What Type of Image Metadata Should I Add To My Scanned Photos?
Published:
After a week of scanning1, I’ve finally managed to scan 616 old family photos which stretch over decades. While I could just share them with my family, not having metadata would make a poor file management experience. Given that, I thought I’d make a project where I attempt to make a good interface for adding all that metadata. This is much harder than scanning the pictures, even if you ones one of the many online photo digitization services since you’ll still need to know what happened and when.
Before writing a bunch of code, I thought I’d figure out what data I wanted to capture in the first place and this is what I came up with. Thanks to exiftool for providing a list of EXIF, XMP and mwg (common) tag descriptions.
| Tag | Tag Type | Description |
|---|---|---|
DateTimeOriginal | mwg | When the photo was originally taken |
CreateDate | mwg | When the photo was digitized (for this I’ll use the day I scanned the photo) |
| GPS Tags/City, State and Country | exif/mwg | Where the photo was taken. It’ll just be latitude and longitude and the city if I can find it. |
FileSource | exif | Where the file came from, probably 2 (reflection print scanner) for all of them |
HostComputer | exif | The computer it came from, probably Farai's PC |
Description | mwg | Some notes on the image. iCloud doesn’t include this in the images for some reason. |
Orientation | mwg | Don’t know if it’s needed but it might be nice to have. |
Keywords | mwg | For tagging |
Collections | mwg | For grouping similar groups of photos |
Regions/apple_fi | mwg | For face tagging. It doesn’t sound hard to have though I’m unsure of the benefit of doing so since my phone will scoop them up anyway. |
These seem like the most sensible tags to collect. While compiling this list, I came across the Metadata Working Group which aims to improve metadata interoperability between media formats. The MWG didn’t write this but a comment I found on exiftool’s website goes like this:
A note to software developers: Re-inventing your own private tags instead of using the equivalent tags in standard XMP namespaces defeats one of the most valuable features of metadata: interoperability. Your applications mumble to themselves instead of speaking out for the rest of the world to hear.
Not that I was going to roll my own tags, but it’s great to have a solid starting off point. They also have a document on how you should work with metadata [ARCHIVE PDF 1.63MB].
As a sidenote, it’s scary just how much metadata is in all these images. Sure websites are smarter to strip off the particularly sensitive bits of metadata before posting them onto the web, but uploading an image gives a platform a lot to work with.
Which could have been less had I used another scanner which was smart enough to identify separate images. ↩︎