[[5-1 Time-Domain Representations]] > [[root mean square (RMS)]] In signal processing, the signal energy is usually defined as the variance of the amplitudes of the signal: $ \text{Energy}(x(n)) = E[(x(n)- \mu)^2] $ where $\mu$ is the average of the signal: $ \mu = E[x(n)] $ The energy is usually estimated over some segment of the signal using a [[window function]]. The energy is usually measured in [[deciBel]] using either [[deciBel Full Scale (dBFS)]] or [[decibel to overload (dBov)]].