I upped all of the IVS sensitivity setting on the -IL camera to 10 and am reporting back. The side-by-side S3 is, in effect, the equivalent of a control group in an experiment. I also made the IVS global length calibrations. Mt gut feeling is that this didn't affect anything. After a week, there's good news and bad news:
Good news:
1. For detecting objects, the -IL is now almost as good as the S3. It had only a handful of misses, and once or twice detected something the S3 missed. One consistent detection fail is that the -IL missed every time the S3 detected a mouse running through a trip line. I naturally don't know if the S3 detected mice reliably, or only occasionally.
2. The -IL is way better than the S3 at not falsely detecting blowing grass and tree limbs. The S3 is horrible with these.
Bad news:
1. The -IL false trips when slow moving shadows are present. Every morning as the sun is rising, it repeatedly false triggers as the shadows move. It happens over a 2 hour period in the early morning, abruptly starting, then abruptly ending. There have been as many as 50 false triggers over the 2 hours. The majority of the false triggers happen when the IVS decides that a particular rock plus a shadow is one of its objects. What seems horribly wrong is that even if the shadow makes the rock look like a human, vehicle, or animal, the rock isn't moving. My understanding of tripwires is that they trigger only when an object moves across the trip line.
I'm going to move the rock out of the FOV to see if that helps.