Pulse Gen · Calibrated signal generation

Transmit the signal you planned.

Pulse Gen turns your projector and amplifier calibration into drive-limited test signals, offshore-wind pile-driving sound and marine-mammal calls, plays them through an NI DAQ, and records the measured voltage, current and hydrophone response in the same project.

Windows

Pulse Gen · Pile Driving Noise
Pulse Gen pile driving noise settings with impact profile, strike rate and broadband components

Calibration first

Every level starts from the hardware you are driving.

Import the projector transmitting voltage response (TVR) and the amplifier gain curve, enter their voltage limits, and Pulse Gen uses them in every generation, playback and safety calculation. When a requested level would exceed a stored limit, Pulse Gen reduces the drive and records the limiting in the file notes.

  1. Create or select a project folder for generated files and sessions.
  2. Import the projector TVR and amplifier gain curves with their voltage limits.
  3. Configure one NI DAQ output and the sensor inputs: hydrophone, voltage probe, current probe or accelerometer.
  4. Generate a signal and review its summary: actual drive, limiting and the curves used.
  5. Play it, monitor the measured inputs and record the session.

How levels are set

For synthetic sources such as pile driving and marine-mammal calls, Pulse Gen predicts pressure from the complete generated spectrum using the amplifier gain and projector TVR at every frequency. The target is the maximum 125 ms Hann-window broadband RMS level, in dB re 1 µPa at 1 m. After the digital, amplifier and projector voltage limits are applied, the final level and any shortfall are recomputed from the saved WAV itself.

Calibrated generation needs measured curves that cover at least 99% of the generated drive energy. If they don’t, Pulse Gen reports the band it could cover and the unsupported fraction, instead of stretching a low-frequency projector’s curve over whistles or clicks. Each file’s sidecar keeps a copy of the curve points and DAC voltage used, so later edits to the curves can’t change what the file claims.

Estimates, stated as estimates: predictions are linear and magnitude-only. They don’t model phase, nonlinear distortion, directivity, propagation or the receiving hydrophone, so confirm the achieved level with a calibrated measurement channel.

Signal generation

Test signals for trials, commissioning and calibration.

Tone bursts

Repeated single-frequency pulses with cycle-based ramp-in and ramp-out tapers and integer-cycle finishing.

Frequency sweep

A stepped series of tone bursts between low and high frequencies, inside the calibrated band.

LFM and HFM

Linear and hyperbolic frequency-modulated pulses with selectable envelope shading.

Multitone

Up to five tones with explicit frequencies and relative phases, driven by SPL target or direct amplifier-input voltage.

Tonals with AM

A carrier tone with controllable amplitude-modulation frequency and depth.

Adaptive Showcase

Demonstration sequences chosen from the selected projector and amplifier curves, with resonance keep-away.

Synthetic sources

Play back realistic noise sources on demand.

Vessel

Propeller harmonics with frequency wander, flow and cavitation noise, and diesel, electric, gasoline or turbine engine signatures.

Wind and tidal turbines

Gearbox, direct-drive and floating offshore wind or tidal-turbine profiles with blade-pass modulation and foundation resonance.

Pile driving

Impact or vibratory driving, with offshore-wind styles, strike-by-strike diagnostics and a seed for repeatable output. More below.

Scuba

Breathing cycles, bubbles, regulator hiss, fin movement and gear noise.

Marine mammals

Synthesised calls for 15 species from learned parameter models. More below.

Offshore-wind pile driving

Pile-driving sound, measured the way guidance reports it.

Choose an offshore style for impact driving and Pulse Gen sets the strike envelope, decay and rate. The ring, thump and noise band follow the combined response of your amplifier and projector, and your manual changes survive switching styles or curves.

StyleDecay constantStrikes/minCharacter
Standard monopile90 ms50Sharp onset, sustained low-frequency body
Heavy monopile120 ms40Fuller thump and longer ring
Jacket pin-pile60 ms60Compact body and stronger onset texture
Soft start90 ms40Standard shape; the first eight strikes ramp up
Distant / reverberant180 ms50Softer onset, longer tail and delayed copies

Strike-by-strike diagnostics

  • Per-strike metrics: modelled pressure and envelope, the central 90% energy interval, peak SPL, T90 RMS SPL and single-strike SEL.
  • Spectrum check: the intended shape against the predicted acoustic output, with the projector’s TVR peak lobe shaded and a measure of how much spectral energy moved.
  • Output limits: requested against achieved level, showing which of the DAQ, amplifier or projector limits applied.
  • Progression: ramp the level up across the sequence and insert pauses every set number of strikes.
  • Spectrum strategy: optimise for the projector, or preserve a reference source shape with up to 24 dB of inverse compensation inside the calibrated band.
  • Electrical impedance: projector resistance, reactance and V/I phase can be overlaid, clearly labelled as electrical rather than acoustic phase.
Simulation presets, not specifications: the style values are design choices, not measurements of a particular pile or installation. Soft start is illustrative, not a mitigation protocol, and the distant style doesn’t calculate transmission loss. A narrow-band projector can reduce the audible difference between styles.

Marine-mammal calls

Synthesised calls for 15 species.

