Hydrophones & Underwater Recording: Physics, Hardware, and Calibration

A hydrophone is an underwater transducer that converts acoustic pressure fluctuations in water into electrical voltage signals. Just as a microphone senses sound pressure in air, a hydrophone captures sound waves propagating through liquid media. However, because seawater has an acoustic impedance approximately 3,600 times greater than air, underwater recording equipment requires specialized transducer materials, low-noise preamplifiers, rigid waterproof encapsulation, and precise laboratory calibration.

Understanding underwater recording requires examining the piezoelectric materials that drive hydrophones, the mathematical standards governing sensitivity calibration, and the ambient noise floors of ocean soundscapes.

Piezoelectric Transducer Physics

Most scientific hydrophones rely on the direct piezoelectric effect, primarily utilizing polycrystalline ceramic elements such as lead zirconate titanate (PZT):

[ Acoustic Pressure Wave (p) ] ──> [ Ceramic Element Compression ]
                                                 │
                                                 ▼
[ Electrical Voltage (V) ] <── [ High-Impedance Preamplifier ] <── [ Surface Polarization Charge (q) ]

1. Pressure to Charge: When an acoustic wave exerts mechanical stress on a polarized PZT crystal, the crystal lattice distorts, creating an electric dipole moment and generating an electrical charge:

q = d₃₃ · F

(where d₃₃ is the piezoelectric charge coefficient in C/N, and F is mechanical force). 2. Voltage Generation: This accumulated surface charge produces an open-circuit voltage proportional to acoustic pressure:

V = g₃₃ · p · t

(where g₃₃ is the piezoelectric voltage coefficient, p is acoustic pressure in μPa, and t is crystal thickness). 3. Preamplifier Impedance Matching: Because piezoelectric ceramics have extremely high output impedance at low frequencies, the sensor must be coupled directly to an internal or adjacent high-input-impedance (typically greater than 100 MΩ) field-effect transistor (FET) preamplifier before the signal travels down long underwater cables.

Hydrophone Sensitivity and Calibration

Hydrophone voltage sensitivity (M_h) measures the electrical output generated per unit of acoustic pressure:

M_h = V_rms / p_rms    (Volts per micropascal, V/μPa)

In logarithmic decibels, sensitivity is expressed relative to one volt per micropascal:

S_h = 20 · log₁₀(M_h)    (dB re 1 V/μPa)

Typical scientific hydrophones exhibit sensitivities ranging from -165 dB re 1 V/μPa (high sensitivity for quiet oceanic ambient noise) to -210 dB re 1 V/μPa (lower sensitivity designed to record intense seismic airguns or nearby dolphin echolocation clicks without clipping).

Calibration Methodologies

To ensure quantitative accuracy, hydrophones are calibrated using three standard techniques:

• Reciprocity Calibration: An absolute calibration standard requiring no known reference sensor, utilizing three reversible transducers where one acts as an acoustic projector, one as a hydrophone, and one as a reciprocal transducer.
• Comparison Calibration: Measuring the test hydrophone alongside a secondary laboratory reference hydrophone (traceable to national standards institutes like NIST or NPL) in an acoustic test tank.
• Pistonphone Calibration: Used for low-frequency calibration (e.g., 5 Hz to 250 Hz) inside an enclosed fluid-filled chamber driven by a motorized oscillating piston.

Nyquist-Shannon Sampling for Delphinid Bioacoustics

A common error in marine bioacoustics is selecting sampling rates based on human speech limits (44.1 kHz or 48 kHz).

According to the Nyquist-Shannon sampling theorem, continuous analog signals must be sampled at a rate strictly greater than twice the highest frequency component present (f_s > 2 · f_max) to avoid irreversible aliasing artifacts:

Minimum Sampling Rate: f_s > 2 · f_Nyquist

As documented by Lammers, Au, and Herzing (2003), odontocete acoustic energy spans up to and beyond 130 kHz:

• Tonal Whistles: Fundamentals up to 24 kHz; harmonics extending beyond 50–100 kHz.
• Echolocation Clicks: Energy peaks between 40 kHz and 130+ kHz.

Recording delphinid echolocation biosonar without aliasing requires hardware capable of sampling at 192 kHz, 256 kHz, or 384 kHz, coupled with sharp analog anti-aliasing low-pass filters prior to analog-to-digital conversion (ADC).

The Ocean Ambient Noise Floor (Wenz Curves)

Underwater recordings must always be interpreted against the background acoustic noise of the ocean. In 1962, Gordon Wenz published the foundational empirical curves (the Wenz curves) defining ocean ambient noise spectra:

Spectral Noise Level (dB re 1 μPa² / Hz)
100 ┌─── Heavy Commercial Shipping (5 – 50 Hz)
    │     ╲
 80 ┼──────╲─── Wind & Wave Breaking Noise (500 Hz – 20 kHz)
    │       ╲     (Correlated with Beaufort Sea State)
 60 ┼────────╲──────────────────────
    │         ╲                      Thermal Molecular Noise Limit
 40 ┼──────────╲──────────────────── (Above 50 – 100 kHz)
    └───────────┴───────────┴───────────┴───────────┴───────────>
    10 Hz       100 Hz      1 kHz       10 kHz      100 kHz (Frequency)

1. Low Frequencies (5 Hz – 100 Hz): Dominated by distant commercial shipping traffic and tectonic seismic events.
2. Mid Frequencies (500 Hz – 25 kHz): Dominated by local surface weather—breaking waves, spray, bubble oscillation, and rain impact—strongly correlated with Beaufort sea state.
3. High Frequencies (above 50 kHz): Dominated by molecular thermal agitation of water molecules, which increases by 20 dB per decade with frequency:

NL_thermal ≈ -15 + 20 · log₁₀(f_kHz)    (dB re 1 μPa² / Hz)

Practical Considerations for Marine Field Work

• Flow Noise & Mooring Dynamics: Water rushing past hydrophone casings creates turbulent pseudo-sound (low-frequency pressure fluctuations that do not propagate). Moorings utilize streamlined acoustic fairings and compliant bungee cords to decouple sensors from surface wave motion.
• Biofouling: In warm coastal waters, barnacles, algae, and encrusting bryozoans colonize hydrophone boots within weeks, altering transducer acoustic impedance. Long-term deployments utilize non-toxic silicone coatings or copper tape shrouds.
• Dynamic Range (Theoretical vs. Real-World Performance): While modern autonomous recorders feature 24-bit delta-sigma analog-to-digital converters (theoretically providing up to ~144 dB of quantization dynamic range), real-world system dynamic range is constrained to approximately 105 dB to 120 dB. The practical limit is governed by the preamplifier's self-noise floor, input thermal noise (Johnson-Nyquist noise), and the converter's Effective Number of Bits (ENOB, typically 18 to 20 bits in low-power marine instruments). Understanding this distinction is vital for setting preamplifier gain: researchers must ensure faint biological signals clear the analog noise floor without risking digital saturation during close-range odontocete click trains.