Fixed time slicing bug, and plotting bug. Verified on small sample size with previous results

I removed the "slice window" because it was redundant and there was bug that caused the time windows not to be centered on the pick. I also cleaned up the debug plotting so it won't break with different time windows.

Right now I have the time window set to 12s, so it grabs data 12s before and 12s after the pick. This is because the pick can move up to 3 s, and the window for the phase and distance classification is +- 7 seconds. So potentially, we would need 10 s after the pick time. The extra couple seconds are for the tapering prior to filtering.