Video editing with Python
There are still no major breaking changes compared to v1.0.3. Expect them to come in the next dev update.
Any new changes made to the master branch will now be instantly reflected at https://moviepy.readthedocs.io, which is where documentation for all versions will be in the future. [#1328]
Install with pip install moviepy --pre --upgrade.
pix_fmt parameter in VideoFileClip, VideoClip.write_videofile(), VideoClip.write_gif() that allows passing a custom pix_fmt parameter such as "bgr24" to FFmpeg [#1237]change_duration parameter in Clip.set_fps() that allows changing the video speed to match the new fps [#1329]ffmpeg_parse_infos() and VideoFileClip now have optional decode_file parameter that ensures that the detected duration is correct, but may take a long time to run [#1063, #1222]ffmpeg_parse_infos() and VideoFileClip now use fps metadata instead of tbr to detect a video's fps value [#1222]FFMPEG_AudioReader.close_proc() -> FFMPEG_AudioReader.close() for consistency with FFMPEG_VideoReader [#1220]ffmpeg_tools.ffmpeg_extract_subclip creating clips with incorrect duration metadata [#1317]OSError: MoviePy error: failed to read the first frame of video file... would occasionally occur for no reason [#1220]UnicodeDecodeError crash when file metadata contained non-UTF8 characters [#959]This development version introduces many bug-fixes and changes. Please note that there may be large backwards-incompatible changes between dev versions!
The online documentation has not been updated to reflect the changes in the v2.0.0 branch, so for help on how to use the new features please refer to the docstrings in the source code.
Install with pip install moviepy --pre --upgrade.
config_defaults.py and MoviePy still cannot autodetect the binaries, you will need to switch to the new method using enviroment variables. [#1109]==. This checks whether every frame of the two clips are identical.env file [#1109]encoding parameter in SubtitlesClip [#1043]ffmpeg_stabilize_video() function in ffmpeg_tools
temp_audiofile_path parameter in VideoClip.write_videofile() to specify where the temporary audiofile should be created [#1144]VideoClip.set_layer() to specify the layer of the clip for use when creating a CompositeVideoClip [#1176]ffmpeg_parse_infos additionally returns "video_bitrate" and "audio_bitrate" values [#930]VideoFileClip or AudioFileClip through videoclip.reader.bitrate and audioclip.reader.bitrate [#930]vfx.scroll arguments w and h have had their order swapped. The correct order is now w, h but it is preferable to explicitly use keyword arguments. in the output file name of ffmpeg_extract_subclip() when targetname is not specified [#939]config_defaults.py is no longer possible [#1109]config.get_setting() and config.change_settings() functions [#1109]AudioClip.to_audiofile() -> use AudioClip.write_audiofile()
VideoClip.to_videofile() -> use VideoClip.write_videofile()
VideoClip.to_images_sequence() -> use VideoClip.write_images_sequence()
concatenate() -> use concatenate_videoclips()
verbose parameter in AudioClip.write_audiofile(), ffmpeg_audiowriter(), VideoFileClip(), VideoClip.write_videofile(), VideoClip.write_images_sequence(), ffmpeg_write_video(), write_gif(), write_gif_with_tempfiles(), write_gif_with_image_io() -> Instead of verbose=False, use logger=None
verbose_print() -> no replacementcol parameter in ColorClip() -> use color
VideoClip.write_videofile() with write_logfile=True, errors would not be properly reported [#890]TextClip.list("color") now returns a list of bytes, not strings [#1119]TextClip.search("colorname", "color") does not crash with a TypeError [#1119]vfx.even_size previously created clips with odd sizes [#1124]IndexError in vfx.freeze, vfx.time_mirror and vfx.time_symmetrize [#1124]rotate() with a ColorClip no longer crashes [#1139]AudioFileClip would not generate audio identical to the original file [#1108]TypeError when using filename instead of txt parameter in TextClip [#1201]IndexError: list index out of range when using concatenate_videoclips [#646]resize with a non-constant newsize to a clip with a mask would remove the mask [#1200]color_gradient() would crash with ValueError: The truth value of an array with more than one element is ambiguous [#1212]Note that this is likely to be the last release before v2.0, which will drop support for Python versions 2.7, 3.4 & 3.5 and will introduce other backwards-incompatible changes.
Notable bug fixes:
slide_out effect works [#795]Fixed bugs:
Merged pull requests:
Closed issues:
ModuleNotFound: No module named 'imageio\_ffmpeg', or imageio v2.5.0 is breaking ffmpeg detection in config #906
Merged pull requests:
Closed issues:
Moviepy now supports much better progress bars, using Proglog. Support for custom callbacks and more!
Fixed bugs:
Merged pull requests:
Closed issues:
with\_mask\(\) #786
pcm_s24le codec for .wav files (#769)fullscreen parameter to preview (#773)
Merged pull requests:
fullscreen parameter to preview #773 (tburrows13)progress_bar parameter added to write_images_sequence
VideoFileClip objects now have a rotation attributeValueError: operands could not be broadcast together with shapes bug fixed (#334)
Requirements adjustments #530
Documentation build failures #525
Index is out of bounds - AudioFileClip #521
Youtube videos fail to load in documentation #536
unicodeDecoderError by running the setup.py during moviepy pip install #532
Should we push another version? #481
Add matplotlib example to the user guide? #421
Fails to list fx after freezing an app with moviepy #274
Documentation doesn't match ffmpeg presets #232
Fixed bugs:
Implemented enhancements:
Merged pull requests: