2017 ESA Annual Meeting (August 6 -- 11)

COS 98-3 - Building a better recording studio: Effects of sound absorption in songbird laboratory environments on song quality perception

Wednesday, August 9, 2017: 2:10 PM
E147-148, Oregon Convention Center
Kate T. Snyder1, Christopher L. Barnobi2 and Nicole Creanza1, (1)Biological Sciences, Vanderbilt University, Nashville, TN, (2)Environmental Acoustics, Dudek
Background/Question/Methods

For a study on song learning capabilities in zebra finches (Taeniopygia guttata), housing must be developed that limits the effects of an unnatural environment on song transmission. The study requires that juvenile songbirds be exposed to a single adult male's song (tutored) for 60 days of song learning susceptibility, using accuracy of song reproduction calculated using Sound Analysis Pro software as a metric of learning capability. Juveniles are capable of very high accuracy in song reproduction, so even very minute unintentional differences in song may be mistakenly learned. Female songbirds are also able to detect very small differences between songs and demonstrate a preference for one song over the other in a choice test. During the tutoring phase, the adult male and juvenile male will be co-housed in a recording chamber and all songs will be recorded. Typically, these chambers consist of a hard plastic container lined with sound-absorbing fabric or foam. This study seeks to 1) analyze the effectiveness of certain sound absorption treatments to prevent reverberation in the chamber and 2) evaluate a female's ability to distinguish between songs recorded in each of the environments via female choice tests.


Results/Conclusions

Sound reverberation in the range of frequencies common in zebra finch song is reduced in foam-lined chambers compared to bare chambers. This results in similarity scores that vary by several percent (calculated by Sound Analysis Pro) for the same song recorded in an open (minimal reverberation) space versus in a bare chamber. This is a significant difference when compared to a typical variance of 5-10% in song similarity scores across live-tutored juveniles. Thus, juveniles may learn and reproduce songs inaccurately if the recording environment is not properly sound treated, and the potential variance is on a scale that is biologically relevant. When adult females are presented with a choice test between two dummy male finches each playing a recording of one of the songs, they have a slight preference against the song recorded in the bare chamber vs the songs recorded in open space and sound-attenuated chambers. These results highlight the importance of using sound absorption treatments in enclosed spaces when recording audio to be finely analyzed or to be used in behavioral tests.