I created SFTs for the same 1800-second segment using window sizes 512 and 1024.
The following lines were suppressed during run D:
The power plots describe the difference between power computed from SFT coefficients. The SFT plots describe the distance between two coefficients as complex numbers.
Two flavors of plots are available for comparison:
The averaging was produced using the current and previous 63 bins. The reason it is not centered on the bin is because this was easier to implement with current code.
All plots are logarithmic.
Power |
||
Entire file |
||
|
|
|
150 Hz - 1000 Hz |
||
|
|
|
650 Hz - 670 Hz |
||
|
|
|
200 Hz - 400 Hz |
||
|
|
|
240 Hz - 270 Hz |
||
|
|
|
250 Hz - 255 Hz |
||
|
|
|
251.25 Hz - 251.35 Hz |
||
|
|
|
263 Hz - 263.1 Hz |
||
|
|
|
SFT |
||
Entire file |
||
|
|
|
150 Hz - 1000 Hz |
||
|
|
|
650 Hz - 670 Hz |
||
|
|
|
200 Hz - 400 Hz |
||
|
|
|
240 Hz - 270 Hz |
||
|
|
|
250 Hz - 255 Hz |
||
|
|
|
251.25 Hz - 251.35 Hz |
||
|
|
|
263 Hz - 263.1 Hz |
||
|
|
|