RawRefinery

RawForge is an open-source command line application for raw image quality refinement and denoising.

Currently in alpha release.

Install from pip:

pip install rawforge

New: ONNX backend!

NVIDIA: For cuda (tested, must have CUDA installed):

pip install rawforge[cuda]

AMD and NVIDIA on Windows:

pip install rawforge[directml]

All others, including M1+ Macs (CoreML)

pip install rawforge[cpu]

For web:

pip install rawforge[web]

Useage is the same, but use the CLI "rawforgeonnx".

rawforgeonnx TreeNetDenoiseHeavy test.CR2 test_heavy.dng --cfa 

Example command line syntax:

rawforge TreeNetDenoiseHeavy test.CR2 test_heavy.dng --cfa 

usage: rawforge [-h] [--conditioning CONDITIONING] [--dims x0 x1 y0 y1] [--cfa] [--device DEVICE] [--disable_tqdm] [--tile_size TILE_SIZE] [--lumi LUMI] [--chroma CHROMA] model in_file out_file

A command line utility for processing raw images.

positional arguments:
  model                 The name of the model to use.
  in_file               The name of the file to open.
  out_file              The name of the file to save.

options:
  -h, --help            show this help message and exit
  --conditioning CONDITIONING
                        Conditioning array to feed model.
  --dims x0 x1 y0 y1    Optional crop dimensions.
  --cfa                 Save the image as a CFA image (default: False).
  --device DEVICE       Set device backend (cuda, cpu, mps).
  --disable_tqdm        Disable the progress bar.
  --tile_size TILE_SIZE
                        Set tile size. (default: 256)
  --lumi LUMI           Lumi noise (0-1).
  --chroma CHROMA       Chroma noise (0-1).

Acknowledgments

With thanks to:

Brummer, Benoit; De Vleeschouwer, Christophe. (2025). Raw Natural Image Noise Dataset. https://doi.org/10.14428/DVN/DEQCIM, Open Data @ UCLouvain, V1.

Chen, Liangyu; Chu, Xiaojie; Zhang, Xiangyu; Chen, Jianhao. (2022). NAFNet: Simple Baselines for Image Restoration. https://doi.org/10.48550/arXiv.2208.04677, arXiv, V1.