Can anyone give brief pros and cons for using duration-based expansion Vs event-based expansion for frequency triggered recordings? I'm wishing to split my triggered recordings into 5 sec chunks for analysis. I'm recording with a 'Recording Duration' of 55secs and a 'Minimum Trigger Duration' of 2secs with frequency triggering and trying to understand how these settings interplay with the 'Maximum File Length' setting for Event-based expansion. Thanks
top of page
To test this feature, visit your live site.
1 Comment
bottom of page
It depends how you intend to process them really. If you do Duration-Based expansion you will get a single file that expands to the original recording period. You can then also impose a maximum file length, of say 5 seconds, so that you get 12 files per minute, and then you can also opt to not generate the ones that are silent with no recorded content. All of these files will start on a round second. If you do Event-Based expansion you get a separate file, timestamped to the closest millisecond, for each detection event in the original recording. You can then impose an additional file length limit, forcing long periods to be split up, and then you can also impose that files start and end on round seconds. We normally just expand the whole file using the default Duration-Based settings, or generate separate files for each individual events using the default Event-Based settings. The other settings are there to manage the constraints of different processing chains that often imposed file length limits if they are being used to train detection models.