Carview!

Don't show a runtime warning about the optional ffplay dependency being missing until someone trys to use it
Documentation improvements
Python 3.9 support
Improved efficiency of loading wave files with pydub.AudioSegment.from_file()
Ensure pydub.AudioSegment().export() always retuns files with a seek position at the beginning of the file
Added more EQ effects to pydub.scipy_effects (requires scipy to be installed)
Fix a packaging bug where the LICENSE file was not included in the source distribution
Add a way to instantiate a pydub.AudioSegment() with a portion of an audio file via pydub.AudioSegment().from_file()

Fix bug where ffmpeg errors in Python 3 are illegible
Fix bug where split_on_silence fails when there are one or fewer nonsilent segments
Fix bug in fallback audioop implementation

Fix inconsistent handling of 8-bit audio
Fix bug where certain files will fail to parse
Fix bug where pyaudio stream is not closed on error
Allow codecs and parameters in wav and raw export
Fix bug in pydub.AudioSegment.from_file where supplied codec is ignored
Allow pydub.silence.split_on_silence to take a boolean for keep_silence
Fix bug where pydub.silence.split_on_silence sometimes adds non-silence from adjacent segments
Fix bug where pydub.AudioSegment.extract_wav_headers fails on empty wav files
Add new function pydub.silence.detect_leading_silence
Support conversion between an arbitrary number of channels and mono in pydub.AudioSegment.set_channels
Fix several issues related to reading from filelike objects

Fix bug in passing ffmpeg/avconv parameters for pydub.AudioSegment.from_mp3(), pydub.AudioSegment.from_flv(), pydub.AudioSegment.from_ogg(), and pydub.AudioSegment.from_wav()
Fix logic bug in pydub.effects.strip_silence()

Add support for playback via simpleaudio
Allow users to override the type in pydub.AudioSegment().get_array_of_samples() (PR #313)
Fix a bug where the wrong codec was used for 8-bit audio (PR #309 - issue #308)

Fix pydub.utils.mediainfo_json() to work with newer, backwards-incompatible versions of ffprobe/avprobe (issue #285, PR #290)

Adds support for audio with frame rates (sample rates) of 48k and higher (requires scipy) (PR #262, fixes #134, #237, #209)
Adds support for PEP 519 File Path protocol (PR #252)
Fixes a few places where handles to temporary files are kept open (PR #280)
Add the license file to the python package to aid other packaging projects (PR #279, fixes #274)
Big fix for pydub.silence.detect_silence() (PR #263)

NOTE: Semi-counterintuitive change: using the a stride when slicing AudioSegment instances (for example, sound[::5000]) will return chunks of 5000ms (not 1ms chunks every 5000ms) (#222)
Debug output from ffmpeg/avlib is no longer printed to the console unless you set up logging (see README for how to set up logging for your converter) (#223)
All pydub exceptions are now subclasses of pydub.exceptions.PydubException
The utilities in pydub.silence now accept a seek_stepargument which can optionally be passed to improve the performance of silence detection (#211)
Fix to pydub.silence utilities which allow you to detect perfect silence (#233)
Fix a bug where threaded code screws up your terminal session due to ffmpeg inheriting the stdin from the parent process. (#231)
Fix a bug where a crashing programs using pydub would leave behind their temporary files (#206)

Add new parameter gain_during_overlay to pydub.AudioSegment.overlay which allows users to adjust the volume of the target AudioSegment during the portion of the segment which is overlaid with the additional AudioSegment.
pydub.playback.play() No longer displays the (very verbose) playback "banner" when using ffplay
Fix a confusing error message when using invalid crossfade durations (issue #193)

Provide feedback