![whistle phone calls whistle phone calls](https://www.sevenware.com/Group/Group621/Catalog/H16909.jpg)
2.23 Harbour porpoise events were sub-divided as: a) single click train (HPS), b) multiple click train (HPM), c) close click train (HPC), d) no clear track (HPN), e) possible porpoise click event (HPX). Classification was based on the frequency spectrum and waveform characteristics of individual clicks and patterns of successive bearings within trains (sequential patterns of clicks) ( Figure 2.3). A click train events were reviewed offline and classified as follows: a) harbour porpoise (HP) b) delphinid (D) c) ship noise during changes of course (T) or d) sonar or noise from other vessels (S).
#WHISTLE PHONE CALLS SOFTWARE#
These data were used by the software Rainbow Click (IFAW) to classify harbour porpoise-like and other clicks. 2.22 The HF system digitised the hydrophone signals from H1 and H2 at a sampling rate of 500 kHz. A second detection rate was thereby calculated that was representative of the actual number of dolphin calls recorded: dolphin calls per km (Dolphin Detection Rate II). Data for these selections were imported to the Access database and linked to the GPS data. Relatively faint whistle call contours were often evident on the spectrogram, which were the threshold parameters necessary for automatic identification in Pamguard. To do this we searched intervals of the spectrogram spanning from 60 s before to 60 s after time periods in which that connected regions had been identified in Pamguard.
![whistle phone calls whistle phone calls](https://dqzmfj.bilskanieruchomosci.pl/templates/f16dc396e088c7c707eabe9d7479e7a2/img/ce0eec3d60708e7db292151db1d01ac7.jpg)
We used Raven to manually extract entire whistles, pulsed calls and overlapping sequences of whistles from the spectrogram display ( Figure 2.2). Typically, many connected regions were identified within a single loud whistle sequence. 2.21 Connected regions represented segments of dolphin calls. These data were used to calculate a standardised dolphin encounter rate: verified Pamguard connected regions per km (Dolphin Detection Rate I). The detection was tagged as such in the database table. Using the start times in the database as a guide, each candidate detection was located on the spectrogram and it was verified whether the connect region was indeed part of a distinct dolphin whistle or pulsed call, or an artefact of background noise. Using Raven, a spectrogram was generated of the signal from H1 (1024 pt, 50% overlap, Hann window).
#WHISTLE PHONE CALLS PRO#
Each detection was subsequently verified using the sound analysis software Raven Pro 1.4 (Cornell Lab of Ornithology).