Generate Quadratic Time-Frequency Distributions This function generates various members of quadratic time-frequency distributions. Predefined members can be specified, or a user defined kernel may be used. Usage: tfd = quadtfd( signal, lag_window_length, time_res, kernel [, kernel options] [,fft_length ] ); kernel is one of: 'wvd', 'smoothed', 'specx', 'rm', 'cw', 'bjc', 'zam', 'b' or 'mb'. where tfd The computed time-frequency distribution. size(tfd) will return [a, b], where a is the next largest power of two above fft_length, and b is floor(length(signal)/time_res) - 1. signal Input one dimensional signal to be analysed. An analytic signal is required for this function, however, if signal is real, a default analytic transformer routine will be called from this function before computing tfd. lag_window_length This is the data window length and controls the size of the kernel used for analysis (lag_window_length must be odd). The kernel used will be defined from -(lag_window_length+1)/2 to +(lag_window_length+1)/2 in both time and lag dimensions. time_res The number of time samples to skip between successive slices. kernel The determining kernel function. kernel is a string defining a predefined kernel. Predefined types: 'wvd' Wigner-Ville 'smoothed' Smoothed Wigner-Ville 'specx' Short-Time Fourier Transform 'rm' Rihaczek-Margenau-Hill 'cw' Choi-Williams 'bjc' Born-Jordan 'zam' Zhao-Atlas-Marks 'b' B-distribution 'mb' Modified B-distribution 'emb' Extended Modified B-distribution kernel_options Some kernels require extra parameters: 'smoothed' kernel_option = length of smoothing window kernel_option2 = type of smoothing window, one of: 'rect','hann','hamm','bart' 'specx' kernel_option = length of smoothing window kernel_option2 = type of smoothing window, one of: 'rect','hann','hamm','bart' 'cw' kernel_option = Smoothing parameter 'zam' kernel_option = ZAM 'a' parameter 'b' kernel_option = B-distribution smoothing parameter beta 'mb' kernel_option = Modified B-distribution parameter alpha 'emb' kernel_option = Extended Modified B-distribution parameters alpha, beta 'ckd' kernel_option = Extended Modified B-distribution parameters alpha, beta fft_length Zero-padding at the FFT stage of the analysis may be specified by giving an fft_length larger than lag_window_length. If fft_length is not specified, or is smaller than the lag_window_length, then the next highest power of two above lag_window_length is used. If fft_length is not a power of two, the next highest power of two is used. See Also: analyt