Lapsora

Self-hosted RTSP timelapse web application. Capture frames from RTSP cameras on a schedule, generate timelapse videos automatically, and manage everything through a modern web UI.

Note: This is a vibecoded project — built iteratively with AI assistance. It works well for its intended purpose, but may have rough edges. Bug reports and feature requests are very welcome via GitHub Issues.

Screenshots

Dashboard	Statistics

Settings	Streams

Features

Stream Management

• Multiple source types: RTSP, go2rtc, HTTP snapshot (single JPEG endpoint), and HTTP MJPEG (multipart stream)
• HTTP source auth: basic / digest / bearer / custom-header (covers generic IP-cam CGI, OctoPrint, Frigate, UniFi, Reolink/Amcrest, public webcams)
• Encrypted credential storage (Fernet encryption)
• Stream health monitoring with configurable failure thresholds and live reachability badges
• Live preview and go2rtc WebSocket live view

Capture System

• Scheduled frame capture at configurable intervals
• HDR / synthetic exposure bracketing with Mertens fusion
• Sun-based scheduling (sunrise/sunset offsets via location settings)
• IR-only frame filter — keep only true greyscale/IR night frames (chroma-spread detection), trimming the color frames around dusk/dawn
• Weather data integration (OpenWeatherMap)
• Per-frame Home Assistant sensor snapshotting (for sensor overlays)
• Quality control with frame corruption detection
• Capture gap alerting

Timelapse Generation

• Output formats: MP4, WebM, GIF
• Deflicker processing (luminance curve smoothing)
• Overlay options: timestamps, weather data, multicolor weather icons (WMO-mapped), Home Assistant sensor values, logo/watermark, heatmap blending
• Per-timelapse overlay configuration (style, position, size, opacity)
• Motion blur effect
• Codec selection: H.264, H.265, VP8, VP9
• Quality presets (low / medium / high / ultra)
• Generation queue with cancellation support
• Scheduled generation (daily, weekly, monthly, yearly presets or custom cron)

GPU Acceleration

• NVIDIA NVENC hardware encoding (H.264 / H.265)
• CuPy-accelerated compute for heatmap blending and deflicker

Integrations

• Home Assistant — pick sensors per profile, snapshot their values per frame, and render them as a styled overlay on timelapses
• PrusaLink — drive a capture profile for the duration of a 3D print: start capturing on PRINTING, and on FINISHED auto-generate a timelapse over the exact print window and notify
• go2rtc — restream/transcode source for exotic transports (RTMP/HLS/WebRTC/ONVIF)
• Live reachability badges for each integration in Settings

Notifications

• Apprise integration — supports ntfy, Discord, Telegram, Slack, email, and many more
• 13 configurable event types (capture failure, stream health, timelapse lifecycle, retention summary, low disk, capture gap, and PrusaLink print started/finished/failed)
• Real-time Server-Sent Events (SSE) for in-app notifications

Retention & Cleanup

• Cron-based cleanup schedules per profile
• Age-based frame and timelapse deletion
• Orphan file cleanup

Statistics Dashboard

• Storage usage trends over time
• Capture activity charts
• Per-profile storage breakdown
• Timelapse generation summaries

Profile Templates

• System and custom preset templates
• Apply templates to quickly create new capture profiles

Quick Start

Docker

# Clone the repository
git clone https://github.com/jumpingmushroom/lapsora.git
cd lapsora

# Configure environment
cp .env.example .env
# Edit .env and set a strong SECRET_KEY

# Start with Docker Compose
docker compose -f docker/docker-compose.yml up -d

Open http://localhost:8000 in your browser.

With GPU Support (NVIDIA)

docker compose -f docker/docker-compose.yml -f docker/docker-compose.gpu.yml up -d

Requires the NVIDIA Container Toolkit.

Configuration

Environment Variables

Set variables with the LAPSORA_ prefix (e.g., LAPSORA_SECRET_KEY) or in a .env file.

Variable	Default	Description
LAPSORA_SECRET_KEY	auto-generated	Fernet encryption key for stored credentials
LAPSORA_DATA_DIR	data	Directory for frames, timelapses, and database
LAPSORA_DATABASE_URL	sqlite:///data/lapsora.db	SQLite database path

In-App Settings

Configured through the web UI under Settings:

Settings are organized into tabs (General, Integrations, Notifications, Branding):

Setting	Description
Location	Latitude/longitude for sun-based scheduling and weather
Health Check	Failure thresholds and monitoring intervals
Notifications	Apprise URLs and per-event toggles
go2rtc	Server URL for go2rtc integration
Home Assistant	Base URL + long-lived access token for sensor overlays
PrusaLink	Printer host + API key for 3D-print timelapses
Branding	Upload a logo/watermark for timelapse overlays
Time Format	12-hour or 24-hour display
Capture Gap Alerting	Alert when captures are missed

Architecture

Single Docker container running:

• FastAPI — REST API backend serving the built frontend
• SvelteKit 5 (Svelte 5 runes) — Responsive dark-themed web UI with Tailwind CSS
• SQLite — Metadata and configuration storage
• APScheduler — Capture scheduling, timelapse generation, and cleanup jobs
• FFmpeg — RTSP frame capture and video encoding
• OpenCV — HDR processing, deflicker, heatmap generation

API Reference

All endpoints are under /api. Full CRUD is available for most resources.

Area	Endpoints	Description
System	GET /health, GET /storage, GET /system/info	Health, storage stats, GPU info
Generation Queue	GET /generations/active, GET /generations/queue, DELETE /generations/{id}	Monitor and cancel timelapse generation
Streams	CRUD /streams/, POST /streams/{id}/test, GET /streams/{id}/preview	Stream management, testing, preview
Profiles	CRUD /streams/{id}/profiles, POST /profiles/{id}/enable, POST /profiles/{id}/disable	Capture profile management
Captures	GET /profiles/{id}/captures, GET /captures/{id}/image, DELETE /captures/{id}	Frame management
Timelapses	GET /timelapses, POST /profiles/{id}/timelapses/generate, GET /timelapses/{id}/video	Timelapse management and generation
Timelapse Schedules	CRUD /timelapse-schedules/, POST /timelapse-schedules/{id}/trigger	Automated generation schedules
Cleanup Schedules	CRUD /cleanup-schedules/, POST /cleanup-schedules/{id}/trigger	Retention policy management
Profile Templates	CRUD /profile-templates/, POST /profile-templates/{id}/apply	Preset templates
Notifications	GET /notifications/, GET /notifications/stream	Notification history and SSE
Settings	`GET	PUT /settings/location, GET
Branding	`POST	GET
Statistics	GET /statistics/summary, GET /statistics/storage-trend, etc.	Dashboard data

Development Setup

Backend:

cd backend
python -m venv .venv
source .venv/bin/activate
pip install -e .
uvicorn app.main:app --reload

Frontend:

cd frontend
npm install
npm run dev

The frontend dev server proxies API requests to the backend at localhost:8000.

Contributing

See CONTRIBUTING.md for guidelines.

License

This project is licensed under the GNU Affero General Public License v3.0.