BUG: use dpss instead of slepian in audiofileIO.py

note that this merely fixes an AttributeError,
`AttributeError: module 'scipy.signal' has no attribute 'slepian'`

There's a deeper issue with whether I'm calling `dpss` with the
right values, see
#91

src/hvc/audiofileIO.py

@@ -182,7 +182,7 @@ def __init__(self,
                     if window == 'Hann':
                         self.window = np.hanning(self.nperseg)
                     elif window == 'dpss':
-                        self.window = scipy.signal.slepian(self.nperseg, 4 / self.nperseg)
+                        self.window = scipy.signal.windows.dpss(self.nperseg, 4 / self.nperseg)
 
         if freq_cutoffs is None:
             self.freqCutoffs = None