In Figure 11, the distribution curve flattens out at about 100 percent of the current thresholds for the first window. For the second window, the distribution flattens out at about 200 percent of the current thresholds. Because the thresholds are linearly proportional to the specified threshold factors (THRESHOLD_FACTOR parameter), increasing the specified threshold factors by 0 and 100 percent effectively moves the thresholds down the trace by 0 and 100 percent, respectively. Figure 12 shows the stacked normalized amplitude distribution after increasing the threshold factors used for Figure 11. The threshold factors for the first window are unchanged and those for the second window are increased by 100 percent. The data has been re-normalized and the thresholds have been reset again at 100 percent on the plot. Neglecting numerical round-off, the data from 1 to 200 percent in Figure 7 has been compressed 1 to 100 percent in Figure 8 due to the new thresholds. The data after the thresholds is much flatter in Figure 8 than in Figure 7. Figure 12. The stacked normalized amplitude distribution after increasing the threshold factors used for each window in Figure 7 by 0 and 100 percent, respectively. Notice the number of frequency bands greater than the threshold is considerably smaller and the distributions are flatter after 100 percent than that seen in Figure 7.