How to handle scanned images?

I just had a user ask

I have many source photos scanned from slides and printed photos. These all have no meta data, and file creation dates far in the future of when the photo was taken. What should I do with these files to ensure PS can sort them correctly by date? Do I have to add meta data, or change the file creation time, or something else? Many thanks in advance.

How PhotoStructure pulls out captured-at

PhotoStructure has a bunch of heuristics to try to pull in a “captured-at” date for your files:

Solutions

Inferring the date via directory or file names

The easiest way to handle scans will probably to use folders that contain a “YYYY-MM-DD” string, like “Trip to Santa Barbara 2022-05-03”. If a file has a “YYYY-MM-DD” bit, that will override the date encoded in the parent directory. You can also use “YYYY-MM-DD HH:MM” if you have timestamps.

Adjusting date parsing via settings

There are a bunch of settings to support more “fuzzy” datestamp matching. See fuzzyDateParsing and fuzzyYearParsing for details.

Clearing metadata

If your scanning software includes the date that the image was scanned (which is never the correct time), that will “win” over any datestamp that PhotoStructure finds in the directory or filename.

You can use ExifTool to either remove these dates:

exiftool -AllDates=  path/to/image.jpg

(that’s a space after -AllDates=, which means “erase all dates found in the metadata”).

Fixing metadata

Or, if you prefer, rather than just erasing the metadata, you can update it to be correct:

exiftool `-AllDates=2018-02-18` path/to/image.jpg

which sets the date to February 18, 2018. ISO 8601 datestamps are accepted.

See also

I am scanning some old images and afterwards I may want to change dates and descriptions etc.
What Exif editor with a GUI would be recommended for Windows 10?

In addition where should information best go, Exif title, description IPTC tags or where ever?
I am confused.
Can anyone assist?

PhotoStructure in general tries its best to find metadata in standard tags, and knows about “synonym” tags (like captured-at tags).

Currently, the title, subject, heading, and description tags aren’t rendered in the interface. When I code that up (and add that to the search index), PhotoStructure will have 2 new StringArraySettings (one for “title”-ish tags, and one for “description”-ish tags). The first non-blank entry will “win”.

TL;DR: use what ever tag is the default for your application: we’ll be able to handle it correctly (either out of the box, or by tweaking a setting or two).