I have calculated spectral entropy Hf and ADI as well as BI, NDSI, AEI, ACI for one minute acoustic recordings for ~ two weeks (one week in March and one in April) with the default settings of scikit-maad.
I plot their mean hourly values over a day and try to describe the soundscape accordingly. But what when Hf is opposite to ADI? Shouldn't acoustic diversity and spectral entropy agree? At this same point I become high BI and high NDSI, so biophony is prevailing , but what else should I take into consideration? This sample has a strong frequency band 0-1000 Hz, due to wind and sheep bells.