cine-reader

Cross-platform readers for Vision Research Phantom .cine files in Python and MATLAB.

GitHub repository:

• https://github.com/rverleur/cine-reader

This repository contains:

• a Python package, cine_reader
• a MATLAB reader, Matlab/Cine.m
• shared C sources and runtime unpack libraries for packed 10-bit payloads

The PyPI distribution name is cine-handler. The import package remains cine_reader.

Preferred once the package is published on PyPI:

pip install cine-handler

You can also install from a GitHub release, from the Git repo, or from a local checkout.

1. Open the Releases page for the repo.
2. Download the wheel for the release you want: cine_handler-X.Y.Z-py3-none-any.whl
3. Install it with:

pip install cine_handler-X.Y.Z-py3-none-any.whl

You can also install directly from a release asset URL:

pip install "https://github.com/rverleur/cine-reader/releases/download/vX.Y.Z/cine_handler-X.Y.Z-py3-none-any.whl"

You can install the source distribution from the release too:

pip install cine_handler-X.Y.Z.tar.gz

pip install "git+https://github.com/rverleur/cine-reader.git"

pip install .

For editable development installs:

pip install -e .

• Python package code lives under python/src/cine_reader/.
• Packed 10-bit decode prefers bundled native unpack libraries and falls back to a NumPy implementation if a native helper is unavailable.
• Raw color CFA/Bayer cines load as 2D sensor mosaics by default.
• Set debayer=True to debayer every loaded frame.
• Set remove_dead_pixels=True to repair dead pixels every loaded frame.
• red_pixels, green_pixels, and blue_pixels expose raw CFA sensor samples with NaN at non-matching color sites.

See python/README.md for Python usage and python/API.md for API details.

MATLAB files are kept in-repo and are not installed by the Python wheel.

Use the MATLAB reader directly from this checkout or from a source archive:

1. Clone or download the repository.
2. Add Matlab/ to your MATLAB path.
3. Build mex_unpack10bit_cached if you need packed 10-bit support on your local MATLAB installation.

See Matlab/README.md for MATLAB setup and usage.

.
├── pyproject.toml
├── README.md
├── python/
│   ├── README.md
│   ├── API.md
│   ├── examples/
│   ├── tests/
│   └── src/cine_reader/
├── Matlab/
│   ├── README.md
│   ├── API.md
│   ├── Cine.m
│   ├── C_Files/
│   └── private/
├── c_src/
├── sample_data/
└── docs/reference/Cine File Format.pdf

The Python wheel and the MATLAB repo include runtime unpack helpers for packed 10-bit CINE data:

• unpack_data_win32.dll
• unpack_data_win64.dll
• unpack_data_elf64.so
• unpack_data_arm64.dylib

If a matching native helper cannot be loaded, Python falls back to the NumPy decoder.

• Format reference: docs/reference/Cine File Format.pdf
• Local smoke-test sample path: sample_data/TrimmedCine.cine

The sample cine file is intentionally gitignored and not distributed in the repository.