harmony generation that does not suppress determination based on an unintentional strum
- Methods that do not involve receiving at least one polyphonic electrical signal from a multi-stringed instrument
- Detecting unintentional strums without comparing received notes to templates based on open string tuning

## The clever bit

The innovation lies in the system's ability to not only generate harmonies but also to intelligently filter out unwanted musical 'noise,' specifically unintentional strums on stringed instruments, ensuring the generated harmony remains musically coherent.

## Real-world examples

1. Music production software (DAWs) with auto-harmony features
2. Digital audio workstations (DAWs)
3. Live performance accompaniment systems
4. Virtual instrument plugins

## Why it matters

This patent addresses a core challenge in music production: creating convincing instrumental accompaniments and harmonies automatically. It provides a technical solution for software and hardware that can assist musicians, from hobbyists to professionals, in arranging and performing music by intelligently adding harmonic layers.

## Frequently asked questions

### What does How to Automatically Generate Musical Harmonies from Audio cover?

This 2012 patent describes a system that listens to music and automatically generates harmony notes to accompany a melody, even detecting and ignoring accidental strums on stringed instruments.

### Who owns patent US 8168877?

Harman International Industries Canada Ltd owns this patent, granted in 2012.

### When does this patent expire?

This patent is expected to expire on October 2, 2027, when the invention enters the public domain.

### What is patent US 8168877 cited by?

This patent has been cited by 24 later patents that build on its ideas.

### What problem does this patent solve?

This patent addresses a core challenge in music production: creating convincing instrumental accompaniments and harmonies automatically. It provides a technical solution for software and hardware that can assist musicians, from hobbyists to professionals, in arranging and performing music by intelligently adding harmonic layers.

### What does this patent NOT cover?

Generating harmonies without analyzing the spectral content of an accompaniment signal

**Full plain-English explainer:** https://patentbrief.org/patent/us/8168877/musical-harmony-generation-from-polyphonic-audio-signals

**Original patent:** https://patents.google.com/patent/US8168877

---

_Source: PatentBrief — https://patentbrief.org. Patent facts are from public records; the plain-English explanation is PatentBrief's._


## Related patents

Semantically similar inventions in the PatentBrief corpus:

- [How Real-Time Pitch Correction and Auto-Tune Technology Works](https://patentbrief.org/patent/us/5973252/auto-tune-pitch-correction) — A method for detecting the pitch of a musical note and adjusting it to a target frequency in real time using auto-correlation.
- [How Music Apps Learn What You Don't Want in Playlists](https://patentbrief.org/patent/us/12277178/media-content-item-recommendation-system) — This patent describes how a music streaming service learns what kinds of songs or artists a user dislikes for their playlists by tracking what they repeatedly ignore, then uses that information to avoid recommending similar things in the future.
- [How Digital Audio Compression Works](https://patentbrief.org/patent/us/5579430/digital-encoding-process) — A foundational method for compressing digital audio by transforming sound into spectral data and using variable-length codes to store it efficiently.
- [How Smart Speakers Know You're Talking to Them After a Command](https://patentbrief.org/patent/us/11361763/detecting-system-directed-speech) — This patent describes how a smart speaker system can tell if follow-up speech is meant for it, even without a "wake word," by analyzing voice activity and partial speech recognition results using an AI model.
- [How Computers Automatically Label Different Speakers in Audio Recordings](https://patentbrief.org/patent/us/6424946/amazon-personalized-recommendations) — A method for identifying and labeling speakers in audio recordings, even if the system has never heard the person speak before, by grouping similar voices and asking a user to name them.
