BF: fix for soundfile float division bug in duration

As it turns out it was inefficient to create sf.info(filename) for
this anyway. We now do tha calculation ourselves which means only
opening the file once

psychopy/sound/backend_sounddevice.py

@@ -360,7 +360,7 @@ def _setSndFromFile(self, filename):
         self.sampleRate = f.samplerate
         if self.channels == -1:  # if channels was auto then set to file val
             self.channels = f.channels
-        info = sf.info(filename)  # needed for duration?
+        fileDuration = float(len(f))/f.samplerate  # needed for duration?
         # process start time
         if self.startTime and self.startTime > 0:
             startFrame = self.startTime*self.sampleRate
@@ -371,19 +371,18 @@ def _setSndFromFile(self, filename):
         # process stop time
         if self.stopTime and self.stopTime > 0:
             requestedDur = self.stopTime - self.t
-            maxDur = info.duration
-            self.duration = min(requestedDur, maxDur)
+            self.duration = min(requestedDur, fileDuration)
         else:
-            self.duration = info.duration - self.t
+            self.duration = fileDuration - self.t
         # can now calculate duration in frames
         self.durationFrames = int(round(self.duration*self.sampleRate))
         # are we preloading or streaming?
         if self.preBuffer == 0:
             # no buffer - stream from disk on each call to nextBlock
             pass
         elif self.preBuffer == -1:
-            # no buffer - stream from disk on each call to nextBlock
-            sndArr = self.sndFile.read(frames=len(self.sndFile))
+            # full pre-buffer. Load requested duration to memory
+            sndArr = self.sndFile.read(frames=int(self.sampleRate*self.duration))
             self.sndFile.close()
             self._setSndFromArray(sndArr)