Signal Floor
Light intensity detection reaches a definitive boundary when individual pixels in a charge coupled device accumulate sufficient charge to fill their full well capacity. Photometric saturation occurs when the rate of incident photons exceeds the capacity of the semiconductor potential well to convert that flux into a measurable electrical signal. Once the pixel reaches this limit, excess electrons spill into adjacent photo-sites or drain into the substrate.
This overflow results in a flattened data peak where intensity increments produce no further output. Calibration protocols identify this threshold as the gain setting peak before the onset of non-linear response.
Response Limit
Sensor electronics determine the maximum voltage output allowed before the analog to digital converter clips the incoming data. Photometric saturation limits the usable dynamic range of an optical system by defining the ceiling of usable grey levels. Designers prevent data loss by adjusting aperture or integration time to ensure the peak irradiance stays below the sensor well depth.
Manufacturers specify this performance threshold as the point where the deviation from linear output exceeds a defined tolerance.
Charge Overflow
Electrons generated by high photon flux accumulate until the potential well physical barriers fail to contain them. Photometric saturation triggers the blooming effect where excess charge leaks into neighboring pixel rows during extended exposure. Vertical streaks appear in captured imagery as a direct consequence of this electron leakage along the column registers.
Modern interline transfer sensors incorporate drains to mitigate this spillover, though even these mechanisms fail under extreme illumination conditions.
Data Validity
Signal quantification terminates at the boundary where the sensor output ceases to track the linear relationship between light intensity and voltage. Analysts apply dark frame subtraction and flat field correction to distinguish between true optical saturation and fixed pattern noise or thermal bias. Data recorded beyond this peak carries no information about the variance in light intensity across the scene.
Accurate measurements require the signal to remain strictly within the range defined by the sensor read noise floor and the saturation limit.