Release v1.1.0

• Added: support for non single precision floating-point formats
  in audresample.remix()
• Changed: raise a RuntimeError in audresample.resample()
  when a non single precision floating-point input signal is given
  instead of converting it silently