During TEOSS workshops we will be aiming at collecting Orthopteran sound recordings of well-identified species with a minimum set of metadata. Our aim is to train recorders for making possible automatic species recognition based on sound recordings for a larger part of the European Orthoptera fauna.
PART I. INTRODUCTION. GETTING SOME CONCEPTS.
Recording devices and microphones
During the workshop we will be recording with the following types of devices:
● Good digital recorder with or without external microphone
● Cheap digital recorder using internal microphone
● Phone with ultrasonic microphone
● Phone with non-ultrasonic microphone
● Phone with internal microphone
In combination with using different sample rates, this will yield a great variation in quality of sound recordings. We aim to let every participant use every device, to be able to experience the difference in (technical) handling, ease of use and recording quality.
Minimum recording quality and size
We only will be recording digitally, with a sampling frequency at or above 44.1 kHz (up to 354 kHZ) and bit depth of 16 bits (or maybe 24 bits). We will save our files uncompressed, meaning in WAV-format, so not compressed in e.g. mp3, m4a, aac, flac, or ogg. We must be aware that some analysis software only accepts files with 16-bit depth and that online platforms accept also higher bit depth files (table 1).
Table 1: Handling of sound recordings on different web platforms.
High quality (stereo) recordings easily become very large. Online platforms accept only small thus short files (except Xeno-canto that accepts 128Mb, see table 2). This means that high-quality sound recordings may last only a few seconds, which may be too short to record the full courtship song of a Gomphocerine grasshopper, for example. This is one of the reasons we decided to go for uploading sound recordings to Xeno-canto (XC) during TEOSS.
Table 2: Maximum duration of sound recordings of different quality, based on maximum file sizes of 10, 20 and 128Mb.
Metadata
We need to add a minimum set of metadata to the sound recordings. This is important for several reasons:
● When including locality date and coordinates, the sound recording will be shared through GBIF, making the recording available as an observation for research and conservation.
● Including information on the recording device and quality may help to understand the representation of the song in the recording.
● Including temperature and time of day during recording is important to know under what conditions species sing and how song characters may depend on temperature.
PART II. WORKFLOW IN THE FIELD
Participants of the workshop will be instructed beforehand about this workflow and how to use the recorders and microphones available. Also, field forms to note metadata and tubes to temporarily collect specimens will be handed out. The workflow in the field has the following steps.
Reaching the location
● Travel to field location. Note Metadata: COUNTRY / PROVINCE / MUNICIPALITY / LOCALITY.
● Hand out recorders Note Metadata: RECORDER/MICROPHONE.
● Attend instructions before field work (e.g. on the time to be back or species to expect or search for).
● Synchronize date/time between Digital Recorder / Smartphone / Camera. Note Metadata: DATE / TIME.
● Basic approach: One sound = one record = one line of data in the paper field form.
● In summary, fieldwork involves 5 successive steps : 1) Recording sounds and 2) taking Pictures and/or 3) collecting the specimen and 4) filling Observation datalog (aka “OBSERVATION”) (using the app Obsmapp or iObs (Observation.org)) and 5) filling the paper form (aka “FORM”).
● Silently participants start to approach carefully singing specimens and start collecting records and note down all relevant metadata. (Note: some metadata may be extracted in the lab from observation.org). Note Metadata: TEMPERATURE and maybe a note if the specimen is in a sunny site or in the shade).
Fieldwork sheet -ready to download-
Sound recording procedure
● Step 1: Record one sound sequence (two situations: various sound files with short recordings or only one long sound file).
o Select a proper recording and note down the name of this recording. Note Metadata: RECORDING NAME.
o Test various recording setups (close distance with handheld recorder vs long distance with use of stick ; check microphone and recorder options) which generate different records.
o Note the sound type (calling song, rivalry song courtship song, linked to behavior, Note Metadata: SONG TYPE.
● Step 2: Compulsory, take pictures of the specimen as a way of identifying the species or keeping a voucher of the specimen:
o With the smartphone (consider to allow the geolocation of images in the smartphone setups);
o With a dedicated camera (very often, geolocation is not available, look carefully at correct Date and Time setups, especially if you traveled from a different time zones).
o You could also take a blank picture (e.g. sky) between two consecutive specimens, especially if they belong to the same species, in order to ease lab work.
● Step 3: Eventually, capture the specimen with a net for close-up pictures or for collecting purposes. Collected specimens should be labeled in the field with the date, locality, name of collector (traditional entomology). Add time of recollection and any detail for ease of future referencing in the lab.
● Step 4: Use the phone’s APP (Obsmapp/iobs) to record the observation in the field. If the APP is not used in the field, you will have the possibility to record the observation in the website, back in the lab.
● Step 5: COMPULSORY. Fill the paper form with the sound file name and relevant information that allow you to fill in the lab the projects’ database.
PART III. BACK IN THE LAB
In the lab or hotel room/lobby, the following activities take place, preferably shortly after return from the field:
Sound work
o Copying the selected sound recordings to a local computer (instructions follow elsewhere). Store files of every day in a separate folder to be sure nothing gets mixed up.
o Shortening of longer sound recordings in Audacity (preferably sound recordings start with the onset of the song that has been recorded and are not larger than 128Mb). No other edits are needed.
o If necessary adjust the name of the file (e.g. when two species are in a recording after each other and you want to upload both as separate recordings or if filenames of different recorders may turn out to be the same).
Identification work
o Further identification of unidentified specimens (using photos, collected specimens or analyzing the sound recording).
o Identification will be updated in the corresponding FORM (the paper form you used in the field) and OBSERVATIONS (online on the web server).
Citizen Science web server work (Observation.org)
o Check that observations have been uploaded.
o Add pictures taken with a camera, sometimes after cropping them.
o Update the identification of the observation if necessary.
o Update the data (stage, sex, etc.)
o copy the XC number into the remarks field of the observation.
o Copy the URL of the observation into the remarks field of the Xeno-canto sound recording metadata.
o Don’t forget to add “TEOSS1” in the remarks field in XC!
This project has received funding from the European Union’s Horizon Europe Research and Innovation programme within the framework of the TETTRIs Project funded under Grant Agreement Nr 101081903.
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or REA. Neither the European Union nor REA can be held responsible for them.