Grunts and you can deep grunts both integrate repetitive points. Because these repeated factors differed considerably towards a couple grunt models, we entitled them in a different way: ‘pulses’ to own grunts, and you will ‘voice cycles’ for strong grunts. I used the program PRAAT 5.4.01 () to your voice analyses.
We chose large-high quality grunts and deep grunts of the only and those in the brand new investigation out of sound services, which had a signal-to-sounds proportion regarding dos or even more on the about three pulses/voice cycles towards higher amplitude. To accomplish this, we opposed the fresh new voice pressure of your heartbeat/period to the 3rd highest amplitude towards the sound pressure out of around three at random chose situations regarding the background looks within 0.5 s up until the grunt otherwise deep grunt. Should your sound stress of that pulse/years is at least two times as higher as records appears, we analysed the brand new attributes of the grunt otherwise strong grunt. On analysis of your own functions of your own grunt brands, i noticed four parameters: step 1. level of pulses/cycles for every single sound, 2. duration of this new sound, step three. quantity of pulses/time periods for each and every second, 4. principal frequency.
In order to quantify what number of pulses/schedules for every sound, i noted all evident pulse/course about wave form each and every grunt within zero crossing following highest height on the pulse/years and you may counted the latest designated zero crossings. To determine the lifetime of a sound, we mentioned the full time amongst the noted zero crossings of the basic and last discernible heartbeat/cylcle. So you can calculate what amount of pulses/schedules for each and every next, we split what amount of pulses/cycles by lifetime of the fresh voice. To search for the prominent volume, i investigated the three loudest pulses within a sound toward volume for the higher voice pressure and you will grabbed the average of these three frequencies.
With the study away from sound properties getting presses and you may plops, we only used sounds by which we are able to demonstrably identify the new sound-generating fish. I revealed clicks and you will plops using one or two details: 1. Dominating frequency, dos. sound pressure level difference in down and better frequencies.
To determine the dominant frequency of your voice, we investigated the power spectral range of the latest click otherwise plop to possess this new regularity towards high sound pressure level. We derived the power spectrum throughout the zero crossing of one’s waveform involving the large and you may lowest amplitude. In order to estimate the http://www.datingranking.net/nl/mytranssexualdate-overzicht/ new sound force differences, we substracted the fresh voice strain of fifth harmonic off brand new voice strain of dominant regularity.
Assessment out of voice services
On comparisons away from sound functions, i basic averaged the knowledge for men songs to the personal height. We were incapable of do this for ladies, as there are no way of a couple of times distinguishing individual lady within the the fresh video clips reliably.
To have ticks and you will plops, i first tested to have gender-certain distinctions of one’s analysed services
We compared the newest principal frequency and you may years anywhere between men grunts and strong grunts to determine differences between both phone call systems. I up coming checked to own differences between the brand new both sorts of unmarried-heartbeat musical.
For statistical analyses, we first investigated the properties of the tested sounds for normality using Shapiro-Wilk tests. If data were normally distributed according to Shapiro–Wilk test (P > 0.05), we used t-tests to examine the differences in sound properties. If the Shapiro–Wilk test showed a significant deviation from a normal distribution (P < 0.05), we log-transformed the data to achieve normality, or used Mann–Whitney U tests where a normal distribution could not be achieved by data transformation. For the statistical analysis of sounds we used R (Version 3.3.1, We assumed a difference between sound properties to be significant if the P-value of the respective test was < 0.05.