Signal recovery presents potential problems for most researchers and engineers at one stage or another in an experiment. Most often, the problem is either a very weak signal, or a very low signal to noise ratio. A common example is recovering a signal from a detector such as a PMT or photodiode, where the signal may be as low as a single photon per second, or may be buried in background scatter. There are three different signal recovery techniques pertinent to this discussion. A lock-in amplifier for phase-synchronous detection, a boxcar averager for gated integration or a photon counter for a statistical counting approach. Which method or instrument is best depends on many factors, including signal intensity, time and frequency distribution, the various noise sources and their time dependence and frequency distribution.