nef2fits

Python package to convert from Nikon Electronic Format (RAW photos) to the astronomical Flexible Image Transport System.

Usage

Inside of python you just do:

from nef2fits import nef2fits
nef2fits("./path/to/file.nef")

And a "./path/to/file.fits" file would be created. The FITS file will have 4 images, one for each filter of the Bayer matrix of the NEF image (Red, Green 1, Green 2, Blue). Observation date, camera model, saturation value, exposition time and other EXIF tags are transferred.

Custom header values can be provided as a list of (key,value) or (key,value,comment) items.

The package provides a shell script to convert:

nef2fits convert file1.nef file2.nef file3.nef --prefix "converted" --header header.json

Where the file head.json could look like

[
    ["TELESCOP", "Meade LX200", "Model of the telescope used"],
    ["OBSERVER", "Your name here"],
    ["OBJECT", "M42"]
]

and those will be transferred to the three newly created FITS files, inside a new folder name "converted", specified by the --prefix option.

You can also "watch" a directory, and the program will convert any .nef files that it encounters into FITS, possibly in a recursive manner. Name changes are alse tracked, but deletion of the .nef files will not result into deletion of any FITS files.

The syntax is similar:

nef2fits watch ./images --prefix "converted" --header header.json

The watch command accepts the same options as the convert command.

Installation

For now, pip install . on this repo.

Dependencies

• astropy for the FITS handling
• rawpy to extract the numeric data of the image
• piexif to extract the image metadata
• watchdog to automatically detect new images and convert them.

Roadmap and ideas

• Watchdog script to automatically convert files on a directory.
• Warning if the max pixel value is near the saturation level.
• Publish to PyPI.
• Shell option to modify the object regex.
• Callable argument to extract header elements from the data itself, not only the NEF exif. For example, max, min, sky values...
• Option to automatically use a local header.json file if found on the same folder on the image.
• Option to separate the R, G1, G2 and B channels into different fit files rather than one file with 4 arrays.