Photometric linearity

The degree to which an instrument's photometric response follows the expected proportional relationship within a specified range.

When a sample is diluted proportionally, its measured absorbance should change proportionally within the applicable range. Departures from this relationship cause the same conversion formula to produce different errors in different concentration regions.

Linearity concerns the shape of the response

Photometric linearity describes whether the response to a series of known transmittance or absorbance levels maintains the specified functional relationship. It requires multiple levels, a defined fitting model and evaluation of residuals, rather than a comparison at a single measurement point.

Photometric accuracy concerns closeness to reference values; linearity concerns whether response remains proportional as the signal changes. Several points can lie on a straight line while the whole line is biased, so good linearity does not imply high accuracy.

Why the high-absorbance region tends to curve

As absorbance rises, the intended transmitted light rapidly weakens. Stray light adds a relatively fixed signal, preventing apparent transmittance from falling as expected and often causing absorbance to rise more slowly. Nonlinearity in the detector, amplifier and analog-to-digital conversion can also change the curve.

A short path length can move the sample into a lower-absorbance region, but the effective path length itself must be reliable. If different paths are used separately, combining all results into one line must not obscure the calibration and switching conditions of each path.

Dilution linearity is not universal proof

Diluting a real sample also changes salt concentration, solvent proportions, chemical equilibria and scattering. It therefore tests the sample and procedure as well as the instrument. Traceable reference materials are more suitable for evaluating the photometric scale; real-sample series verify a specific method.

Evaluate residual patterns, range endpoints and repeated measurements, not just the correlation coefficient. A very high coefficient can conceal endpoint bias and cannot establish that a method remains linear outside the tested range.

Photometric linearity needs a defined range, material and error criterion, not a straight line without conditions.