The Limitations of Traditional Audio Signal Processing

Typical Digital Signal Preprocessing Methods

• Several popular implementations in various languages
• De facto standard for digital signal processing applications
• Constraints due to frequency and time dependency
• Outputs with misrepresentations and unintended byproducts

Reconsidering the Biological Models

• Frequency domain conversion without window sizes
• Retains time domain information

Applications

Seismic Audio Analysis

Collaborative research in oil discovery audio analysis

• Signal extraction from drilling rigs, logging trucks, and boreholes
• Correlation on primary signal with sensors
• Handles noise and harmonics

Sound Isolation

Business competition entry for hearing aid technology:

• Software IP: Simplified algorithm, smartphone app, low cost
• Hardware IP: Small circuit footprint, ultra-low power
• Real-time sound separation (e.g., piano strings vs. key taps)

Note: Extracted only the string sounds, leaving the hammer sound with the voice

Voice Color and Intonation

• Extracts voice color with minimal training
• Intonation removal and manipulation
• Voice compression without sacrificing frequency bands