January 24, 2022

Bird songs, calls and other bird sounds

In some habitats, to observe directly a bird may be extremly difficult; the identification of a bird relies then on identifing its call.
There are different methods to identify a bird through its song, call, noise, etc:

1) training the ear by listening to the calls and songs of birds properly identified and recorded in audio material (Cd, tape, etc.) that you can reproduce as often as you wish, and then compare the learned and memorized song and sound with that that you hear in the field in situ.
You may also record the sound, call or song during your field trip, reproduce it later at home and then compare it with the audio of the referenced songs.
For comparison, choose the audios from species that you most likely will encounter in the habitat you are going to visit. In this way, you narrow the spectrum of songs you have to learn. It is thus important, you gather first information on what type of habitat are in the region you are going to visit and what species can be found in such habitats. Open sources audios of birds can be found in: https://xeno-canto.org/
This plattform allows you to download audios or hear them online. You may also download the sound frequency spectra of the audios.

2) analysing and comparing the sound frequency spectrum of the call, song, sound of the bird you have heard with the sound frequency spectrum of audios recorded from properly identified species.
The pattern of the sound frequency spectrum is species specific. Although some birds also may include in their species specific songs some elements from other species...
This method requires additonal equippment apart from your "ear" and "brain" and the tape recorder, CD-player or whatever apparatus you may use to reproduce audios; at least a device with which you can record the bird song in situ is necessary (your smartphone, a small recorder...) and a software that converts your audio into a sound frequency spectrum.

Open source software or applications that converts your audio into a frequency spectrum which you can then print or save and compare with referenced frequency spectrums are, for example:

2a)Raven Lite (https://ravensoundsoftware.com/software/raven-lite/).
This software was conceived by the laboratory of Ornithology from the Cornell University specially for animal sounds. There are also more professional versions from Raven whose licence you may acquire through payment. In Raven Lite, there is an upper frequency upon which the programm won´t be able to analyse and represent graphically the sound frequency spectrum. This upper limit, however, is above the spectrum of bird song. Thus, you may be able to represent graphically the sound frequency spectrum of all the birds you may record.

2b) The software developed by Audacity (www.audacityteam.org).
You download the softfare to your computer and run it. Besides of analysing and representing graphically the frequency spectrum, this software allows you also to cut pieces out of your audio, reduce noise and much more.

3) using a program of automatic bird song recognition online through the application of BirdNET (https://birnet.cornell.edu/api/)
You upload online your audio; the application analyses then the audio and proposes you with an likelyhood estimation the identity of the bird species. The more "clean" the sound is, the less background noise there is and the louder the sound of the specimen of interest is, the higher is the likelyhood of the correctness of the proposed bird identity.

4) identifying the bird call through sharing your audio with users of citizen science plattforms. This is a friendly and easy way to identify the bird whose call, sound, song etc. you have recorded. You upload your audio in such a plattform, share it with all the community and allow birders, trained users or experts to identify the bird for you ;-). Appart from using for this purpose www.inaturalist.org you may also upload your audio in https://xeno-canto.org/.

Recognizing birds through their song is fascinating and opens you the pandora box in at a first glance an apparently empty landscape...Viel Spaß! :-)

Posted on January 24, 2022 11:09 AM by murzillo murzillo | 2 observations | 2 comments | Leave a comment