Snack sound toolkit

64-bit builds are now supported. A new -regressionorder option was added for the speatures command. Python installation was simplified through a setup.py script. Updates were made for the ALSA 1.0 API and bugfixes were made to the AIFF header parsing routine, Mac OS X echo effects, the pitch command for short sounds (ESPS method), mean calculations for dBPowerSpectrum, and for handling busy audio devices under Linux. Better spectral estimations were implemented. A queued sound flooding bug was fixed, a playback was cleaned up, a memory leak was corrected, and out-of-bounds checks were added.

This release adds high level sound objects with flexible storage
management and streaming support, multiple simultaneous playback and
recording threads, a script compilation tool for easy deployment using
stand-alone executables, filters to process sound objects or use
during playback for on-the-fly operation, realtime
visualization/spectrum analysis with waveforms, spectrograms, and
spectrum sections, and Postscript support. All audio data is now
handled as floating point internally for fast and accurate
computation.