# spectralCrest

Spectral crest for audio signals and auditory spectrograms

## Syntax

## Description

specifies options using one or more name-value arguments.`crest`

= spectralCrest(`x`

,`f`

,`Name=Value`

)

`[`

returns the spectral peak and spectral mean. You can specify an input combination from any
of the previous syntaxes.`crest`

,`spectralPeak`

,`spectralMean`

] = spectralCrest(___)

`spectralCrest(___)`

with no output arguments plots the
spectral crest.

If the input is in the time domain, the spectral crest is plotted against time.

If the input is in the frequency domain, the spectral crest is plotted against frame number.

## Examples

## Input Arguments

## Output Arguments

## Algorithms

The spectral crest is calculated as described in [1]:

$$\text{crest}=\frac{\mathrm{max}\left({s}_{k\in [{b}_{1},{b}_{2}]}\right)}{\frac{1}{{b}_{2}-{b}_{1}}{\displaystyle \sum _{k={b}_{1}}^{{b}_{2}}{s}_{k}}}$$

where

*s*is the spectral value at bin_{k}*k*.*b*_{1}and*b*_{2}are the band edges, in bins, over which to calculate the spectral crest.

## References

[1] Peeters, G. "A Large Set of Audio Features for Sound Description (Similarity and Classification) in the CUIDADO Project." Technical Report; IRCAM: Paris, France, 2004.

## Extended Capabilities

## Version History

**Introduced in R2019a**