raw2mono: RAW to Monochrome Conversion Scripts

This repository contains two Python scripts designed to convert RAW image files into monochrome (black & white) images. One script outputs a DNG file, while the other produces a TIFF file. Both scripts:

• Extract each color channel (R, G1, G2, B) from the Bayer pattern.
• Apply Rec.709 luminance coefficients to generate a single-channel (grayscale) image.
• Multiply the result by a small factor to utilize the 16-bit range effectively.
• Copy relevant metadata from the original RAW file using exiftool.

Note on Compatibility

• DNG Output: The resulting DNG files are recognized correctly in darktable and will allow automatic lens detection and lens corrections. However, Adobe software (Lightroom, Photoshop) currently does not recognize these monochrome DNG files properly.
• TIFF Output: The TIFF files are widely recognized, including by Adobe Lightroom, which will also auto-detect the lens for lens corrections.

Table of Contents

1. Overview of Scripts
2. Requirements
3. Installation
4. Usage
   • Converting to Monochrome DNG
   • Converting to Monochrome TIFF
5. How It Works
6. License

---

Overview of Scripts

1. raw2mono_dng.py

• Description: Converts RAW files to a single-channel (monochrome) DNG.
• Key Steps:
  1. Reads the RAW file using rawpy.
  2. Extracts color channels (R, G1, G2, B).
  3. Applies Rec.709 luminance coefficients to get a grayscale image.
  4. Multiplies by 4 and converts to a 16-bit array.
  5. Uses pidng to create a DNG container.
  6. Copies EXIF data from the original RAW using exiftool.

2. raw2mono_tiff.py

• Description: Converts RAW files to a single-channel (monochrome) TIFF.
• Key Steps:
  1. Reads the RAW file using rawpy.
  2. Extracts color channels (R, G1, G2, B).
  3. Applies Rec.709 luminance coefficients to get a grayscale image.
  4. Multiplies by 4 and converts to a 16-bit array.
  5. Writes the result as a TIFF using tifffile.
  6. Copies EXIF data from the original RAW using exiftool.

---

Requirements

Make sure you have the following installed:

• Python 3.6+ (tested on Python 3.10).
• pip for installing Python packages.
• rawpy (Python RAW image processing library).
• numpy.
• exiftool (external command-line tool for metadata).
  • Ensure exiftool is accessible in your system’s PATH.
• pidng (required only for the DNG script).
• tifffile (required only for the TIFF script).

You can install the Python dependencies via:

pip install rawpy numpy pidng tifffile

Note: If you do not need to convert to DNG, you may skip installing pidng. If you do not need TIFF output, you can skip tifffile.

---

Installation

1. Clone this repository or download the two script files (raw2mono_dng.py and raw2mono_tiff.py).

2. Install dependencies as described above.

3. Confirm exiftool is installed and available in your PATH. For example:

   exiftool -ver

   This should display the installed version of exiftool.

---

Usage

Both scripts use a straightforward command-line interface with --input and --output arguments.

Converting to Monochrome DNG

python raw2mono_dng.py --input path/to/input.raw --output path/to/output.dng

Parameters:

• --input: The path to your source RAW file.
• --output: The path (and filename) for the resulting DNG file.

Example:

python raw2mono_dng.py --input IMG_0001.NEF --output IMG_0001_monochrome.dng

Once completed, the script will:

1. Read the RAW file (IMG_0001.NEF).
2. Convert it to a single-channel grayscale image using the Rec.709 coefficients.
3. Save it as IMG_0001_monochrome.dng.
4. Copy metadata from the original file via exiftool.

Converting to Monochrome TIFF

python raw2mono_tiff.py --input path/to/input.raw --output path/to/output.tiff

Parameters:

• --input: The path to your source RAW file.
• --output: The path (and filename) for the resulting TIFF file.

Example:

python raw2mono_tiff.py --input IMG_0001.CR2 --output IMG_0001_monochrome.tiff

Once completed, the script will:

1. Read the RAW file (IMG_0001.CR2).
2. Convert it to a single-channel grayscale image using the Rec.709 coefficients.
3. Save it as IMG_0001_monochrome.tiff.
4. Copy metadata from the original file via exiftool.

---

How It Works

1. Reading RAW:

   • We use rawpy to decode the RAW file into a numpy array (raw_data).
   • This array represents the sensor’s Bayer pattern data.

2. Extracting Channels:

   • We slice the raw mosaic into the four color channels: R, G1, G2, and B.
   • For example, R = raw_data[0::2, 0::2], B = raw_data[1::2, 1::2], etc.

3. Grayscale Conversion:

   • We apply the Rec.709 luminance coefficients:
     [ L = R \times 0.2126 + \frac{(G1 + G2)}{2} \times 0.7152 + B \times 0.0722 ]
   • This produces a single-channel float array.

4. 16-Bit Scaling:

   • The float array is multiplied by 4 (arbitrary scaling to better use the 16-bit range).
   • The result is converted to 16-bit integers (np.uint16).

5. Writing to DNG or TIFF:

   • DNG: We use pidng to embed the single-channel data in a DNG file structure.
   • TIFF: We use tifffile to write out a single-channel grayscale TIFF (minisblack).

6. Metadata Copy:

   • Finally, we use exiftool to copy all tags from the original RAW into the new DNG or TIFF. This preserves capture time, camera make/model, and other relevant EXIF data.