metadata

license: other
license_name: custom
license_link: >-
  https://huggingface.co./datasets/RoyalCities/Vocal_Textures_Dataset/blob/main/LICENSE.md
language: en
tags:
  - audio
  - music-generation
  - sample-generation
  - vocals
  - choir
  - fine-tuning
  - stable-audio
pretty_name: Vocal Textures Dataset

Vocal Textures Dataset

Introduction

The Vocal Textures Dataset is a curated collection of vocal choir audio samples designed to facilitate the fine-tuning of machine learning models, particularly in the domain of audio generation. This dataset is especially useful for hobbyists and researchers who are working on projects related to music production, AI-generated audio, and similar fields.

This dataset can be used to train non-commercial models and is structured to make it easy to navigate different types of vocals based on gender, musical key, and tempo. It features standardized file names and can be used out of the box for fine-tuning with Stable Audio Open.

Dataset Breakdown

Vocal Type	Major Progression (4 bars)	Major Progression (8 bars)	Minor Progression (4 bars)	Minor Progression (8 bars)
Male	420	420	420	420
Female	420	420	420	420
Ensemble	420	420	420	420

Total .wav files: 5040
Total duration: 980.64 minutes
Average Length: 11.67 seconds
Total Size: 14.50 GB
Sample Rate: 44100 Hz

Dataset Structure

The dataset is organized into the following folder structure:

Vocal_Textures_Dataset/
├── Ensemble/
│   ├── Major/
│   │   ├── 100BPM/
│   │   │   ├── Female Vocal Texture, chord progression, A major, 100BPM, 4 bars.wav
│   │   │   ├── Female Vocal Texture, chord progression, A major, 100BPM, 8 bars.wav
│   │   │   ├── Female Vocal Texture, chord progression, A# major, 100BPM, 4 bars.wav
│   │   ├── 110BPM/ (similar structure as 100BPM)
│   │   ├── 120BPM/ (similar structure as 100BPM)
│   ├── Minor/ (similar structure as Major)
├── Female/
│   ├── Major/
│   │   ├── 100BPM/ (similar structure as Ensemble > Major)
│   │   ├── 110BPM/ (similar structure as 100BPM)
│   ├── Minor/ (similar structure as Major)
├── Male/
│   ├── Major/ (similar structure as Ensemble > Major)
│   ├── Minor/ (similar structure as Major)

Key Signatures

This dataset is balanced in covering all 12-tone key signatures written with sharps only.

Minor Scales
A minor, A# minor, B minor, C minor, C# minor, D minor, D# minor,
E minor, F minor, F# minor, G minor, G# minor

Major Scales
A major, A# major, B major, C major, C# major, D major, D# major,
E major, F major, F# major, G major, G# major

BPM and Bars

This dataset also features equal coverage in the following BPMs and Bars.

BPMs
100BPM, 110BPM, 120BPM, 128BPM, 130BPM, 140BPM, 150BPM

Bars
4 bars, 8 bars

Example Files

Here are a few example file names from the dataset:

Female Vocal Texture, chord progression, A major, 100BPM, 4 bars.wav
Ensemble Vocal Texture, chord progression, B minor, 128PM, 8 bars.wav
Female Vocal Texture, chord progression, A# major, 110BPM, 4 bars.wav
Male Vocal Texture, chord progression, D# minor, 140BPM, 8 bars.wav

Usage

This dataset is ideal for training machine learning models, particularly in the area of audio generation. Below are a few examples of how you might use the dataset:

Fine-Tuning Models: Use the audio samples to fine-tune models such as the one available at RC_Vocal_Textures.
Music Production: Producers can use the .wav files in their projects, including commercial songs, films, and multimedia projects.
Research and Experimentation: Ideal for non-commercial research projects exploring AI-generated music or audio synthesis.

Licensing

This dataset is licensed under a Custom License based on CC BY-NC 4.0. You may use, share, and adapt the dataset for non-commercial purposes, including the training of machine learning models. The .wav files may also be used in commercial creative works such as music production. However, the dataset may not be used for commercial machine learning purposes, including the creation of derivative models intended for commercial use.

For the full terms, please refer to the LICENSE file included in the dataset.

Citation

If you use this dataset in your research or project, please cite it as follows:

RoyalCities. (2024). Vocal Textures Dataset. Hugging Face. https://huggingface.co./datasets/RoyalCities/Vocal_Textures_Dataset/