ENH: Add trough_threshold parameter to feature.sat.pitch

[ralphpeterson]

I was trying to compute the fundamental frequency of some ultrasonic vocalizations and noticed that the pitch function (vocalpy.feature.sat) returned wonky results. It turns out that the librosa yin function referenced in pitch has an additional argument, trough_threshold that when changed helps produce better f0 estimates (see example below with black line showing f0 estimate from librosa.yin). Not sure exactly the underlying reason, but nonetheless we should have access to this argument in vocalpy.

In this PR I simply added trough_threshold as an argument in the pitch function.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (d58d43f) 79.75% compared to head (a07aa8e) 79.75%.
Report is 2 commits behind head on main.

❗ Current head a07aa8e differs from pull request most recent head 875f669. Consider uploading reports for the commit 875f669 to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #117   +/-   ##
=======================================
  Coverage   79.75%   79.75%           
=======================================
  Files          29       29           
  Lines         968      968           
=======================================
  Hits          772      772           
  Misses        196      196           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[NickleDave]

Awesome, thanks so much @ralphpeterson.

We definitely needed to add this. The visual is quite helpful -- maybe we should add something like that to docs at some point, e.g. a plot in the feature.sat.pitch docstring.

Overall this looks great, I just made a couple tweaks, like adding the same parameter to feature.sat.similarity_features.

Merging in! Much appreciated