Harmonic-Frontier-Audio/Inhales_and_Reverse_Phonation_Preview

Harmonic Frontier Audio – Inhales and Reverse Phonation (Preview, v0.9)

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

Inhales and Reverse Phonation (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 speech-related ingressive airflow gestures and reverse phonation behaviors — non-lexical vocal primitives produced with inward airflow rather than conventional egressive airflow.

The full dataset encompasses:

• sharp involuntary-style gasps with abrupt ingressive onsets
• controlled sustained inhalations with minimal voicing
• voiced ingressive phonation in low and mid registers
• whispered ingressive airflow without vocal-fold vibration
• continuous inhale → exhale transitions in low and mid registers
• continuous exhale → inhale transitions in low and mid registers
• reverse-vowel articulations with stable mouth shape and inward airflow
• stable, speech-like reverse phonation without musical stylization

These characteristics make the dataset valuable for AI speech and voice modeling, phonetics research, reverse-phonation synthesis, airflow-direction modeling, expressive audio generation, respiratory gesture 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/inhales-reverse-phonation

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

---

🌬️ About Inhales and Reverse Phonation

Ingressive airflow occurs when air moves inward through the vocal tract during a respiratory or vocal gesture, reversing the direction used in ordinary exhaled speech. It can appear in natural behaviors such as gasps and controlled inhalations, as well as in voiced or whispered reverse-airflow gestures.

Reverse phonation extends this behavior into speech-adjacent vocal production, including voiced inhales, whispered inhales, reverse-vowel articulation, and continuous transitions between ingressive and egressive airflow. The dataset captures these behaviors without overtone techniques or musical stylization.

These phenomena are foundational to:

• airflow-direction-aware speech and voice synthesis
• reverse-phonation and ingressive-voice classification
• respiratory and expressive gesture modeling
• modeling transitions between inward and outward airflow
• phonetic study of voiced and unvoiced ingressive behaviors
• embodied and multimodal agent sound generation

This dataset presents a neutral, non-linguistic, non-performative representation of ingressive airflow and reverse phonation.
It is not designed as a collection of musical reverse-vocal techniques or semantic speech, but rather to isolate acoustic primitives underlying gasps, inhalations, reversed articulation, and bidirectional airflow transitions.

---

📂 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 3 representative audio files, selected to demonstrate:

• a sharp gasp — a brief involuntary-style inhalation with rapid pulmonic ingressive airflow and an abrupt onset
• a whispered inhale — sustained ingressive airflow without vocal-fold vibration
• a low-register inhale → exhale transition — a continuous shift from ingressive to egressive airflow

---

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.
• Performer positioned approximately 3.5 inches from the microphone, with the microphone approximately 10–15 degrees off-axis.
• Burst gestures were recorded at approximately 1–2 seconds, sustained conditions at 4–6 seconds, and airflow transitions at 6–8 seconds.
• Voiced conditions used comfortable low and mid register zones rather than prescribed exact pitches.
• Performances remained speech-related and non-lexical, with consistent microphone distance and no musical stylization or overtone techniques.

---

🌈 Spectrogram Preview

Below is a spectrogram illustrating the broadband ingressive airflow energy, abrupt gasp transients, sustained reverse-airflow texture, and changing spectral structure characteristic of inhales and transitions between ingressive and egressive airflow:

⚡ Usage

This preview pack is designed for:

• Evaluation of Harmonic Frontier Audio dataset quality and structure
• Testing AI and DSP systems that model ingressive airflow, reverse phonation, gasps, and airflow-direction transitions
• Research in phonetics, speech synthesis, respiratory audio, and speech-adjacent reverse vocal behavior
• Creative and technical applications requiring controlled inhaled, reversed, and bidirectional vocal-airflow primitives

👉 Note: This is not a full dataset.
The complete Inhales and Reverse Phonation dataset includes sharp gasps, controlled inhales, voiced and whispered ingressive phonation, bidirectional inhale/exhale transitions, and reverse-vowel articulations across relevant low and mid register conditions, and is available for licensing.

---

💡 Full Dataset Availability

This is a preview pack of the Inhales and Reverse Phonation 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/Inhales_and_Reverse_Phonation_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

• Human Vocality Primitives Series (Previews)
• Celtic Constellation Series (Previews)
• Extended Vocal Techniques Spectrum (Previews)
• Novelty Gems Cabinet (Previews)

(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 (March 2026) – Initial Preview Pack release for Inhales and Reverse Phonation.
See CHANGELOG.md for detailed version history.

---

Citation

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

Pullen, B. (2026). Inhales and Reverse Phonation Dataset (Preview) [Data set]. Harmonic Frontier Audio. Zenodo. https://doi.org/10.5281/zenodo.21989831

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

BibTeX

@dataset{pullen_2026_inhalesandreversephonation_preview,
  author       = {Blake Pullen},
  title        = {Inhales and Reverse Phonation Dataset (Preview)},
  year         = {2026},
  publisher    = {Harmonic Frontier Audio},
  version      = {0.9},
  doi          = {10.5281/zenodo.21989831},
  url          = {https://doi.org/10.5281/zenodo.21989831}
}