Skip to content

Dataset · ccmusic-database

Instrument timbre evaluation base

ccmusic-database/instrument_timbre

Audio recordings of 61 musical instruments with expert timbre ratings and spectrograms.

What's inside

Based on the National Timbre Testing Database: participants with musical training evaluated the sound of 37 Chinese and 24 Western instruments across 16 characteristics (bright, dark, husky, etc.). Alongside audio and text labels in Chinese and English, the dataset includes 10 reports with spectrogram analysis. Total size of the set is under 1,000 records.

What you can do with it

  • Training musical instrument classification and timbre analysis models
  • Creating automated sound tagging systems based on subjective characteristics
  • Multimodal research at the intersection of audio and spectrogram images
Tasks
Audio classification
Languages
Chinese, English
Size
up to 1K rows
Downloads this month
472
Likes
34

How to load it

from datasets import load_dataset

ds = load_dataset("ccmusic-database/instrument_timbre")