Lightroom: Change file names in LR 4, but still sees ALL pics from same card as NEW...

  • 3
  • Question
  • Updated 6 years ago
  • (Edited)
...even though I have selected "don't import suspected duplicates."

Example: I start with freshly formatted card, take 100 photos, import 100 photos, rename them on import to LR - THEN take another 100 pics ON SAME CARD, go to import, but LR library lists all 200 files on the card as new, even though only 100 are new.

Is this because of the renaming? Not a biggie, but it irks me...
Photo of Paul Repage

Paul Repage

  • 1 Post
  • 0 Reply Likes

Posted 6 years ago

  • 3
Photo of jdv

jdv, Champion

  • 728 Posts
  • 56 Reply Likes
Photos do not have an absolute identifying GUID in the metadata, so a heuristic is used to determine duplicates on import. The filename is one of those heuristics because it usually contains a reasonably (but not guaranteed) unique ID. (This is one of the reasons I always preserve the original file number in the photo names.)

Since false positives (mistakenly identifying a new file as already imported and greying it out) are not tolerated, we have to err on the side of letting the human (who, in this case has modified the primary mode of identification) make the decision.
Photo of Rob Cole

Rob Cole

  • 4831 Posts
  • 386 Reply Likes
Lr could do a better job of duplicate detection - no doubt about it.

I mean, if there is any uncertainty after considering original filename, and capture date/time, and camera serial number ..., one could always do a byte-by-byte comparison of image data to be 100% certain.

Note: the byte-by-byte comparison would only be necessary if the data length is exactly the same, which it almost never is when photographs are different.

Yeah, it'd slow things down if a card full of duplicates gets inserted, and the 100% certainty method is employed, but it could be an import option, which is normally unchecked, since 99.999999999% of the time, if the original filename is the same, and other critical metadata, and image data length is the same, then it's a duplicate - no need for 100% certainty.

Other software is certainly handling it better, or so I've heard...