This glossary defines key terms used throughout V-Listen and in the broader field of Noise, Vibration, and Harshness (NVH) engineering. Terms are listed alphabetically.
| Term | Definition |
|---|---|
| Aliasing | Distortion that occurs when a signal is sampled at less than twice its highest frequency component (Nyquist criterion). Aliased frequency components appear folded into the analysis bandwidth as false peaks. Prevented by anti-aliasing filters and adequate sampling rate. |
| Auto Spectrum | The magnitude-squared representation of the FFT output, proportional to signal power at each frequency bin. Also called the power spectrum or auto-spectral density when normalized by frequency resolution. Units: (engineering unit)². |
| Averaging | The process of combining multiple spectral estimates to reduce variance. Common modes: linear (RMS), exponential (time-weighted), and peak hold (maximum retention). More averages yield smoother spectra at the cost of reduced time resolution. |
| BPFI / BPFO | Ball Pass Frequency Inner / Outer race. Characteristic frequencies at which rolling elements pass over a defect on the inner or outer race of a rolling element bearing. Fundamental to bearing condition monitoring via envelope analysis. |
| BSF | Ball Spin Frequency. The rotational frequency of a rolling element (ball or roller) about its own axis within a bearing. A defect on the rolling element surface produces impacts at 2×BSF. |
| Cepstrum | The inverse Fourier transform of the logarithm of a spectrum. Transforms periodic spectral patterns (harmonics, sidebands) into peaks in the quefrency domain. Variants: real, power, and complex cepstrum. Key tool for gear and echo diagnostics. |
| CMIF | Complex Mode Indicator Function. A technique for identifying natural frequencies and closely spaced modes from a set of FRFs. Computed as the singular values of the FRF matrix at each frequency line. Peaks in the CMIF indicate modal frequencies. |
| Coherence | A frequency-domain measure of the linear relationship between two signals, ranging from 0 (no linear correlation) to 1 (perfect linear correlation). Low coherence indicates nonlinearity, noise contamination, or unmeasured input sources. Used to assess FRF quality. |
| Cross Spectrum | The product of the FFT of one signal with the complex conjugate of the FFT of another. Contains both magnitude and phase information, representing the frequency-domain correlation between two signals. Foundation for FRF and coherence computation. |
| Damping Ratio (ζ) | A dimensionless parameter characterizing how oscillations decay after excitation. Expressed as a fraction of critical damping. Typical values in mechanical structures: 0.1%–5%. Estimated from FRF bandwidth (half-power method) or modal curve fitting. |
| dB (Decibel) | A logarithmic unit for expressing ratios. In acoustics: L = 20·log10(p/pref) with pref = 20 μPa. In vibration: reference depends on quantity (1 μm/s² for acceleration). Provides a compressed scale matching human perception. |
| DOF | Degree of Freedom. In modal analysis, specifies a measurement point and direction
(e.g., node 5, +Z direction). Written as 5Z+. Each FRF connects a response DOF
to a reference (excitation) DOF. |
| ENBW | Equivalent Noise Bandwidth. The bandwidth of an ideal rectangular filter that would pass the same noise power as the actual window function. For Hanning: ENBW = 1.5·Δf. Critical for correct power spectral density scaling. |
| Envelope Spectrum | The frequency spectrum of the amplitude envelope of a band-pass filtered signal, obtained via Hilbert transform demodulation. Reveals modulation frequencies caused by repetitive impacts in bearings and gears. |
| FFT | Fast Fourier Transform. An efficient algorithm for computing the Discrete Fourier Transform (DFT), converting time-domain signals to frequency-domain representation. Complexity: O(N·log N) vs. O(N²) for the direct DFT. |
| FRF | Frequency Response Function. The ratio of the output (response) spectrum to the input (excitation) spectrum of a system. Contains magnitude and phase. Types include receptance (displacement/force), mobility (velocity/force), and accelerance (acceleration/force). |
| FTF | Fundamental Train Frequency. The rotational frequency of a bearing cage (retainer). Cage defects produce sub-synchronous vibration at FTF and its harmonics. |