Marine Mammal Generation synthesises new calls from portable parameter models learned from annotated recordings. Tonal models capture pitch contours, amplitude envelopes, harmonics and spectral shape; click models capture centre frequency, bandwidth, pulse duration and interval patterns. Each call is built fresh, so the models contain no source audio.

Calls are scaled with the same calibrated level method as other synthetic sources. A continuous soundscape option keeps a quiet background running between calls and lets each call decay naturally, for playback that sounds less like isolated clips.

Spectrogram of a synthesised dolphin whistle generated by Pulse Gen
Spectrogram of a synthesised dolphin whistle (prototype model).
SpeciesTraining unitsNotes
Humpback whale306Tonal and textured calls; rare call families have little support
Blue whale128Low-frequency calls
Fin whale128Low-frequency calls; mostly unspecified call labels
Right whale128Upcalls
Dolphin96Wideband tonal calls, including whistles
Long-finned pilot whale75Tonal and whistle calls
Orca36Distinct tonal contours; pulsed and click calls not yet modelled
Sperm whale42Clicks with measured coda interval patterns
Cuvier’s beaked whale45Ultrasonic click trains; 192 kHz or higher output
Gervais’ beaked whale47Ultrasonic click trains; 192 kHz or higher output
Sowerby’s beaked whale43Ultrasonic click trains; 384 kHz or higher output
Sei whale11Limited prototype
Beluga whale9Small source collection
Gray whale2Limited reference parameters
Minke whale1Single reference; limited support

Ultrasonic species

Beaked-whale models require a minimum sample rate rather than shifting calls into an audible band. Your output hardware must support that rate and bandwidth.

Checked against Analyze

Generated files can be scored by Analyze’s trained classifiers as the predicted emitted sound, as the electrical drive signal, or as a real recording, each reported separately.

Call-based messaging

Encode short text (up to 128 bytes) as a sequence of synthesised calls, with a CRC check, and decode it from a saved WAV. Offline and unencrypted.

Acoustic prototypes, not certified species emulators: the models reproduce call acoustics, not animal identity, behaviour or song order. Species with few training units, such as gray, minke and beluga, need more independent recordings. A classifier score is not proof of realism; use independent references, spectrograms and listening before relying on a species model.

Playback, monitoring and recording

See the measured response while you transmit.

The playback monitor shows the commanded signal beside the measured inputs: voltage and current probes, hydrophones and accelerometers, across one or more NI DAQs. Input diagnostics flag missing samples and suspicious floating channels before they mislead a result.

  • Live traces and spectra for each input, with electrical impedance |V/I| from a voltage and current probe pair.
  • Measurement window that selects the highest-energy interval for detailed review.
  • Full-rate multichannel WAV recording, optional per-channel files and interval logs in CSV and SQLite.
  • Playlists and scheduled playback with per-file play counts, waits and start and stop times.
  • GPS logging from a serial receiver or a TCP NMEA stream, kept with the session.

Specialist tools

For the harder parts of a trial.

Harmonic cancellation

Measure and lock correction tones for the second to fifth harmonics of a fixed fundamental, with before-and-after spectra.

Propeller tone targets

Calculate shaft-rate and blade-pass targets, then measure the output-to-hydrophone transfer and latency response.

CTD propagation optimizer

Scan source depth and frequency against an imported CTD profile to find the combinations predicted to reach farthest.

Acoustic messaging

Encode text messages into acoustic WAV signals with a passphrase, or as offline marine-call sequences, and decode supported message WAVs.

WAV stitching

Join compatible WAV files in a chosen order, with optional gaps and a map of section boundaries.

Projects and sessions

Generated files, playback sessions, metadata and GPS records are organised beneath the active project.

Pulse Gen CTD propagation optimizer inputs for source level, receiver threshold, depths, frequency range and loss presets
CTD Propagation Optimizer: an engineering estimate for planning, not a ray-trace solver.

Editions

The full toolkit, or only the tools a project needs.

Pulse Gen is available as a full edition or as a tailored edition built around a specific project—for example, tone generation and pile-driving noise only. Every edition includes playback, DAQ setup, calibration curves, projects and GPS logging, and comes with a user manual written for the tools it contains.

Pricing: provided on request. Ask about pricing and tell us which tools and hardware you plan to use.

Pulse Gen questions

What teams usually ask first.

Does Pulse Gen need hardware?

Generating signals and reviewing files can be done without an attached DAQ. Playback, live acquisition and calibration need compatible NI DAQ hardware with the NI-DAQmx driver installed.

Is the requested SPL what the projector will produce?

Not necessarily. The requested level is a calibration-based estimate, and the drive can be reduced when stored limits are reached. Confirm the achieved level with a calibrated measurement channel.

How realistic are the marine-mammal calls?

They are acoustic prototypes learned from annotated recordings, not certified species emulators. Support varies by species: some models have over a hundred training units, others only one or two. Check them against independent recordings before relying on them.

Which operating system does it run on?

Windows.

Planning a sea trial or commissioning a projector?

Tell us about the projector, amplifier, DAQ and signals you need, and we’ll recommend an edition.

Ask about pricing