A process of modifying a time series of data whereby the beginning and ending 10% of the data points are smoothly reduced in amplitude to approach zero at the ends. Such conditioning of raw data is often necessary before performing a spectral or Fourier analysis. If this were not done, it is possible that spurious spectral signals might be introduced because the Fourier analysis assumes that the finite-duration time series is cyclic (which could cause a step discontinuity between the end of the series and the beginning of the repeated series). However, any time that data are modified, including the conditioning described here, the validity of the resulting analysis may be questioned.