Skip to content

Latest commit

 

History

History
156 lines (124 loc) · 4.42 KB

README.md

File metadata and controls

156 lines (124 loc) · 4.42 KB

Multicam Podcast Editor

A command-line tool designed to automate video editing tasks, with a focus on podcasts, multicam recordings, and creative video projects. It uses audio analysis and advanced editing techniques to simplify the creation of professional-quality videos with minimal manual effort.

Features

  • Automatic Multicam Editing: Seamlessly switches between multiple camera angles based on audio levels, highlighting the active speaker or key moments.
  • Short Format Multicam Editing: Generates concise clips from multicam setups, perfect for highlights or social media.
  • Video Cutting: Allows precise removal of unwanted segments by specifying time ranges.
  • Automatic Jumpcuts: Detects and eliminates pauses to improve pacing.
  • Transcription and Captioning: Transcribes audio and adds customizable captions.
  • Music Video Creation: Combines images and MP3s into music videos with transitions.
  • Audio Enhancements: Applies filters for improved audio quality (e.g., noise reduction, equalization).

Installation

  1. Clone the Repository:

    git clone https://github.com/horvay/multicam-podcast-editor.git

  2. Install Dependencies: Use uv to install the required Python packages:

    uv sync

    Install ffmpeg and auto-editor for audio/video processing:

    # On Ubuntu/Debian
sudo apt-get install ffmpeg
# On macOS with Homebrew
brew install ffmpeg
# Install auto-editor
uv pip install auto-editor

  3. Verify Setup: Check that everything works by running:

    uv run main.py -h

    This shows the help menu with command-line options.

Usage

Run the tool with uv run main.py followed by specific flags and arguments. For all options, use:

uv run main.py -h

Automatic Multicam Editing

Edit a multicam video with a main video and individual angles:

uv run main.py -m -i combined_vid.mp4 -i person1.mp4 -i person2.mp4 -i person3.mp4
  • -m: Multicam editing mode.
  • -i: Input files (first is the main video, followed by angles).

Add screenshare videos:

uv run main.py -m -i combined_vid.mp4 -i person1.mp4 -i person2.mp4 -si screenshare1.mp4

Short Format Multicam Editing

Create a short clip with a start time (and optional end time):

uv run main.py -s 127 -ti 148 -i combined_vid.mp4 -i person1.mp4 -i person2.mp4
  • -s: Start time (seconds).
  • -ti: End time (seconds; defaults to 60 seconds after start if omitted).

Cutting Videos

Remove segments by specifying time ranges:

uv run main.py -c 1.0 2.0 -c 3.0 4.0 -i input_video.mp4
  • -c: Start and end times (seconds) to cut.

Automatic Jumpcuts

Remove pauses:

uv run main.py -j -i input_video.mp4
  • -j: Jumpcut mode.
  • Adjust silence threshold (default 0.75 seconds):
uv run main.py -j --jump-cuts-margin 0.5 -i input_video.mp4

Transcription and Captioning

Transcribe a Video

Generate a transcript:

uv run main.py -t -i input_video.mp4
  • -t: Saves transcript to output/input_video.mp4.csv.

For multicam, create a chat-style transcript:

uv run main.py -t -i combined_vid.mp4 -i person1.mp4 -i person2.mp4
  • Output: output/transcript.txt.

Add Captions

Add captions using a CSV:

uv run main.py -cv -i input_video.mp4 -csv output/input_video.mp4.csv
  • -cv: Captioning mode.
  • -csv: Transcription CSV file.

Customize captions:

uv run main.py -cv -i input_video.mp4 -csv output/input_video.mp4.csv -ct 2 -cp "200, 100" -cs "300, 500" -fs 80
  • -ct: Caption type (1: across, 2: multiline).
  • -cp: Position (x, y).
  • -cs: Size (width, height).
  • -fs: Font size.

Music Video Creation

Create a music video from directories:

uv run main.py -mv -i music_dir -i art_dir -i reminders_dir -th thumbnail.png
  • -mv: Music video mode.
  • -i: Directories for music, art, reminders.
  • -th: Optional thumbnail.

Audio Enhancements

Enhance audio for podcasts:

uv run main.py -ape -i input_video.mp4
  • -ape: Podcast enhancements.

For music:

uv run main.py -ame -i input_video.mp4
  • -ame: Music enhancements.

Additional Notes

  • Output Files: Saved in output/ with names like final_<random>.mp4 (customizable via --output-name).
  • Temporary Files: Processed in temp/, cleared each run.
  • Customization: Use --threads (default: 10) or --hi-def for 1080p.