Audio Piano Triads Dataset

Created by: Agustín Macaya Valladares
Date: May 5th, 2021

- Dataset contains 43.200 examples of piano triads in .wav format.

Details:
- Sample rate: 16000 Hz.
- Data type: 16-bit PCM (int16).
- File size: Each example has a file size of 128 kB (5.53 GB for complete dataset).
- Duration: 4 seconds.
- Sound: Piano (digital).
- Chords were played by a human on a velocity-sensitive piano keyboard.
- 3 seconds pressed, 1 second released.

- 3 octaves (2,3,4).
- 12 base notes per octave: Cn, Df, Dn, Ef, En, Fn, Gf, Gn, Af, An, Bf, Bn. (n is natural, f is flat).
- 4 triad types per note: major (j), minor (n), diminished (d), augmented (a). No inversions.
- 3 volumes per triad: forte (f), metsoforte (m), piano (p).

- 10 original examples per combination of octave, base note, triad type, and volume. (103*12*4*3 = 4.320 examples).
- x10 data augmentation for each example (4.320 * 10 = 43.200 total examples).
- Data augmentation through random temporal and amplitude shifts.
- Metadata is in the name of the chord. For example: "piano_3_Af_d_m_45.wav" is a piano chord, (3) 3rd octave, (Af) A flat base note, (d) diminished, (m) metsoforte, 45th example.

Note:
- The audios are in 16-bit PCM (int16) data type to reduce the file size. This means that the dynamic range of values in the array is -32768 to 32768, integers. To normalize the audios in the range -1 to 1 just divide by 32768.