| Hanning Window | A raised-cosine window function applied to FFT blocks to reduce spectral leakage. The default and recommended window for most continuous signal analyses. Provides a good balance between frequency resolution and leakage suppression (−31 dB sidelobes). |
| IEPE | Integrated Electronics Piezo-Electric. A sensor interface standard where the transducer contains a built-in charge amplifier powered by a constant-current source (typically 2–4 mA) from the data acquisition device. Also known as ICP (PCB Piezotronics trademark). |
| Loudness | A psychoacoustic metric quantifying the perceived intensity of sound. Measured in sone (linear scale) or phon (dB-referenced). Computed according to ISO 532-1 (Zwicker method) or ISO 532-2 (Moore-Glasberg method). 1 sone = loudness of a 1 kHz tone at 40 dB SPL. |
| MAC | Modal Assurance Criterion. A scalar value (0 to 1) comparing two mode shape vectors. MAC = 1 indicates identical shapes; MAC = 0 indicates orthogonal shapes. Used to correlate experimental and analytical (FEA) mode shapes or to check measurement consistency. |
| Natural Frequency | The frequency at which a structure tends to oscillate when disturbed and allowed to vibrate freely. Each natural frequency corresponds to a mode shape. Also called eigenfrequency or resonance frequency (though resonance technically occurs at a slightly different frequency when damping is present). |
| Nyquist Frequency | Half the sampling rate. The maximum frequency that can be represented without aliasing in a sampled signal. All analysis results are limited to frequencies below the Nyquist frequency. |
| Order | A frequency component that is a fixed multiple of the shaft rotational speed. The 1st order equals the shaft frequency; the nth order equals n times the shaft frequency. Order analysis tracks these components as speed varies. |
| Overlap | The percentage of shared samples between consecutive FFT blocks. Compensates for data loss at the tapered edges of window functions. Typical values: 50% for Hanning, 66.7% for Flat Top. Increases the effective number of averages for a given signal length. |
| PSD | Power Spectral Density. The auto spectrum normalized by the frequency resolution (Δf), giving power per unit frequency (engineering unit²/Hz). Independent of FFT block size, making it suitable for comparing spectra computed with different resolutions. |
| Quefrency | The independent variable (horizontal axis) of the cepstrum, with units of time (seconds). Represents the period of a periodic pattern in the frequency spectrum. Derived by rearranging the letters in "frequency." |
| Sharpness | A psychoacoustic metric quantifying the perceived high-frequency content of a sound. Measured in acum. A sound with more energy above 3 kHz has higher sharpness. Computed from the specific loudness distribution with a weighting function emphasizing high critical bands. |
| Specific Loudness | Loudness per critical band (Bark). The distribution of loudness across the auditory frequency scale. Plotted as sone/Bark vs. critical band rate (Bark). Integration of specific loudness over all Bark bands yields the total loudness in sone. |
| Spectrogram | A time-frequency representation showing how the spectral content of a signal changes over time. Displayed as a color map with time on the horizontal axis, frequency on the vertical axis, and color representing amplitude. Computed from successive FFT blocks (STFT). |
| TNR | Tone-to-Noise Ratio. A metric for quantifying the prominence of tonal components in a noise signal, per ECMA-74 / ISO 7779. Compares the power of a tonal peak to the surrounding noise floor within a critical band. Higher TNR indicates a more prominent (potentially annoying) tone. |
| Tonality | A psychoacoustic attribute describing the degree to which a sound is perceived as tonal vs. noise-like. Computed per the Aures/Terhardt model or the Hearing Model approach. Measured in tu (tonality units). Strongly correlated with annoyance perception. |
| Wavelet Transform | A time-frequency analysis method that uses scaled and shifted versions of a mother wavelet to decompose a signal. The Continuous Wavelet Transform (CWT) provides adaptive resolution: fine frequency resolution at low frequencies and fine time resolution at high frequencies. |
| Windowing | The application of a tapering function to a finite block of time-domain data before FFT computation. Reduces spectral leakage caused by the discontinuity at block boundaries. Common windows: Hanning, Hamming, Flat Top, Kaiser-Bessel, Rectangular. |
© 2026 V-Listen. All Rights Reserved.