Abstract

The use of a photomultiplier to measure the reduction in counting statistics noise (from the usual N1/2 form) which is expected to occur for squeezed coherent light is shown to lead to no reduction or to an increase of the noise unless the number of photons in a fundamental measuring time of the photomultiplier is very large. This fundamental time is estimated to be less than 10−13 sec for ordinary detectors. A technique for using squeezed states for determining the time a photomultiplier takes to detect the presence of a photon is presented