Ffmpeg-normalize

* Code cleanup, add option to set format and audio codec.

* Add option for no prefix, fixes 20.

* Handle multiple spaces in path; fixes issue 18.

* Handle spaces in path, fixes 12.

* Update README.rst.

* Change default level back to -26.

* Typo in README example.

* Update documentation.

* Bump to v0.2.0.

* Support for multiple files and output directories.
* Support merging of audio with input file
* Set audio codec and additional options
* User-definable threshold
* Better error handling and logging
* Deprecates avconv

* Change default level back to -28.

* Merge pull request 15 from auricgoldfinger/master.

Add extended normalisation options

* Add extended normalisation options.

- add program option to write output in a separate directory in stead of
prefixing it

- add program option to merge the normalized audio in the original
(video) file rather than creating a separate WAV file

- change the maximum setting: will now normalize so that max
volume is set to 0, adjusted with the given level.
e.g. : -m -l -5 will increase the audio level to max = -5.0dB

- improve verbose logging: number of files are written to the
info log

- improve performance: check first whether the output file
exists before calculating the volume levels + not modifying
the file if the adjustment < 0.5dB (level is never exactly 0)

* Update README, fixes 11.

* Check for Windows .exe, fixes 10.

* Check path and fix 9.

* Merge pull request 8 from benjaoming/master.

Add MANIFEST.in

* Bump version.

* Add manifest to include missing files in sdist.

* Merge pull request 6 from jetpks/master.

Fixed ffmpeg v2.6.3 compatibility and docopt config

* Updated to work with ffmpeg v2.6.3, and fixed broken docopt config.

ffmpeg update:

ffmpeg v2.6.3 puts mean_volume on stderr instead of stdout, causing
`output` in `ffmpeg_get_mean` to be completely empty, and no match for
mean_volume or max_volume to be found.

Fixed by adding `stderr=subprocess.PIPE` in both Popen calls in
`run_command`, and combining stdout and stderr on return. We already
exit with non-zero return, so combining stderr/stdout shouldn't cause
any poor side-effects.

docopt config:

- args['--level'] was not recognizing its default because there was
an errant comma between -l and --level, and it needed <level> after
the arguments.
- Fixed spacing for --max
- Removed quotes around 'normalized' so single quote characters don't
end up in the output file names.

* Removed Windows carraige returns from __main__.py.

* Merge pull request 5 from mvbattista/master.

Installation update to ffmpeg

* Installation update to ffmpeg.

* Update to ffmpeg.

* Update HISTORY.rst.

* Update to ffmpeg.

* Merge pull request 4 from benjaoming/rename.

Rename project

* Make at least one file mandatory.

* Rename project and remove pyc file.

* Merge pull request 2 from benjaoming/docopt-setuptools-avconv.

Docopt, Setuptools, avconv compatibility

* Use docopt.

* Use normalize-audio when using avconv because it doesn't have a way to measure volume.

* Functional setup.py, communicate with avconv/ffmpeg about overwriting.

* Also detect avconv.

* Use a main function instead.

* Add a history for the project.

* Move to more unique module name.

* Update README.rst.

* Change the README to rst (PyPi)

* Delete .gitignore.

* Update README.md.

* Various improvements, fixes 1.

* License.

* Livense.

* Update README.md.

* Merge branch 'master' of https://github.com/slhck/audio-normalize.

* Initial commit.

* Initial commit.