Manually specify that two images should be considered one asset

Thank you for your hard work and I look forward to really getting to grips with my photo library thanks to PhotoStructure.

Over the years I have accumulated multiple versions of some images through uploading certain ones to web services such as Google Photos or Facebook and then downloading them again later for backup purposes. In most cases these web services have stripped the metadata from the images, leaving them with no EXIF tags and only the downloaded date/time. In almost all cases the downloaded image is somewhat compressed and much lower in resolution than the original.

Obviously, PhotoStructure considers these re-downloaded backups to be separate assets from the original photos, where those exist. What I’d like to do is associate the backups with the original image asset so that they all show up together in the interface and I can pick the best version.

You might think I should just delete the re-downloaded files but, unfortunately, some of the original files have been lost. This means I can’t delete the whole Facebook or Google or Flickr folder without losing some images entirely.

Thanks for reading. I don’t think I’ve duplicated an existing request but I apologise if I have.


PhotoStructure should handle most cases where a downsampled image has been stripped of metadata: it relies on the image hash and dominant colors to find matching variations in this case.

Can you email me a pair of images (both the original and downsampled) so I can take a look? I may be able to assist the matching heuristics to make things “just work.”

I’ve thought a bit about how I can support “manual” deduping, as well: in other words, letting the user tell PhotoStructure that these 2 assets are actually the same (and the converse, that the current asset is actually incorrectly aggregating more than 1 photo or video). If anyone’s seen a slick UX for this task, please share!