Frequently Asked Questions

General Questions

What is Starship Analyzer?

Starship Analyzer is a Python toolkit for extracting, analyzing, and visualizing telemetry data from SpaceX Starship launch webcasts using computer vision and optical character recognition.

What data can it extract from videos?

The tool can extract:

• Speed and altitude measurements
• Engine ignition status for all Raptor engines
• Fuel level data for LOX and CH4 tanks
• Timestamps and mission events
• And calculate derived metrics like acceleration and G-forces

Is this an official SpaceX tool?

No, Starship Analyzer is an independent, open-source project and is not affiliated with or endorsed by SpaceX.

Installation Questions

What are the minimum system requirements?

• Python 3.8 or higher
• 4GB RAM (8GB+ recommended)
• Approximately 500MB disk space plus space for videos
• Any modern operating system (Windows, macOS, Linux)

Do I need a GPU to run Starship Analyzer?

No, a GPU is not required but is highly recommended for optimal performance. With a CUDA-compatible NVIDIA GPU, OCR processing can be 5-10x faster.

I'm getting errors during installation. What should I do?

Check the Troubleshooting page for common installation issues. If your issue isn't covered, create a GitHub issue with details about the error and your system configuration.

Usage Questions

Which Starship flights can I analyze?

You can analyze any SpaceX webcast that includes telemetry overlays. The tool is optimized for official SpaceX webcasts but may work with other sources if they use similar telemetry displays.

How accurate is the data extraction?

The accuracy depends on several factors:

• Video quality (higher resolution = better results)
• Telemetry overlay visibility
• Video frame rate
• Engine visibility in the frame

Generally, the tool achieves 90-95% accuracy for telemetry data and 95-99% accuracy for engine status detection under optimal conditions.

Can I compare data from different flights?

Yes! One of the key features is the ability to compare performance metrics across different Starship test flights.

How long does processing take?

Processing time depends on:

• Video length and resolution
• Your hardware specifications
• Selected processing options

As a rough estimate:

• With GPU: ~1-2 minutes per minute of HD video
• Without GPU: ~5-10 minutes per minute of HD video

Where are the results saved?

By default, all results are saved in the results directory in the project folder. This includes:

• CSV files with raw extracted data
• PNG/SVG plots and visualizations
• Log files for troubleshooting

Technical Questions

How does the engine detection work?

The engine detection system:

1. Uses predefined pixel coordinates for each engine location
2. Analyzes the brightness at these coordinates
3. Applies a threshold to determine if an engine is firing
4. Tracks changes in engine status over time

What OCR technology does the tool use?

Starship Analyzer uses EasyOCR, which is built on PyTorch. This provides reliable text recognition for extracting numeric data from the telemetry overlay.

Can I use the tool programmatically in my own Python scripts?

Yes! The application is modular and its components can be imported and used in other Python scripts. See the API Documentation for more details.

How can I improve processing performance?

To improve performance:

• Use GPU acceleration if available
• Process frames at intervals rather than every frame
• Reduce video resolution if accuracy isn't critical
• Increase batch size if you have sufficient RAM
• Trim videos to only include relevant portions

Does the tool transmit any data to external servers?

No, Starship Analyzer operates entirely locally. It does not transmit any data to external servers or share information with third parties. All processing happens on your own computer.

Contribution Questions

How can I contribute to the project?

See the Contributing page for detailed information on how to contribute. You can help with:

• Code contributions
• Bug reports
• Feature requests
• Documentation improvements
• Testing and feedback

How do I report a bug?

You can report bugs by creating an issue on the GitHub repository with detailed information about the problem, steps to reproduce, and your system configuration.