Page 1 of 1

Duplicates that aren't

Posted: Sat Feb 11, 2017 3:26 pm
by geecee
I'm having a problem getting Duplicate Cleaner to accurately report duplicate images. As a test, I have a single folder with two images in it. The images are visually identical, but differ in their dpi. When I calculate their hash (md5 or sha-1) by hand I get different values. Also, the file sizes are slightly different (3.87 MB vs 3.86 MB). However, when I run an Image Mode search with Photo Similarity set to "Exact Match" (and no other options checked) the tool always tells me the files are duplicates. I see the same behavior no matter which Content Comparison Type I choose.

FWIW, the images are jpeg's and I'm running version 4.0.4 of Duplicate Cleaner.

Any ideas?

Re: Duplicates that aren't

Posted: Sat Feb 11, 2017 4:10 pm
by DigitalVolcano
The 'Exact match' in image mode hashes just the image pixels. This hash is different from the file hash which can differ due to image tags, etc.
DPI is just an image tag, and doesn't necessarily mean that the images are different on a pixel level.

If you scan them in Regular mode you should get two different hashes.