1.3 Audio Recordings

Here are some events found on recordings in language documentation corpora:

  • Singing: Songs, typically folk songs, sung in the target language.
  • Storytelling: Stories, both traditional and anecdotal.
  • Conversations: Dialogue between two or more speakers.
  • Speeches: More formal speech, prepared for public speaking.
  • Elicitation: Answers to prompts made by the documenter, such as translations for specific sentences or reciting word lists.

Equipment for High-Quality Recording

While we cannot always reach our goals, we attempt to make recordings that are optimal for analysis, revitalization, and preservation.  Take a look at the following guides and think about recommendations for best practices for video and audio recording for language documentation, paying attention to:

  • portability and durability of recording devices
  • recording formats 
  • battery life and reliability
  • reducing background noise like loud insect sounds, overhead fans, or strong wind
  • microphones suitable for varied situations (one speaker, conversations, large hall performances)
  • external versus internal microphones
  • ensuring  the stability of the device, and consistency in the quality of the recording by using a tripod with your phone or other devices. 

Guides: 

Vermont Folklife Center:  https://www.vermontfolklifecenter.org/field-recording-in-the-digital-age

Queensland Studies Authority: https://www.qcaa.qld.edu.au/downloads/p_10/snr_atsi_languages_11_documenting.pdf

 

Recording Remotely

It is not always feasible to record in person. In these situations, there are still options for collecting recordings remotely.

Multiple online virtual meeting applications allow for recorded sessions.  One consideration with these applications is that audio could be unclear if users are working with an older laptop or inexpensive computer microphone. Recording virtually is not ideal for research focused on phonetics. Overall, though, recording with a virtual meeting application seems to be working well for field researchers.  

You can read more about this here:  https://www.vermontfolklifecenter.org/remote-recording

There are several options for virtual meeting platforms that allow recording:

Zoom Conferencing  Free options available for virtual recorded meetings

WebEx                      Free options available for virtual recorded meetings

GoToMeeting            No free options, but fairly inexpensive for basic features

There are also other options online that could be used, such as Google Meet, but recording features typically require payment.

Recording on Android phone

A useful app has been designed for the purpose of phone-based language documentation: LIG-Aikuma: A Mobile App to Collect Parallel Speech for Language Documentation

picture of text describing the LIG-AIKUMA application for language documentation

The description of LIG-Aikuma from their website

The app can be downloaded to either an Android phone or a computer. The download is free. Follow these basic instructions from their website:

  • From the Android device
    1. Enter the url address in your web browser.
    2. Confirm the installation.
      • Settings -> Security tab -> Check the “Unknown sources” button

After downloading the app you must grant the app access to your device’s (1) Microphone (for recording speech), (2) Phone (to share recorded files), (3) Location (to position the recording session on the map), and (4) Storage (to store recordings). 

  • Start experimenting with the ‘Elicitation’ option. 
    • You have the option to download wordless picture books and create individual elicitation .WAV files for each image. 
  • Once you are happy with your elicitations you can translate each one after selecting the ‘Translating’ field and entering the relevant metadata.
    •  For these purposes, translating consists of creating a ‘respeaking’ audio file to correspond to the elicitation file. 
    • Both the elicitation and the respoken files are saved in the same folder and are named according to their function,  Date-Time_ Language_trad, and  Date-Time_ Language_rspk, respectively. There are files that provide the alignment of the two. These end in _trad.map and/or _rspk.map. 
  • When your project is ready for export, select the ‘Share’ field, select all desired files, press “Share”, and select the desired destination. 
    • Additionally, you have the option of transferring data to your git repository.

Saving Audio Files

Two of the most common audio file types are.WAV and.MP3.  It is recommended to save original audio files in a .WAV format.  This is because .MP3 files are compressed and information about some sounds is lost in the compression process.  You can always make an .MP3 copy of the original file .WAV file. Why bother with.MP3? Simply because .MP3 files are much easier to stream and download.  

The Value of Audio

Use the power of audio recordings to save for posterity those precious moments in community life that you would like shared with future generations.  Once made available via archiving these sound files will allow language learners to hear the cadence, sentences, and words of authentic interactions.  Once transcribed, translated, and analyzed, these audio files will be the source of much information for language learning and discovery of language structure.

References

Data Collection App for Language Documentation: Lig-Aikuma. (2020). Retrieved July 22, 2020, from lig-aikuma.imag.fr