We show how to correct polarization measurements performed with non-id
eal polarizers. The formulas presented in this note can be used to per
form precise polarization measurements using inexpensive polarizing sh
eets. The simplest method to measure the state of polarization of (com
pletely polarized) light, is using a rotatable polarizer and a photome
ter, as shown in Figure 1. Specifically, one measures the output inten
sities along the axes of maximum and minimum polarizer transmission, I
-max and I-min. When I-min = 0, one has linearly polarized light; when
I-max = I-min, the light is circularly polarized, and, in general, wh
en I-max not equal I-min, the light will be elliptically polarized wit
h ellipticity omega = arctan ([I-min/I-max](1/2)). When the polarizer
is not perfect-the polarizing degree of an inexpensive polarizer is no
t 100%-instead of measuring the projection of the incident field along
the transmission direction of the polarizer, what we actually measure
is a sum of projections along the transmission direction and its perp
endicular, weighted by the respective coefficients. This modifies the
measurements sometimes significantly. For example, we could obtain I-m
in not equal 0 even for linearly polarized light. In this note, we pro
pose a simple device-built with inexpensive polarizing sheets-to deter
mine the state of polarization of light that compensates for the defic
iencies introduced by non-ideal polarizers.