ONSETNSGT - Onset-based nonstationary Gabor transform
Usage
[c,g,shift,M,Ls] = onsetnsgt(f,thre,short,max_win,win_length) [c,g,shift,M,Ls] = onsetnsgt(f,thre,short,max_win) [c,g,shift,M,Ls] = onsetnsgt(f,thre,short) [c,g,shift,M,Ls] = onsetnsgt(f,thre) c = onsetnsgt(...)
Input parameters
f | The signal to be analyzed (single channel only) |
thre | Peak-picking threshold |
short | Shortest allowed window length |
max_win | Maximum number of different windows used |
win_length | Window length for the onset STFT analysis |
Output parameters
c | Cell array of transform coefficients |
g | Cell array of analysis windows |
shift | Vector of time shifts |
M | Number of frequency channels |
Ls | Original signal length |
Description
This is a wrapper function for the scaleframe nonstationary Gabor transforms with onset detection based adaptation. Given a signal f, this wrapper computes the spectral flux onset detection function based on a regular discrete Gabor transform with redundancy 16 using a Hann window of length win_length. A simple peakpicking algorithm determines the significant maxima in the spectral flux function. Those are assumed to be the onsets in f.
From this onset sequence, a scaleframe nonstationary Gabor system will be constructed and the corresponding analysis performed by nsgt_real.
Note: The current wrapper only supports the threshold parameter thre of the onset detection algorithm. To obtain optimal results, the remaining parameters need to be fine tuned as well. An experienced user should use onsetdet, nsgsclwin and nsgt_real on separately instead. Also see the Onset How-To included in the toolbox.
References:
S. Dixon. Onset detection revisited. In Proceedings of the 9th International Conference on Digital Audio Effects, volume 120, pages 133-137, 2006.
P. Balazs, M. Dörfler, F. Jaillet, N. Holighaus, and G. A. Velasco. Theory, implementation and applications of nonstationary Gabor Frames. J. Comput. Appl. Math., 236(6):1481-1496, 2011.