Home

Calculations
Contacts
Conversions
Help
Links
Subjects

Signal Processing

The autocorrelation function gives a measure of the extent to which a signal correlates with a displaced version of itself, as a function of the displacement.
Bandpass Filter
Bandstop Filter
Bandwidth-Time Product: The bandwidth-time product is used in calculating the confidence limit of a measurement. Confidence limits describe the uncertainty in measuring the level of random signals in finite periods of time.
Bartlett Window
Bessel Filter
Cepstrum: A cepstrum is a time history defined as the power spectrum of the logarithm of the power spectrum. This you can interpret as the spectrum of a spectrum. It extracts periodicities in the spectrum, such as harmonic families and sidebands. The imaginary part is the Hilbert transform.
Chirp: A chirp is a sound that has a rapid increase in frequency.
Coherence: A two channel function which is used when an output (e.g. displacement) is caused by more than one input (e.g. two unrelated forces at different positions). The coherence function shows how much of the output is caused by each of the inputs, as a function of frequency.
Colourmap: This is a method of plotting noise or vibration versus engine speed and frequency simultaneously. Normally frequency is plotted along the x-axis, engine speed along the y-axis and the level is shown by the colour (the rpm and frequency axes can be plotted the other way).
Correlation: If two variables are in some way dependent on each other or related to each other, they are said to be correlated.
Covariance: Covariance functions define the degree of alikeness between two time histories as a function of time shift (cross covariance) or the degree to which a function is correlated with itself as a function of time delay (autocovariance).
Crest Factor

Peak value divided by the R.M.S. value, used as a measure of the severity of a segment of data.

The cross-correlation function gives a measure of the extent to which two signals correlate with each other as a function of the time displacement between them.

Cross Spectrum

The cross spectrum is the forward Fourier transform of the Cross Correlation function.

DFT

The digital (sampled) version of the fourier transform. Any block size can be used. This technique involves long computations, and normally the Fast Fourier Transform is used instead due to its quicker computation.

Down Sample

Elliptic Filter

FFT

A method developed in 1965 which allows a Fourier Transform to be computed very quickly. This method relies on the number of data points in the time series being a power of 2 (e.g. 512, 1024, 2048). Most modern frequency analysers now use this process.

Filter

A device which removes or attenuates certain frequencies present in a voltage signal. Low pass filters cut out high frequencies, high pass filters cut out low frequencies and bandpass filters cut out frequencies above and below a selected frequency band.

Frequency Domain

Hamming Window

An amplitude weighting of the time signal used with gated continuous signals to give them a slow onset and cut-off in order to reduce the generation of side lobes in their frequency spectrum.

Hanning Window

An amplitude weighting of the time signal used with gated continuous signals to give them a slow onset and cut-off in order to reduce the generation of side lobes in their frequency spectrum.

Spectrogram

Time History

A continuous record of the variation of a physical quantity (e.g. displacement, acceleration, force, etc.) with time. The record may be stored in a chart recorder, magnetic tape, computer disc or any other storage medium.

Up Sample

Waterfall

A method of plotting engine vibration or noise data in three-dimensional form.

Wavelet

The wavelet allows a rapidly changing time history (e.g. door slam noise) to be investigated in the frequency domain.

Windowing

An amplitude weighting of the time signal used with gated continuous signals to give them a slow onset and cut-off in order to reduce the generation of side lobes in their frequency spectrum.