How SPECTER Detects Paranormal Activity: The Anomaly Index Explained

Most ghost hunting tools ask you to watch a dial and guess. SPECTER takes a different approach: it fuses six live measurements into a single confidence score and logs the evidence for you. Here is exactly how that works.

One number that answers one question

The Anomaly Index is SPECTER's primary readout. It is a single percentage that answers the only question that matters in the moment: is something happening right now? Every other meter exists to provide context. The Anomaly Index is the decision signal.

It is a weighted blend: the strongest candidate region in the frame contributes half the score, a collapse in field coherence contributes thirty percent, and deviation from the live baseline contributes the rest. When the blended score crosses a threshold you set, SPECTER fires a detection event and archives the evidence automatically.

The six metrics behind it

Luminance Flux

Mean brightness of the frame plus its change from a rolling average. A sudden drop means something occluded the lens or passed between the sensor and a light source. A spike may correlate with reported flash phenomena.

Spectral Density

Shannon entropy of the color histogram, a measure of chromatic complexity. A sudden collapse means a large uniform object entered frame. A sudden expansion means something with a different color palette appeared, including forms with a shifted near-infrared signature.

Particulate Flow Vector

Dense optical flow across the whole image, reporting the dominant direction and speed of movement. This is the system's primary detector of movement that bypasses the visible spectrum, because it measures displacement in the image plane, not light.

Field Coherence

The structural integrity of movement in the medium. A clean smoke cloud drifts as one coherent body. When something moves through it, the smoke separates and coherence collapses. This is the most sensitive detector of presence in a fog investigation, and it carries the highest weight in the index.

Baseline Deviation

The Euclidean difference between the current frame and a live, continuously updated picture of "normal" for the room. It is not a static reference shot at session start, it adapts as the environment settles.

The Anomaly Index itself

The composite of the above. You choose how sensitive it is with three detection classes.

Detection classes

• CLASS-I (40%): low-confidence anomaly. Archive entry created, keep monitoring.
• CLASS-II (60%): strong anomaly. Archived with a depth map badge if a depth sensor is active.
• CLASS-III (80%): exceptional anomaly. Full evidence package, all active metrics snapshotted.

It watches so you do not have to

The most important part: this runs whether or not you are looking at the screen. The archive logs every event autonomously. Anomalies that happen while you are reacting to something else, setting up gear, or reviewing a previous capture are still logged in full. And SPECTER never post-processes evidence frames, what the sensor recorded is what the archive contains, to preserve chain-of-custody integrity.