Graph embedding: Difference between revisions
en>Helpful Pixie Bot m ISBNs (Build KC) |
No edit summary |
||
| Line 1: | Line 1: | ||
In [[signal processing]], the '''high frequency content measure''' is a simple measure, taken across a signal [[spectrum]] (usually a [[STFT]] spectrum), that can be used to characterize the amount of high-frequency content in the signal. The magnitudes of the spectral bins are added together, but multiplying each magnitude by the bin "position" (proportional to the frequency). Thus if ''X''(''k'') is a discrete spectrum with ''N'' unique points, its high frequency content measure is: | |||
: <math> | |||
\mathrm{HFC} = \sum_{i=0}^{N-1} i|X(i)| | |||
</math> | |||
In contrast to perceptual measures, this is not based on any evidence about its relevance to [[human hearing]]. Despite that, it can be useful for some applications, such as [[onset detection]]. | |||
The measure has close similarities to the [[spectral centroid]] measure, being essentially the same calculation but without normalization according to overall magnitude. | |||
==References== | |||
* P. Brossier, J. P. Bello and M. D. Plumbley. ''[http://aubio.piem.org/articles/brossier04realtimesegmentation.pdf Real-time temporal segmentation of note objects in music signals]'', in Proceedings of the International Computer Music Conference (ICMC 2004), Miami, Florida, USA, November 1–6, 2004. | |||
* Masri, P. (1996). ''Computer modeling of Sound for Transformation and Synthesis of Musical Signal.'' PhD dissertation, University of Bristol. | |||
[[Category:Digital signal processing]] | |||
{{Signal-processing-stub}} | |||
Revision as of 13:17, 30 November 2013
In signal processing, the high frequency content measure is a simple measure, taken across a signal spectrum (usually a STFT spectrum), that can be used to characterize the amount of high-frequency content in the signal. The magnitudes of the spectral bins are added together, but multiplying each magnitude by the bin "position" (proportional to the frequency). Thus if X(k) is a discrete spectrum with N unique points, its high frequency content measure is:
In contrast to perceptual measures, this is not based on any evidence about its relevance to human hearing. Despite that, it can be useful for some applications, such as onset detection.
The measure has close similarities to the spectral centroid measure, being essentially the same calculation but without normalization according to overall magnitude.
References
- P. Brossier, J. P. Bello and M. D. Plumbley. Real-time temporal segmentation of note objects in music signals, in Proceedings of the International Computer Music Conference (ICMC 2004), Miami, Florida, USA, November 1–6, 2004.
- Masri, P. (1996). Computer modeling of Sound for Transformation and Synthesis of Musical Signal. PhD dissertation, University of Bristol.