Fig. 4From: Bias correction for direct spectral estimation from irregularly sampled data including sampling schemes with correlationMean of the primary estimates (no correction) of the power spectral density from 1000 signals. a Ideal random and independent samples from continuous time with a continuous distribution of sampling intervals, b random samples from continuous time with a continuous distribution of sampling intervals with processor dead time, c independent outliers in equidistant time with a discrete distribution of sampling intervals, and d correlated data gaps in equidistant time with a discrete distribution of sampling intervals (au, amplitude unit; tu, time unit)Back to article page