Vladimir Dergachev, tenaya webpage

First run of computing 1800-second SFTs

Analysis of double-valued line at 200-400hz

Power spectrum of L1 during S1, with Rayleigh statistic.

Power spectrum of L1 during S2, with Rayleigh statistic.

Power spectrum of L1 during S2, with Kolmogorov-Smirnov statistic.

First run of computing calibrated 1800-second SFTs

First run of computing calibrated 1800-second SFTs, coincidence plots.

Second run of computing calibrated 1800-second SFTs, 182 segments.

Third run of computing calibrated 1800-second SFTs, 1653 segments.

Cross section plots of third run SFTs

Comparison between Xavier's and Vladimir's SFTs

SFT-3 with calibration, fourth run, L1 data

SFT-3 with calibration, fourth run, cross section plots

Comparison between SFTs with window sizes 512 and 1024

SFT-3 with calibration, fourth run, L1 data, non-parametric estimation.

Comparison between Xavier's SFTs with Direchlet kernel sizes 16 and 64

Influence of line suppression on comparison of window tail sizes:

• Comparison between SFTs with window sizes 512 and 1024, no suppression
• Comparison between SFTs with window sizes 512 and 1024, run A
• Comparison between SFTs with window sizes 512 and 1024, run B
• Comparison between SFTs with window sizes 512 and 1024, run C
• Comparison between SFTs with window sizes 512 and 1024, run D
• Comparison between SFTs with window size 1024, no suppression versus run D
• Comparison between SFTs with window sizes 512 and 1024, run D, using averaged norm
• Comparison between SFTs with window sizes 512 and 1024, no suppression, using averaged norm
• Comparison between SFTs with window sizes 512 and 1024, with low- and high-pass filter
• Comparison between SFTs with window sizes 512 and 1024, with low- and high-pass filter and line suppression
• Comparison between SFTs with window tail sizes 512 and 1024, with correct filtering
• Comparison between SFTs with window tail sizes 20480 and 40960, with correct filtering and frequency window
• Comparison between SFTs with window tail sizes 20480 and 30000, with correct filtering and frequency window
• Comparison between SFTs with window tail sizes 20480 and 30000, with correct filtering and frequency window, different plot styles
• Comparison between SFTs with window tail sizes 20480 and 21504, with correct filtering and frequency window

Flowchart of make_sft code - this is the program used to generate the above data.

Stitched plots for H1

512 versus 1024 window sizes for high-pass only data

SFT-3 with calibration, fourth run, more cross section plots

Testing using generated data:

• Fake data, run 1
• Fake data, run 2
• Fake data, run 3, window comparisons
• Fake data, run 4, two sine waves
• Fake data, run 5, off-center sine waves

SFT-3 with calibration, third run, H1 data, full statistic

Fifth run, cross section plots

Fifth run, L1 plots