Adding and Editing an Image's metadata fields

[honorcb]

Lets have a discussion about the metadata that the Grid /BBC Images associates with images.

There seem to be several threads to follow,

Editing standard Metadata - do we allow editing of all metadata for people with correct permissions?
Adding standard Metadata
Adding new Metadata for example Accessibility Text for an image. ( Have been listening to discussion in IPTC photo metadata group, about adding embeded Accessibility Text to images metadata, The group lobbing for the addition is the Smithsonian. )

Format of Metadata being added , and how the "rules" appear to users

[paperboyo]

This is very sensible functionality. Would be great to have a list of fields that need to be editable (that currently aren’t) and the list of fields that should be possible to be added if they don’t exist. Would be great to have these list sorted by priority.

Some loose thoughts:

There was some work done towards it:
[adding] bbc#52
[editing] bbc#54

We should try to avoid a situation where something is added but not to all the places and not fully-fledged. Example of remaining cleanup work still to be done (after the initial #2667).

There are several considerations:

1. Data models and UI/UX helpers for different data types
   1. arrays (keywords, labels etc, see bottom of here)
   2. dates, date ranges, friendly date spellings etc
   3. structured fields (eg. location)
2. Functionality needs to be added to:
   1. Upload page
   2. Viewer (image page)
   3. Browser (search page)
3. We should share the code (and UI too) for at least pts. 2 ii & 2 iii. There is no reason for the differences (some fields are not visible in Browser that are in Viewer!).
4. We may want to control the fields shown via config (?) (eg. Archivists may want to see more fields?). We already have this for Show all metadata, maybe we need it for main UI too?

I think, sorting out remaining issues with #3083 could be a good intro in some of the complexities and challenges I tried to describe above.

[honorcb]

The initial list I was thinking about were the very basic information of an image, but can be missing

Title
Description
Location

[paperboyo]

Title

Editable when exists. Not possible to add on upload, in Viewer or in Browser. When exists – visible in all three places.

Description

Editable always. Possible to add in Viewer and Browser, if absent – filled with filename on upload; required (image invalid without). Visible in all three places.

Location

Only legacy (but still most popular with agencies) schema used. Non-editable, not possible to add. Visible only in Viewer. Needs structured data editing (not simple strings).

[honorcb]

link to Smithsonian. presentation

https://s3-us-west-1.amazonaws.com/groupsioattachments/25866/84860041/3865/0?AWSAccessKeyId=AKIAJECNKOVMCCU3ATNQ&Expires=1629719019&Signature=7TZIzqtYVbGE14yrjdxglUWuyT8%3D&response-content-disposition=inline%3B+filename%3D%2220210811_IPTC_Smithsonian-NMNH.pdf%22

[paperboyo]

Sadly, the link doesn’t work for me, @honorcb.

[honorcb]

it show a picture of an elephant , with this Alternative Text

A dark grey taxidermied African Elephant on display
in the National Museum of Natural History’s rotunda.
and the accessiblitity ". Extended Description"
A taxidermied male African Elephant on display
in the National Museum of Natural History’s
rotunda. The elephant is standing, facing to the right,
with his trunk extended. His front left foot is bent at
the knee, to mimic stepping forward. The skin is
dark grey, and the two tusks are a yellowed white.
The elephant is mounted on a raised platform made
of bands of green and mottled white marble that
has the words “Elephants in Danger: Your
Choices Make a Difference” embedded into the
stone. Behind the elephant, the Hope Diamond
gallery on the second floor and the Fossil Hall on the
first floor of the rotunda are visible. In front of each
of the exhibits hall entrances, there are two
columns of mottled white marble, matching the
granite on the elephant’s base.

[paperboyo]

So Alt text specifically is an interesting case. Some thoughts:

1. There is no standard schema for it
2. Agencies do not supply it, but their Description habits align quite closely (because the closer the description, the more chances of a sell), so in practice, Description makes quite a good Alt text most of the time
3. Example above is an amazing Alt text. Whether it’s too descriptive would depend on the audience and the use case. But, realistically, we won’t get as good Alt text written by agencies for all pics. Expecting our users to do that also seems a bit unrealistic (?). Maybe future image recognition will help (but will probably be able to describe the image on the fly with adjustable degree of verbosity, instead of relying on static metadata).

For now, I think integration with CMSes should, maybe, prefill Alt text with existing Description (ours prefills Caption, leaves Alt text empty, but allows quick copying of Caption into Alt text), but I’m not sure providing an additional (empty) field in Grid would be helpful.

cc @SiAdcock FYI

[paperboyo]

Editing Location done in #3703. Possible to add in Browser/Viewer, but not (visible) in Upload.