It occurred to me to test the audio response of the system I used to analyse the trees - which was to shoot video on an old ipod and play it through the laptop speakers into Audacity. This time I shot video of tones generated in Audacity to see what happened. First is 1000Hz. It clearly peaks there but there is some white noise behind.
Next was 10000Hz. The peak is in the correct place but notice that the decibels are so much lower. The white noise on both seems to be about -80dB.
As an experiment I tried the highest available tone of 22000Hz. I can't hear that. Neither can the microphone. This might look loud but it is all below -73dB so it is just the same white noise. It seems the analysis of the tree might be correct because though quiet, it reached -33dB.
