Why GEN‑B Generates Audio Mathematically Instead of Playing Audio Files

The Mathematics Are the Signal: Inside GEN-B's Algorithmic Tone Generator

Professional reference signals must be predictable, repeatable, and measurable. That principle shaped the audio architecture of GEN-B from the beginning.

GEN-B does not reproduce prerecorded WAV, AIFF, MP3, FLAC, AAC, M4A, or other audio assets. Its tone block is a real-time digital signal generator. Every output sample is calculated mathematically, assembled into a continuous digital stream, and delivered directly to the device audio engine.

This distinction matters. GEN-B is not playing a recording of a test signal. GEN-B is creating the test signal.

From a Mathematical Model to a 48 kHz Digital Stream

A continuous sine wave can be described mathematically by its amplitude, frequency, and phase. In a digital system, that continuous function is evaluated at discrete moments in time.

For a sine wave, a simplified sample equation is:

y[n] = A · sin(2πfn/Fs + φ)

Where:

• A is amplitude.
• f is frequency in hertz.
• Fs is the sample rate.
• n is the sample index.
• φ is phase.

At a sample rate of 48 kHz, the generator calculates 48,000 samples every second. The interval between samples is approximately 20.833 microseconds. Each calculated value becomes part of the outgoing digital waveform.

Frequency, amplitude, phase, channel assignment, and waveform type can therefore be changed by modifying parameters in the synthesis model rather than loading another recording.

Why GEN-B Does Not Use Audio Files

A prerecorded test file can be useful, but it introduces a different architecture and a different set of limitations. The signal is fixed at the moment the file is created. Long-duration playback may require looping. A poorly prepared loop may introduce a discontinuity at the splice point. Compressed formats may add coding artifacts, while every file-based implementation also requires asset storage, version control, loading, and playback management.

GEN-B avoids those dependencies.

Its synthesis engine provides:

• No prerecorded audio.
• No file loading or waveform storage.
• No compression artifacts.
• No loop boundaries or splice points.
• Continuous duration for as long as the generator remains active.
• Immediate parameter changes.
• Repeatable output derived from the same mathematical model.

The Sine Wave: The Universal Reference Signal

The sine wave is the fundamental reference waveform for audio measurement because, ideally, its energy is concentrated at one frequency. This makes it especially useful when the goal is to observe how a system treats a known, controlled input.

In professional workflows, sine waves are used to evaluate or align:

• Gain.
• Audio levels.
• Frequency response.
• Harmonic and intermodulation distortion.
• Noise floor and dynamic range.
• Analog and digital signal paths.
• Channel balance and phase relationships.

A 1 kHz sine wave is particularly common because it lies in the midband of the audible spectrum and can pass through most professional audio systems without being dominated by low-frequency coupling behavior or high-frequency bandwidth limitations.

The waveform itself is mathematically simple, but its value as a reference is enormous: when the input is known precisely, deviations at the output become measurable evidence of system behavior.

The Square Wave: Harmonics, Edges, and System Behavior

A square wave is very different from a sine wave. An ideal symmetrical square wave contains a fundamental frequency and a predictable series of odd harmonics. Its rapid transitions challenge a signal path in ways that a single-frequency sine wave cannot.

Square-wave testing can help reveal:

• Transient response.
• Amplifier behavior.
• Loudspeaker and crossover behavior.
• Digital processing limitations.
• Timing and phase characteristics.
• Ringing, overshoot, and undershoot.
• Bandwidth limitations.
• Signal integrity problems.

No physical audio system can reproduce an infinitely fast transition or an unlimited harmonic series. The resulting changes in edge shape, settling, symmetry, and ringing can therefore expose bandwidth, phase, filtering, or stability problems.

Reference Levels Across Professional Workflows

Professional facilities do not all use the same digital alignment level.

A widely used North American broadcast practice is a 1 kHz reference tone at −20 dBFS. Many European and international workflows use −18 dBFS, commonly associated with EBU alignment practice.

These values are not competing definitions of correctness. They are operating conventions that preserve different amounts of headroom above the reference level. GEN-B supports both so engineers can work with the convention used by the facility, broadcaster, production, or distribution chain under test.

GEN-B also provides commonly encountered analog and digital reference selections, including 0 dBFS and +4 dBu-related workflows. Analog dBu and digital dBFS are different measurement domains; their relationship depends on the alignment convention and equipment calibration used by a particular facility.

White Noise: Equal Power per Hertz

White noise contains energy distributed uniformly per unit of bandwidth. In practical terms, equal-width frequency bands contain approximately equal power.

Because wide portions of the spectrum are excited simultaneously, white noise is useful for:

• Broadband system verification.
• Filter and equalizer evaluation.
• Electronic noise and measurement tests.
• Loudspeaker and component testing.
• Identifying frequency-dependent behavior.

White noise sounds bright because a linear frequency scale contains many more hertz in the upper octaves than in the lower octaves.

Pink Noise: Equal Power per Octave

Pink noise decreases in power density as frequency rises, approximately 3 dB per octave. This produces roughly equal energy in each octave and better corresponds to octave-based acoustic analysis and the way humans perceive frequency distribution.

Pink noise is commonly used for:

• Room tuning.
• Studio and control-room calibration.
• Broadcast monitoring environments.
• Cinema and installed sound systems.
• Loudspeaker optimization.
• Real-time analyzer and equalization work.

White noise and pink noise serve different purposes. White noise is uniform per hertz; pink noise is balanced per octave.

Frequency Selection Is Part of the Instrument

A professional generator must support more than one fixed frequency.

Low-frequency tones help evaluate bass response, coupling, resonance, polarity, and mechanical behavior. Midband tones are useful for general alignment and gain verification. High-frequency tones can reveal bandwidth restrictions, filtering, converter behavior, and other upper-spectrum limitations.

Because GEN-B generates its output algorithmically, changing frequency does not require a different file. The synthesis engine simply evaluates the selected mathematical model using a new frequency parameter.

Continuous Generation, Stable Control

Real-time synthesis provides several operational advantages:

Unlimited signal duration

The waveform is not a finite recording. It continues for as long as the generator is active.

Immediate level and frequency changes

Parameters can be changed directly in the model without searching for or loading another asset.

Repeatability

Given the same mathematical parameters and operating conditions, the engine produces the same reference waveform consistently.

Small asset footprint

The application does not need a library of long audio recordings for every combination of waveform, level, frequency, and channel configuration.

Expandability

New signals can be introduced by adding mathematical models and control logic rather than recording and distributing additional media assets.

A Foundation for Future Test Signals

The same architecture can support more advanced measurement signals, including:

• Linear frequency sweeps.
• Logarithmic sweeps.
• Multitone signals.
• Impulse signals.
• Tone bursts.
• Identification tones.
• Channel-specific test sequences.
• Specialized broadcast and audiovisual reference signals.

The engine remains the same. The mathematical model changes.

Engineering Philosophy

GEN-B was designed as an engineering instrument, not as a collection of media files.

Its tone generator transforms equations into samples, samples into digital waveforms, and digital waveforms into professional reference signals. The result is a flexible architecture built for alignment, testing, education, troubleshooting, and future expansion.

There are no prerecorded tones pretending to be a signal generator.

The mathematics are the signal.