Additional observation: "Moving" reported despite motion energy never approaching threshold
While testing gate thresholds, observed the sensor consistently reporting result=1 (moving)
even when motion energy stayed well below its threshold on every gate throughout, and only
micro/static energy came close to (and briefly touched) its own threshold.
Example (single frame, gate index : motion energy dB / motion threshold dB, micro energy dB / micro threshold dB):
Gate 5: motion 30.29 / 37.99 (7.7 dB below) -- micro 34.99 / 34.99 (at threshold)
Gate 6: motion 30.44 / 37.99 (7.5 dB below) -- micro 34.51 / 34.99 (0.5 dB below)
Sampled 12 consecutive frames (~300ms apart) during a seated, mostly-still period:
- Motion energy never crossed its threshold on any gate, in any frame (consistently 3-9 dB below).
- Micro energy on gates 5-6 repeatedly sat within ~0.5 dB of its threshold, touching it exactly in one frame.
- Result byte was 1 (moving) in every frame where a target was present -- never 2 (still) -- despite
motion never being the thing near its threshold.
Per the datasheet's documented decision logic (any gate's moving energy over its threshold -> moving,
else any gate's static energy over its threshold -> still, else none), this case should have reported
"still," not "moving."
Question: is there additional undocumented logic governing the moving/still classification -- e.g. new
detections defaulting to "moving" for an initial period regardless of which threshold triggered them,
or a different evaluation window than a single frame? Documentation of the actual decision logic beyond
the simple per-frame threshold comparison in Table 5-7 would help us tune thresholds correctly.
