50% of what? The ratio of the dT between the (sensor and the background) and the (sensor and target) shall influence the nature of the distribution as depicted at 2:09. ..or do you somehow ensure that the b/g is at the same temperature as that of the sensor (dT_1 = 0) before carrying out this experiment?
50 percent of normalized graph. We are not interested in absolute values returned from the sensor in this experiment that is why normalisation is done. Ideally the temperature of the object should be high enough to have clear graph. Normalized graph is not much sensitive from the temperature value itself. Definitely, object should be hot enough to be easily distinguished from the background and same time not to saturate the sensor. And yes, as any FIR sensor, it has its internal temperature sensor for compensation. You are right, ideally sensor temperature should be as background, but for most cases with proper object temperature selection, this requirement is not so strict.
50% of what? The ratio of the dT between the (sensor and the background) and the (sensor and target) shall influence the nature of the distribution as depicted at 2:09. ..or do you somehow ensure that the b/g is at the same temperature as that of the sensor (dT_1 = 0) before carrying out this experiment?
50 percent of normalized graph. We are not interested in absolute values returned from the sensor in this experiment that is why normalisation is done. Ideally the temperature of the object should be high enough to have clear graph. Normalized graph is not much sensitive from the temperature value itself. Definitely, object should be hot enough to be easily distinguished from the background and same time not to saturate the sensor. And yes, as any FIR sensor, it has its internal temperature sensor for compensation. You are right, ideally sensor temperature should be as background, but for most cases with proper object temperature selection, this requirement is not so strict.
S value is area or length of spot point?