uw-panopto-downloader 1.1.5

A Python library for downloading videos from UW Panopto

UW Panopto Downloader

A Python library and CLI tool for downloading and processing videos from University of Washington's Panopto. Use at your own risk!

Overview

Tired of taking classes and forgetting everything the next quarter? Tired of having to rewatch the same lecture over and over again only for a few key points? Scared of graduating and losing access to all this institutional knowledge? UW Panopto Downloader might be just what you're looking for!

Key Features

• Video Downloads: Log in to Panopto and batch download videos from course pages
• Audio Conversion: Convert videos to audio format for efficient storage and mobile listening
• Transcription: Generate accurate transcripts with timestamps either locally using OpenAI's Whisper or Google Cloud Speech-to-Text
• Metadata Management: Organize videos with tags, notes, and searchable transcripts
• Store Locally: Keep all your videos, audio, and transcripts in a structured format on your local machine with a SQLite database for metadata indexing

Installation

pip install uw-panopto-downloader

Requirements

• Python 3.12 (exactly 3.12 if you want any hope of using the Whisper model)
• FFmpeg installed on your system
• Google Cloud credentials (for transcription features)

Quick Start

Downloading Videos

uwpd download --url "https://uw.hosted.panopto.com/Panopto/Pages/Sessions/List.aspx#folderID=..."

Converting Videos to Audio

uwpd convert path/to/videos

Transcribing Audio with Google Cloud

uwpd cloud-transcribe path/to/audio.mp3 --language en-US

Transcribing Audio with Local Whisper Model

uwpd transcribe path/to/audio.mp3

Searching Transcripts

uwpd db search "neural networks" --transcript

Command Reference

The CLI is organized into the following main commands:

Core Commands

Command	Description
download	Download videos from UW Panopto
convert	Convert video files to audio format
cloud-transcribe	Transcribe audio files using Google Cloud Speech-to-Text
transcribe	Transcribe audio files using local Whisper model
config	View or update configuration
version	Show the current version

Database Commands

Command	Description
db list	List videos in the database
db search	Search videos in the database
db info	Display detailed information about a video
db tag	Add or remove a tag for a video
db tags	List all available tags with usage count
db delete	Delete a video from the database
db note	Add or update a note for a video
db stats	Show database statistics
db migrate	Migrate existing files to the database

Detailed Usage

Downloading Videos

uwpd download --url "https://uw.hosted.panopto.com/Panopto/Pages/Sessions/List.aspx#folderID=..." \
              --output ~/Videos/CSE142 \
              --workers 3 \
              --headless

This command will:

1. Open a browser window for you to log in
2. After login, automatically extract available videos
3. Prompt confirmation before downloading
4. Use 3 concurrent downloads for speed
5. Store metadata in the local database

Google Cloud Speech-to-Text Transcription

uwpd cloud-transcribe ~/Videos/lecture.mp3 \
                     --output ~/Transcripts/lecture.txt \
                     --language en-US \
                     --credentials path/to/google-credentials.json

Features:

• Automatically splits long audio into chunks separated by silence for better accuracy
• Produces timestamped transcripts for easy reference
• Stores transcripts in searchable database

Configuration

Configuration is stored in ~/.uw-panopto-downloader/config.json. You can update it using:

uwpd config --download-dir ~/Videos \
            --max-workers 4 \
            --headless True \
            --audio-bitrate 192k

Storage Management

The tool uses a structured storage approach:

• Content is stored in ~/.uw-panopto-downloader/content/
• Subdirectories for videos, audio, and transcripts
• Symlinks created to original locations when desired
• SQLite database tracks metadata at ~/.uw-panopto-downloader/metadata.db

Google Cloud Integration

This project leverages Google Cloud Speech-to-Text for high-quality transcription.

For optimal transcription quality, the tool:

1. Automatically converts audio to mono
2. Splits audio at natural silence points
3. Uses word-level time offsets for timestamps
4. Assembles chunks into a cohesive transcript

Development

To contribute to this project:

git clone https://github.com/elimelt/uw-panopto-downloader.git
cd uw-panopto-downloader
pip install -e ".[dev]"

Bumping Version

Use the provided script to bump the version before pushing changes/merging PRs (only for maintainers):

./bump-version.sh patch  # or minor or major

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Created by Elijah Melton ([e-mail])
• Submitted for SWECCathon 2025!