Can I batch-correct RAW exposure to compensate for changing Auto ISO?

Asked 8/10/2018

2 views

2 answers

0

I shot a series of Nikon D800 NEF files for photogrammetry with manual shutter speed and aperture, but I accidentally left Auto ISO enabled. As a result, the images have the same aperture and shutter speed but different ISO values. I want to normalize the files so they have consistent exposure across the set, ideally by reading each file’s metadata and applying an exposure correction based on its ISO relative to a chosen base ISO. Is there a practical way to batch do this programmatically or with image-editing tools?

Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0

Photography Stack Exchange contributor

7y ago

2 Answers

1

It depends on what programming language you have access to. It's easy to read exif data with MATLAB: https://blogs.mathworks.com/steve/2011/03/08/tips-for-reading-a-camera-raw-file-into-matlab/ but it is not free.

With Python or other languages you can 1) convert the nef files to 16 bit tiff; 2) use exiftools (eg. https://smarnach.github.io/pyexiftool/) to read the shutter speed/iso. 3) load the tiff files with a for loop and save the images with Pillow.

Originally by user71741. Source · Licensed CC BY-SA 4.0

user71741

7y ago

0

AI Answer

Generated from our catalog & community — verify before relying on it.

Yes. Because shutter speed and aperture stayed constant, the exposure differences come from the ISO changes, so batch-normalizing the files is a reasonable approach.

A practical workflow from the community answers is:

  1. Read each file’s EXIF metadata to get ISO.
  2. Convert the NEF RAW files to 16-bit TIFF for editing.
  3. Apply a matching brightness/exposure adjustment to each image in a script.

Tools mentioned:

  • Python with ExifTool bindings to read metadata, plus Pillow to load/save images.
  • MATLAB can also read RAW/metadata, but it isn’t free.
  • GIMP or Photoshop may be able to do it, though scripting/batch setup may take more effort.

Your EV idea is broadly correct: compensation should be based on the ratio between each image’s ISO and your chosen reference ISO, using a log2 relationship. In practice, working from RAW or 16-bit TIFF is preferable to preserve image quality and avoid banding.

For photogrammetry, consistent brightness across the set is usually more important than matching camera settings exactly, so this kind of batch correction is appropriate.

UniqueBot

AI

7y ago

Your Answer