Making Songbirding More Accessible

After many years of leading birding outings with a wide range of folks, it didn’t take long for me to meet birders who had trouble hearing certain birds’ songs. I found even more birders who struggled when I worked with volunteers on the Hamilton Bioacoustics Research Project.

For some, hearing loss isn’t just all sounds getting quieter, it’s the higher-pitched, high frequency sounds getting quieter and going silent. Conversation with other people might be no trouble hearing, but a Cedar Waxwing is inaudible.

Podcasts do not usually need to account for this kind of hearing loss: voices will be fine, most music will be fine, perhaps a few soundscapes or sound effects may be affected. The core content will not usually be a challenge for someone with high-frequency hearing loss.

For a podcast focused on birdsong, however, this can be difficult if not impossible to hear. Many birds’ songs reside in the upper frequencies. Many warblers, for instance, have songs ranging between 6 and 9 KHz, and some even higher.

To account for this, “Songbirding Under 5KHz” is a modified version of Songbirding that uses the following process to convert a standard episode into a modified version that takes all frequencies above 5KHz and duplicates them at an octave lower – to one half the frequency.

Steps

The following are steps used in Audacity version 2.3.3, though other versions should be similar. The same process can likely be simulated in other software, so long as the software has something equivalent to Audacity’s “high quality stretching” for pitch changes.

  1. Open the sound file you wish to modify. It should be mixed down already to a single track mp3 file or similar.
  2. Duplicate this track with Edit -> Duplicate.
  3. Select the duplicate track, use Effect -> High Pass Filter… and set the frequency to 5000 at 48 dB. This will remove frequencies below 5,000 on this track.
  4. Next, use Effect -> Change Pitch… and in the Pitch dialog, set -12.0 Semitones (half-steps), or set Percent Change to -50.00 (both are the same). Ensure “Use high-quality stretching” is checked, this is important. When you click “OK” this may take a while depending on the audio length. Typically it will take about a third of the length of the audio in processing time: 10 minutes for every 30 minutes of audio, but this will vary depending on the power of your computer.
  5. Select both the original and the duplicate tracks. Go to Tracks -> Mix and Render.
  6. (Optional) You may want to tack on some introductory narration that explains its been modified, if there’s a chance someone might come across the file without the unmodified original.

The result

Here’s an example episode.