Datasets:

Harmonic-Frontier-Audio
/

Whisper_and_Aspiration_Preview

Tasks:

Other

Modalities:

Formats:

Languages:

Size:

Tags:

Libraries:

License:

Dataset card Data Studio Files Files and versions

xet

Community

Dataset Viewer

Auto-converted to Parquet Duplicate

Split (1)

train · 5 rows

Search is not available for this dataset

audio audioduration (s) 5.63 7

Harmonic Frontier Audio – Whisper and Aspiration (Preview, v0.9)

A high-fidelity human vocal dataset designed for AI training, speech research, and expressive voice modeling.

Whisper and Aspiration (Preview), created by Harmonic Frontier Audio, provides a compact reference set demonstrating the quality, formatting, and metadata conventions used in the Harmonic Frontier Audio Human Vocality Primitives series.

🔎 Summary

This dataset provides high-quality, rights-cleared recordings of whisper phonation and aspiration-based vocal gestures — fundamental airflow-driven vocal behaviors that exist at the boundary between voiced speech and unvoiced breath noise.

The recordings emphasize:

airflow texture
mouth shape variation
fricative articulation
transitions into and out of silence

These characteristics make the dataset valuable for AI speech modeling, phonetics research, voice synthesis, breath modeling, and human-aligned vocal control systems.

Developed by Harmonic Frontier Audio, this preview follows The Proteus Standard™ for dataset provenance, transparency, and ethical AI use.
Learn more about the Proteus Standard → https://harmonicfrontieraudio.com/proteus-standard

Full dataset details and licensing information are available at:
https://harmonicfrontieraudio.com/datasets/whisper-aspiration

If you find this dataset useful, please consider giving it a 🤍 on Hugging Face to help others discover it.

🌬️ About Whisper and Aspiration

Whisper phonation occurs when air passes through the vocal tract without sustained vocal fold vibration, producing noise-shaped speech sounds rather than pitched phonation.
Aspiration refers to breath-driven sound components that accompany or replace voiced articulation, including transitions into silence and near-noise airflow events.

These phenomena are foundational to:

speech production
expressive voice synthesis
breath-aware AI voice systems
phonetic and physiological modeling

This dataset presents a neutral, non-linguistic, non-performative representation of whisper and aspiration.
It is not designed to encode semantic speech content, but rather to isolate acoustic primitives that underlie whisper-based vocal behavior.

📂 Contents

Audio Files (.wav)

Recorded at 96 kHz / 24-bit WAV format
Exported as mono
Fade-ins and fade-outs of 3–5 ms applied for consistency
No compression, normalization, or creative processing applied
High-pass filtered at ~40 Hz to remove subsonic rumble

This preview includes 5 representative audio files, selected to demonstrate:

sustained whisper phonation
neutral and constrained mouth shapes
fricative-based whisper gestures
transitions into silence

Metadata (.csv)

Includes structured fields for:

file name
sound source type
airflow type
phonation type
gesture and articulation descriptors
microphone and recording chain
sample rate, bit depth, and dataset version

Metadata follows the Harmonic Frontier Audio – Foundations schema.

🎤 Recording Notes

Recorded in a treated studio environment using a single-mic setup:
- Microphone: Rode NT1-A condenser microphone
- Recording chain: Rode NT1-A → Zoom F8n Pro
Captured at 96 kHz / 32-bit float, rendered as 96 kHz / 24-bit mono WAV for release.
Natural room tone and subtle breath noise were preserved to retain acoustic realism.

🌈 Spectrogram Preview

Below is a spectrogram illustrating the broadband noise structure, airflow turbulence, and formant-shaped energy characteristic of whisper phonation and aspiration-based vocal gestures:

⚡ Usage

This preview pack is designed for:

Evaluation of Harmonic Frontier Audio dataset quality and structure
Testing AI and DSP systems that model unvoiced or breath-driven sounds
Research in phonetics, speech synthesis, and expressive vocal modeling
Creative sound design involving breath, noise, and vocal texture

👉 Note: This is not a full dataset.
The complete Whisper and Aspiration dataset includes a substantially larger set of whisper, aspiration, and airflow primitives and is available for licensing.

💡 Full Dataset Availability

This is a preview pack of the Whisper and Aspiration Dataset.
The complete dataset is available for commercial licensing.

For licensing inquiries:
📩 info@harmonicfrontieraudio.com

📥 How to Use This Dataset in Python

You can load the Parquet-converted version of this dataset directly with the datasets library:

from datasets import load_dataset

dataset = load_dataset(
    "Harmonic-Frontier-Audio/Whisper_and_Aspiration_Preview",
    split="train"
)

print(dataset)

⚙️ Note: Parquet conversion and load_dataset() support will be available within 2–3 days of publication.

🔗 Explore More from Harmonic Frontier Audio

(All datasets follow The Proteus Standard™ for ethical dataset provenance and licensing.)

📜 License

Released under CC BY-NC 4.0.

Free for non-commercial use, testing, and research
Commercial licensing available via Harmonic Frontier Audio
A formal rights declaration is included in this dataset bundle

📧 Contact

Harmonic Frontier Audio
📩 info@harmonicfrontieraudio.com
🌐 https://harmonicfrontieraudio.com/

🗒️ Release Notes

Version 0.9 (Jan. 2026) – Initial Preview Pack release for Whisper and Aspiration.
See CHANGELOG.md for detailed version history.

Citation

If you use this dataset in your research, please cite:

Pullen, B. (2026). Whisper and Aspiration Dataset (Preview) [Data set]. Harmonic Frontier Audio. Zenodo. https://doi.org/10.5281/zenodo.18228940

ORCID: https://orcid.org/0009-0003-4527-0178

BibTeX

@dataset{pullen_2026_whisperandaspiration_preview,
  author       = {Blake Pullen},
  title        = {Whisper and Aspiration Dataset (Preview)},
  year         = {2026},
  publisher    = {Harmonic Frontier Audio},
  version      = {0.9},
  doi          = {10.5281/zenodo.18228940},
  url          = {https://doi.org/10.5281/zenodo.18228940}
}

Downloads last month: 236

Size of downloaded dataset files:

8.94 MB

Size of the auto-converted Parquet files:

8.93 MB

Number of rows: