We are somewhat luddite and view automatic ID with a healthy degree of scepticism.
When Jude sent her Echometer to be updated for the latest version of the ipad software, it came back with their automatic ID installed for free.
We generally use the Bat box Griffin for stationary recording, so we thought we would compare results.
We usually use Prof John Altrinhgam’s Team’s elegant free software BatClassify developed by Chris Scott . It only works on time expansion calls (which is not designed to run on the Griffin or Echometer Touch) and can be downloaded for free here
This is very good at picking out pips and we focus only anything which is not a pip and analyse the file the lD fashioned way (using Adobe Audition)
So on a rainy Good Friday we tried using both
NB Coffee or tea is an essential prerequisite to analysis Photo Jude Hirstwood
Bob has written an excel spreadsheet which colour codes the results from the Altringham software John Altringham himself says you should only accept recordings at above 0.9 as safe IDs, but lower values are also worth looking at.
Bob’s colour coded spreadsheet. Photo Jude Hirstwood
The EM Touch only identifies the loudest call on each recording and misidentified several calls as Barbastelles when they were just background noise. The BatClassify software is very good at barbs). Neither programme is happy with pip social calls, which they classify as big bats. (a mistake Jude has been known to make on initial inspection). The Batclassify software identifies big bats but does not try to break them down further. The EM touch auto-ID does and is quite good at discriminating them.The time tags on both recorders are not recognised by the BatClassify software
Both programmes are useful for prescreeening recordings , but they do need double checking, particularly as we are using detectors which the software was not designed to use.
Our concern is that in order to save time consultants may not do the double checking by eye but it is a very handy tool to have