Signal Processing Toolbox Release Notes    

Chapter 3
Signal Processing Toolbox 5.0 Release Notes

New Features

This section introduces the new features and enhancements added in the Signal Processing Toolbox 5.0 since the Signal Processing Toolbox 4.3 (Release 11.1).

The Signal Processing Toolbox 5.0 provides a number of important enhancements, and a completely new facility for designing and analyzing filters, the Filter Design & Analysis Tool (FDATool).

This section is organized into the following subsections:

Filter Design & Analysis Tool (FDATool)

The Filter Design & Analysis Tool provides a simple and intuitive graphical interface to most of the Signal Processing Toolbox's digital filter design functions.

To start the tool, type fdatool at the MATLAB command line.


SPTool has received a number of enhancements:

To start the tool, type sptool at the MATLAB command line.

Spectrogram Demo

The Spectrogram Demo (specgramdemo(y,Fs)) is a new demo that displays a spectrogram, a time plot, and a frequency slice of an input signal, y, assuming a sample rate of Fs Hz.

Automatic Order Adjustment in FIR Filter Functions

Most of the FIR filter functions (fir1, fir2, firls, remez, fircls, fircls1) now automatically increase the order of linear phase filters by 1 when necessary to meet the design specifications. For example, a Type II (odd order, symmetric) linear phase filter must have a zero at radians. If a gain other than zero is instead specified at radians in any of the above mentioned functions, the function adds 1 to the filter order and proceeds to design the higher order filter.

A warning message is produced in the command window whenever a function automatically changes the filter order.


The grpdelay function now returns more accurate results for IIR filters.


The hilbert function no longer zero-pads sequences that have a non-power-of-two length, so the transforms are now exact.


The lpc function can now return an estimate of the prediction error variance (power) as a second output.


The xcorr function can now efficiently compute the cross-correlation and autocorrelation of large vectors.

  Known Software Problem Major Bug Fixes