How should I store "description" metadata for maximum consistency, visibility, and portability?

Question

I’m embarking on a project to scan our family’s old collection of slides. My plan had been to write descriptive information about each slide (e.g., when/where/why it was taken, who's in it, etc.) to the three description fields (Exif Image-Description, IIM Caption-Abstract, XMP dc:Description) in a bid to (1) ensure consistency between these fields and (2) make the information visible via as many programs as possible.

Carl Siebert's excellent article summarizes the varying ways different programs read and write the description fields (note particularly how Windows File Explorer reads only the Exif field), and Capture One's help page indicates that the IIM Caption-Abstract field is limited to 32 characters.

I don’t anticipate writing giant amounts of information per slide, but length restrictions are troubling, especially if they're as small as 32 characters. I’m concerned that if I write a description longer than 32 characters and store it in the three description fields, it will be truncated in the IPTC Caption-Abstract, but not in the Exif or XMP description fields, thus giving rise to the inconsistency writing to all three fields is intended to prevent.

I’d be grateful for advice regarding the best way to store general free-form descriptive metadata for scanned images. My slide scans will be distributed to family members with varying levels of technical sophistication using a variety of devices and software. I want to maximize the likelihood that the metadata I store will be visible to everybody, both now and in the future.

Answer 1

I would suggest not worrying about the details and use a Digital Asset Management (DAM) such as Digikam, Darktable, (both are free) or Adobe Lightroom (paid). Adobe Bridge is also free though it's not a full fledged DAM, but it will give an easy GUI to enter metadata.

All of these will embed a full range of metadata in the appropriate locations in EXIF, IPTC IIM/Legacy, and XMP (IPTC Core/Ext) so you don't have to worry about the behind the scene details.

Also, with regards to character limitations, most programs ignore the limitations defined by the IPTC IIM/Legacy specs in writing and reading. EXIF and XMP have no character limits for the most part except in a few very specific tags which aren't widely used.

Answer 2

My first inclination would be to store detailed metadata externally and not fool with Exif because Exif metadata is hard to aggregate across multiple images.

What I mean is getting the descriptions of fifteen pictures at once does not have good tooling.

Instead I would be inclined to add a unique index to the file name of each picture and a separate document listing each picture by file name and description…maybe in a spreadsheet or even a text file.

Then it would be possible to search the descriptions quickly and use the file name to find the picture.

This avoids the limitations of Exif fields, makes it obvious that the pictures are described already, and makes searching easy…searching through the Exif of thousands of pictures will be clunky for most people.

A separate document can even be printed and human readable. It need only contain the relevant information.

Answer 3

I agree with @StarGeek. Just let whatever program you use manage the behind-the-scenes metadata for you.

If you are directly writing to specific fields, pick one that works with whatever programs you use. Keeping multiple fields synchronized is a pain. Even if you are writing a program or script to keep the fields synchronized, you still have to pick one to propagate.

Which one to pick? It doesn't matter as long as it works with the program you use. So much better if you just let the program manage it for you. If you really can't decide, use XMP. If you think something else is better, use that instead.

Some basic information can be stored in folder and file names.

My problem is that I want to store metadata about a scanned image in the image file such that it is likely to be viewable by people with different technical capabilities using a variety of different program over which I have no control.

Just not going to happen with anything short of directly printing captions over images. Even then, you can't force people to read.

I assume that those people may propagate the image to other people via email or text message or who-knows-what, and I want the metadata to remain with the image (unless somebody or some program decides to expressly remove it).

Many image sharing programs/apps remove metadata, regardless of what the user wants. To prevent this from happening, you would need to maintain and share a centralized repository with known behavior when people download the photos.