Spectral Analysis Option
With FlexPro's Spectral Analysis Option, you enter a new world of software
engineering. You'll save precious time by eliminating the programming
or multi-step UI procedures that are normally required for performing
sophisticated spectral analysis. By using real-time 2D and 3D spectral
graphs, the new FlexPro's Analysis Wizard offers immediate visual feedback
when changing algorithms, algorithm parameters, and spectral formats.
Quickly
locate your signal components
FlexPro's Spectral Analysis Option gives engineers and researchers the
power to rapidly find the components of complex signals. A rich set of
spectral analysis procedures helps you make intelligent signal content
conclusions for any application. The built-in spectral analysis procedures
include: FFT, AR, ARMA, Eigenanalysis, Continuous Wavelets, Cross-Spectra,
Coherence, and Transfer Function Estimation.
Identify frequency and power
with Fourier analysis
Get a complete picture of the frequency signature of a signal using
up to five different Fourier spectrum methods. Solve the leakage problem
found with a standard FFT by using one of the thirty built-in data-tapering
windows. The latest innovations in algorithms, adaptive spectra, and
peak determination help you to better characterize the frequency and
power of each signal component. You can even manage unevenly spaced data
with Fourier techniques originally developed by astrophysicists.
Effortlessly
analyse non-stationary data
Simultaneously find the time and frequency localization components of
a non-stationary periodic signal with Short-Time Fourier Transform or
Continuous Wavelet Transform methods. For the CWT, the Spectral Analysis
Option gives you a choice of three adjustable wavelets in order to find
the optimum time-frequency resolution tradeoff.
Principal component modeling
The Spectral Analysis Option offers state of the art methods for isolating
the spectra of the principal components within a signal. These methods
remove of the influence of noise in the AR SVD, ARMA SVD, and Eigen-decomposition
procedures, enabling you to optimize the estimation of narrowband components.
Harmonic
analysis
Advanced parametric sinusoidal modeling is offered with your choice
of frequency estimation methods. The number of harmonics or spectral
peaks can be set directly by count or indirectly by spectral threshold.
Cepstral
Analysis
The Cepstrum and its minimum-phase reconstruction can be used to de-convolve
signals. Its main applications are speech analysis and echo detection.
Features
Fourier Spectral Analysis
- Procedures: windowed Fourier spectrum, periodogram, Fourier multitaper,
spectrum of unevenly sampled data, cepstrum
- Transforms: best exact n
method automatically chosen from four different algorithms (radix2,
prime factor, mixed radix, chirp-Z)
- Spectral formats include: amplitude,
RMS amplitude, amplitude²,
magnitude, magnitude², phase, dB, normalized dB, TISA power, MSA
power, SSA power, variance, complex, real part and imaginary part
- Options
for zero padding and to display white noise critical limits
- Data tapering
windows, 21 fixed width, 9 adjustable width including Kaiser-Bessel,
VanderMaas, Chebyshev, and Slepian DPSS
- Fourier peak detection by
bin interpolation
AR, ARMA and Eigen Spectral Procedures
- Autoregressive (AR) spectral estimators: autocorrelation, maximum
entropy (Burg), least-squares normal equations, least-squares covariance
and modified covariance, SVD principal component AR
- Autoregressive-Moving-Average
(ARMA) spectral estimators, including non-linear optimization and
SVD principal component methods for signal-noise separation
- Eigenanalysis
methods: MUSIC (Multiple Signal Classification), EV (Eigenvector)
- Select
signal and noise sub-spaces for SVD or Eigen-based signal noise thresholding
- Peak
detection by complex roots of AR polynomial or eigenmodes
- Adaptive
spectra using Runge-Kutta algorithm to accurately map sharp spectral
peaks, minimize spectrum length
Time-Frequency Spectral Analysis
- Short-Time Fourier Transform (STFT) spectrum
- Continuous Wavelet Transform
(CWT) spectrum multi-resolution time-frequency techniques
- Wavelet
spectra can be generated with up to 1000 linear or logarithmic frequencies,
range of frequencies can be customized
- Adjustable mother wavelets:
Morlet, Paul, Gaussian Derivative
- Offers capability of ultra high
frequency resolution with very large signals
Harmonic Analysis
- Sinusoid or damped-sinusoid modeling using automatic, Fourier,
AR, Eigen, or Prony algorithms for frequency estimation
- Harmonics table,
THD, SNR, SINAD and de-noised signal
Two-Signal Spectral Analysis
- Fourier windowed cross-spectra and Fourier cross-periodogram
- Coherence,
including SNR spectra
- Fourier domain transfer function
Non-linear methods
- Real cepstrum including "liftering" and minimum-phase reconstruction.
To purchase this FlexPro option, click
here or contact your local Adept Scientific
office. |
RSS Feeds
