Check data and PSD computations from TD to FD

[SSL32081]

Need to ensure the way we obtain FD data and PSDs from the given TD data segment is faithful.

It needs to be robust against different segment durations and regardless of data source (e.g. GWOSC, or numpy arrays, or even frame files).

TODO (all these are to compare with Bilby outputs):

• TD data from GWOSC to FD strain
  • Test different durations
  • Test different TD window
• Long TD data from GWOSC to PSD
  • Test different durations
  • Test different TD window
  • Test PSD method (maybe?)

Ideally, all these should be computed with the same algorithm, and the expectation should be down to machine precision.