Abstract
The temporal structure of animals’ acoustic signals can inform about context, urgency, species, individual identity, or geographical origin. We present three independent ideas to further expand the applicability of rhythm analysis for isochronous, that is, metronome‐like, rhythms. A description of a rhythm or beat needs to include a description of its goodness of fit, meaning how well the rhythm describes a sequence. Existing goodness‐of‐fit values are not comparable between methods and datasets. Furthermore, they are strongly correlated with certain parameters of the described sequence, for example, the number of elements in the sequence. We introduce a new universal goodness‐of‐fit value, ugof, comparable across methods and datasets, which illustrates how well a certain beat frequency in Hz describes the temporal structure of a sequence of elements. We then describe two additional approaches to adapt already existing methods to analyze the rhythm of acoustic sequences of animals. The new additions, a slightly modified way to use the already established Fourier analysis and concrete examples on how to use the visualization with recurrence plots, enable the analysis of more variable data, while giving more details than previously proposed measures. New methods are tested on 6 datasets including the very complex flight songs of male skylarks. The ugof is the first goodness‐of‐fit value capable of giving the information per element, instead of only per sequence. Advantages and possible interpretations of the new approaches are discussed. The new methods enable the analysis of more variable and complex communication signals. They give indications on which levels and structures to analyze and enable to track changes and differences in individuals or populations, for instance, during ontogeny or across regions. Especially, the ugof is not restricted to the analysis of acoustic signals but could for example also be applied on heartbeat measurements. Taken together, the ugof and proposed method additions greatly broaden the scope of rhythm analysis methods.
Highlights
In recent years, the temporal structure or rhythm of animal's acoustic signals has received increasing attention
Three methods have been proposed in the context of bioacoustics for extracting exact beat frequencies in order to describe an isochronous element sequence: (1) Fourier analysis, which decomposes a signal into its sinusoidal components (Burchardt & Knörnschild, 2020; Saar & Mitra, 2008); (2) generate-and-test approach (GAT), where a series of acoustic signals are overlaid with an artificial beat to test which artificial beat frequencies resemble the series best (Norton & Scharff, 2016; Ravignani & Norton, 2017); and (3) interonset interval analysis (IOI), which allows the calculation of beat frequencies by averaging IOIs and transforming this rate into a frequency (Burchardt & Knörnschild, 2020)
Universal goodness-of-fit value that can be applied to any possible description of a temporal structure relying on frequencies; we term it ugof for “universal goodness-of-fit value.”
Summary
The temporal structure or rhythm of animal's acoustic signals has received increasing attention. The first issue, which is independent of the complexity of the structure, is the limitations with which so-called goodness-of-fit values quantifying how well a certain beat frequency describes an element sequence can be compared between species as well as between methods. These values exist for all three above-mentioned methods to extract exact, best-fitting beats (Burchardt & Knörnschild, 2020), but they inflict three problems: (a) They show complex correlations to, among other parameters, the number of elements in a sequence; (b) values differ depending on the method used, which precludes any comparison between studies using different methods; and (c) only one value can be obtained for the whole sequence that is being analyzed, without any information at the element level. The reporting on the ugof is emphasized throughout the manuscript, as it is a true innovation for the field, while the two other methods are additions to already established analyses
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
