top of page

Purchase Strings Dataset

"Strings" is an AI music dataset that showcases the power of string instruments. From violins to cellos, this can be used for the transformative potential of string instruments, whether in generative music, investigating MIR tasks, or pioneering new techniques for processing string instruments.

Untitled design (3).png

Dataset Specifications

Total Audio Tracks: Up to 100k Strings tracks
Type: Instrument (Brass)
File Format: WAV, FLAC, MP3, CSV, JSON

Dataset includes:

  • Duration

  • Key

  • Tempo

  • BPM Range

  • Mood

  • Energy

  • Description

  • Keywords

  • Chord Progressions

  • Timestamps

  • Time Signature

  • Number of Bars

Purchasing License

Annual License
Contact Us
Perpetual License
Contact Us

Free audio sample available

The Strings Dataset is a comprehensive collection of audio tracks paired with detailed metadata, tailored to enhance machine learning applications in the domain of string instruments. This extensive dataset provides a diverse array of musical performances, capturing the nuances and intricacies of string-based compositions across various genres and styles.

Complementing the audio tracks are meticulously curated metadata annotations, offering insights into the musical structure, instrumentation, key signatures, tempo variations, timestamps, and more. These annotations empower machine learning models to grasp the complexities of string-based compositions, facilitating tasks such as generative music composition, music information retrieval (MIR), source separation, and beyond.

The Strings Dataset serves as a versatile resource, providing a platform for exploration and innovation in machine-driven music synthesis and analysis. Whether you're seeking to create expressive compositions, extract meaningful insights from musical data, or advance the boundaries of computational musicology, this dataset offers a wealth of opportunities for experimentation and discovery.

Untitled design (3).png

Contact Us

Thanks for submitting!

bottom of page