Skip to main content

AutoClip

Automagically clip out objects obstructing your sign.

Autoclip gif

Recommended installation procedure (not following these may lead to errors)

  • Prerequisites

    • Install Python 3.12
      • It is highly recommended to install Python for all users. (configurable in the installer on Windows)
    • Install Vapoursynth from here: https://github.com/vapoursynth/vapoursynth/releases
      • Installation through pip may lead to problems, so it's recommended to use the installers from the link above unless you know what you're doing.
      • Recommended to also install for all users. (configurable in the installer on Windows)
  • Installing the Aegisub script

  • First-time Setup

    • When you start Autoclip from Aegisub for the first time, a setup wizard is shown, follow the steps, and always check whether the settings match your system.
    • The wizard is going to run commands on your system, but will always show them and give you a chance beforehand to edit them if necessary

Usage

  1. Time your sign to the whole cut.
  2. Seek the video to a reference frame where, ideally, the sign is unobscured from the foreground object.
  3. Create a rectangular clip that covers your sign across the whole cut. This clip defines the area where AutoClip should be active. Anything outside this clip will not be clipped.
  4. Select all the lines that make up the sign. This can include already fbfed or layered lines. AutoClip recognises lines not based on their order, but their start and end time.
  5. With the video seek head at the reference frame and all lines selected, run „Automation > AutoClip > AutoClip“ and a new AutoClip window shall open.
  6. In the new window, adjust the sliders until you get a satisfactory clip. AutoClip tests the absolute difference of a pixel at the current frame against the reference frame, and clips away the pixel if it has a difference above the specified thresholds.
  7. Click „Apply“ to apply the clip to your subtitles.
  8. Be aware, that not fbf lines are going to be converted to fbf. If you need to apply motion tracking or similar to your sign, make sure you do them before using autoclip.

AutoClip can now merge existing clips with incoming AutoClip:

  1. Time your sign to the whole cut.
  2. Create a rectangular clip that covers your sign across the whole cut. This clip defines the area where AutoClip should be active. Anything outside this clip will not be clipped.
  3. Run „Automation > AutoClip > Set or Unset Active Area“. The active area will only be used if the later AutoClipping are performed at or near the timing of the selected lines.
  4. Remove the rectangular clip and typeset the sign as normal. If there is anything else that needs clipping, for example, clipping text body onto the grain layer, or clipping the text body onto border, it should be performed in this step.
  5. Select all the lines of the sign.
  6. Seek the video to a reference frame where, ideally, the sign is unobscured from the foreground object.
  7. Click „Automation > AutoClip > AutoClip“. The previously set active area will be prioritised over existing clips on the line.
  8. In the new window, adjust the sliders until you get a satisfactory clip as normal, and click „Apply“.
  9. Back to Aegisub, a new window will pop up asking you how you would want to merge the existing clips with incoming AutoClip. This window will pop up for every different layers in the selected lines, and for each layer, you can choose to replace existing clip With AutoClip, apply AutoClip in additional to existing clipping (iclip OR), apply AutoClip to existing clip (iclip Subtract), apply existing clip To AutoClip (iclip Subtract), apply iclip AND, apply iclip XOR, or keep existing clip.
  10. Since active area set in step 3. are only automatically used if the later AutoClipping are performed within the start frame and the end frame of the selected lines, in most cases, you don't need to unset the active area. However, if you want to unset active area, select any line that doesn't contain a rectangular clip and click „Automation > AutoClip > Set or Unset Active Area“.

License

  • AutoClip is released by Zahuczky and Akatsumekusa under BSD 3-Clause License.
  • AutoClip uses Noto Sans Display Medium in the UI. Noto Sans Display is released by Google under SIL OFL 1.1 License.

TODO

  • AutoClip UI shows nested clips but nested clips don't work. I never remembered fixing this bug but it seems that it has fixed itself somehow? Available in 2.1.0.
  • Handle moving signs/tracking data for those
  • More sliders (probably a bunch of other VS filters to fine-tune the clip area)
  • Simplifying the clips to curves and such.
  • Currently only the longest contour gets taken into account. Maybe combine them? Maybe a slider for this?
  • Add clips to existing clips on the line.

Release files for ass-autoclip 2.1.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for ass-autoclip 2.1.1
File Size Uploaded
ass_autoclip-2.1.1.tar.gz 210.2 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for ass-autoclip 2.1.1
File Interpreter ABI Platform
ass_autoclip-2.1.1-py3-none-any.whl Python 3 none any Details

Total release size: 421.6 kB

Release files / ass_autoclip-2.1.1.tar.gz

Download URL ass_autoclip-2.1.1.tar.gz
Size 210.2 kB
Tags Source
SHA-256 checksum
How to use checksums
d14602c2507feb585f8401ad321a5dbbc75d319529cc66b11386176a394bc465
BLAKE2b-256 checksum
How to use checksums
11128f00b8968b76261244d7d6527a8569118d367bf19f549983b5314023b7c1
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.12.3

Release files / ass_autoclip-2.1.1-py3-none-any.whl

Download URL ass_autoclip-2.1.1-py3-none-any.whl
Size 211.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
d0257159054e90560cdadb75d12a64d115c93115d1b22455dfb2f01ee3604a01
BLAKE2b-256 checksum
How to use checksums
fec03d768b74290100b3a27e2159e2a44cb9fd18f0cf9137f8fe49de7ab872fe
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.12.3

Release history Release notifications | RSS feed

This release

2.1.1 This release

2 release files

2.0.5

2 release files

2.0.3

2 release files

2.0.1

2 release files

2.0.0

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page