The zoo soundscape has a number of important implications for animal welfare, management, and conservation. However, despite its importance, the zoo soundscape is yet to be examined in depth using multiple measures. Consistent human presence can influence the zoo soundscape. However, it is difficult to determine the specific impact of human presence, as visitors are usually present during the day when animals are active. The COVID-19 lockdown in 2020 provided a unique opportunity to study zoo soundscapes in the absence of visitors. We compared the sound environment during the 2020 closure period to a comparable open period in 2019 across three zoo aviaries, examining broad band frequency measures of sound pressure levels, sound pressure levels in defined frequency bands, and acoustic indices (Acoustic Complexity Index and Normalized Difference Soundscape Index) to describe the zoo soundscape. Acoustic indices have not, to our knowledge, previously been used in the zoo setting, although they may provide a useful metric to assess sound disturbance. Therefore, we also used this natural experiment to explore how successful these measures may be in assessing disturbances in captive environments. We found a significant effect of human presence on the sound environment; aviaries were generally quieter with less low frequency noise and with a greater proportion of biotic sound during the 2020 zoo closure period. We argue that NDSI could be a useful index for determining anthropogenic disturbance in zoos, although further information on how it is influenced by additional factors, such as human speech, would be beneficial. The use of multiple measures to assess the sound environment in zoos can provide additional information beyond 'loudness', such as frequencies where sound energy is concentrated and characteristics of the soundscape, which could be used to better target management and mitigation.