22w27a

New Features

• New option, --resolution let's users set timeline resolution. This can be used to change the resolution used in video rendering.
• New short -p added as an alias for --player
• The attribute 'ColorSpace' is now written for fcpxml files. #283
• You can now edit different audio tracks even if not all inputs have that amount of audio tracks. #241
• Color space metadata from the first video file is added to output as long the output extension is the same as the first video's.
• Info subcommand now shows color range, color space, color primaries, and color transfer

Bug Fixes

• Fixed a bug where every text object uses the same text size as the first text object
• Fixed a bug when exporting certain timelines to Premiere Pro caused a traceback

Dependencies

• Bump pillow to 9.2.0

Breaking Changes

• Variables centerX and centerY no longer exist. Use position percent units instead (50%, 34%)
• Data streams are no longer copied due to sometimes creating invalid videos.
• info subcommand: Default view now strips underscores out of key strings.
• info subcommand: "aspect ratio" key is now "aspect_ratio" in json export.

🎉 Auto-Editor Reached 1000+ stars

Full Changelog: 22w25a...22w27a

22w25a

New Features

• Different videos (even with different resolutions) can now be combined by auto-editor. Auto-Editor uses --background to set what value to fill in empty space.
• Video rendering with speeds other than 1 now render correctly
• New --player option, set what program to open output files.
• Allow the number type to accept have underscores (_) as spacers and accept fractions. e.g. --frame-rate 30_000/1001

Bug Fixes

• Fix file seeking regression

Full Changelog: 22w24b...22w25a

22w24b

Re-release to add proper Windows files
See Release notes for 22w24a

Full Changelog: 22w24a...22w24b

22w24a

New Features

• Options that use frame_type, range_type can now use minutes and hours units in addition to seconds units
• Video rendering can now handle complex timelines but doesn't combine other videos correctly.

Bug Fixes

• Revert Windows FFmpeg to 5.0
• Properly find FFmpeg on M1 Macs #263
• Disable linking to X11 lib on MacOS FFmpeg #291
• Make chunks data render the right number of frames

Bugs Added

• Video rendering with speed not in (1, 99999) adds wrong number of frames

Full Changelog: 22w22a...22w24a

22w22a

New Features

• Upgrade ffmpeg to 5.0.1 and use stable ffprobe to get info about media
• Add vp9 encoders and decoders to ffmpeg
• Upgrade pillow, don't install yt-dlp by default
• Add --frame-rate option, which allows you to change the output media's fps

Breaking Changes

• With the exception of clip-sequence, Auto-Editor will now make one, and only one, output media file.
• --combine-files is now implicit. Every input file will be combined into one file. This has been implemented for audio files, however, video and subtitle files get Error! Timeline too complex because I haven't gotten around to implementing NLE in those methods yet.

Full Changelog: 22w18a...22w22a

22w18a

New Features

• Add new editing method random, which randomly marks frames as "loud" or "silent" based on a seed.
• Add option --no-seek which disables file seeking when rendering video streams. Using this option may help with desync issues.
• Add "codec context" timebase in info's output

Bug Fixes

• Fix desync issues when exporting as a media file with decimal framerates

Full Changelog: 22w17a...22w18a

22w17a

Features

• Make info subcommand output valid YAML; add timebase information by default
• Add support for GPU specific encoders and add their pix_fmt information #262 #267

Bug Fixes

• Re-add libopenh264.dll to pip wheels #265
• Fix divide by zero error when audio is completely silent #264

Full Changelog: 22w16a...22w17a

22w16a

Features

• Audio files and video's with audio greater than 4 hours can now edited and processed correctly
• Embedded images will now be copied from input to output
• Help messages have been retouched and corrected
• Using 32-bit Python is now supported and will not emit a warning

Bug Fixes

• Fixed a crash when using a speed so fast, that the sped up audio's length is 0
• Stopped random messaged from polluting stdout like "Stream 0 does not have timescale set"

Dependencies

Bump PyAV to 9.2.0

New Contributors

• @chapmanjacobd made their first contribution in #258

Update April 25, 2020

A Windows DLL file was accidentally not included in this pip release, so it is recommended that all Windows users upgrade to 22w17a

Full Changelog: 22w14a...22w16a

22w14a

Features

• Auto-Editor can now download urls in containers other than .mp4

Bug Fixes

• Fix regression in exporting to premiere

Dependencies

Bump pillow to 9.1.0

Breaking Changes

• The option --cut-by-this-audio is removed.

Full Changelog: 22w13a...22w14a

22w13a

New Features:

• Add ability to specify attributes for individual edit methods in the --edit option.
  • One example is --edit audio:threshold=0.03 or audio:stream=1:threshold=0.07
  • See auto-editor --edit --help For all the attributes each edit method has
• Add new editing method pixeldiff
  • pixeldiff marks sections as loud whenever the number of pixels is changed by a certain amount. You can set that amount by settings pixeldiff's threshold attribute. --edit pixeldiff:threshold=30 where 30 is the minimum amount of pixels that needs to be changed in between frames.

Bug Fixes:

• Fix Crash when description has invalid unicode characters (Issue #252)

Breaking Changes:

• --motion-threshold, --md-width, and --md-blur have been removed in favor of the new syntax:
  --edit motion:threshold=2%,width=900,blur=9

Dependencies:

• Bump PyAV to 9.1.0

Full Changelog: 22w12a...22w13a

Update April 5, 2022

--export premiere has a regression which causes Premiere Pro to output: The Importer reported a generic error. This is because samplerate is incorrectly set to None.

22w12a or a future version can be used instead.