Technically speaking, the term "metadata" means "data about data", or simply information that describes other information. When it comes to digital pictures, typical metadata includes:
Date taken - the date the picture was captured (typically by a camera).
Caption - a user-entered descriptive word, sentence, or phase about the picture.
Keywords - separate descriptive words or properties about the picture.
Author/Creator - Identifies the person who took the picture.
Owner - Identifies the person who owns the picture. It's important to note that this is not always the Creator. Also note that. although a picture typically has only one Creator, it can have many Owners.
There are many, many other examples of picture metadata. Also, although this post deals specifically with pictures, this discussion easily extends to all other types of files (digital assets).
Still further, some metadata is "static" and some is "dynamic". Static metadata are items that don't change when the picture is edited, saved, modified, etc. The examples shown above are largely static metadata. Other metadata, like content IDs generated by some hash of the image data, or simply the "date modified" flag, will generally change every time the file is changed somehow. Or at least they *should* change.
Confused yet? It gets better. Although many "standards" exist for metadata, very few software programs completely adhere to them. That's because they all depend upon the goodwill of the software vendor themselves, who quite frankly have bigger fish to fry than fret over some seemingly unimportant adherance to a spec that will not generate a dime more revenue in the near-term. Thus the problem: the world is becoming littered with mountains of digital image files that have illegal, non-compliant, erroneous, or simply no metadata at all.
"So what"" you say? Well, don't say that to the commercial stock photographers. They clearly want to keep track of the very pictures that form the basis for their livelihood. But what about everybody else? Well, you tell me. Go to your computer, and (if you can figure out how) do a search of all your hard drives for .JPG files and sort them by file size. I wagering what you will find is exactly what I find on my computer. Thousands of files of course. But also, many many files with exactly the same size. Now THAT's odd. Why is that?? I'll tell you why...these files are duplicates. Copies of the same image, over and over again, in the same or different directories, partitions, or separate hard drives. They are there because you have NO IDEA how to get rid of them! An effective metadata management system would make this a thing of the past. Further, wouldn't it be nice to know WHICH version of a picture you had? Ever edit a picture, but want to save a backup just in case something went wrong? But you are intentionally creating duplicates! IF you had the proper metadata in these files, the system could automatically manage all this crap for you.
Bottom line. Effective metadata management can make the tangle of duplicates, backups, different versions, different renditions, etc. simple and automatic.
So why don't we have this? If everyone realizes this is a problem, why hasn't something happened? Well, it's simply because doing this right is a heck of a lot of work, requires a lot of collaborative effort among several companies, will take years to implement fully, and ...here is the big one... is is not obvious to most company managers how to make money from the investment! No one wants to pay for it unless they are going to get a big piece of the pie.
Actually...it's worse than that. Current metadata schemes are "fragile". That is, it takes a LOT of hard work to create, and yet one non-compliant program or nasty user can destroy all the good intentions. Since most metadata systems rely upon binary or text data embedded inside the file itself, anybody can write (or use) a program to remove, modify, edit, or erase the metadata that has been placed there. So what good is working hard to put my "Creator" or other metadata in my digital files, if once they leave my computer, any schmuk can simply change the info and make "my" picture his??
There must be a better way. One possibility is embedding a unique identifier INSIDE the IMAGE DATA itself. Companies such as Digimarc and others have been touting this type of digital watermarking for years. Supposedly it survives even edits, so someone could take my picture, crop it, apply "autofix" image processing, color balance, etc., to it, and the watermark would remain.
IF we had a standard system that linked this digital watermark to all the other relevant metadata for an image then we just might be able to claw our way out of this digital mess about to be foisted upon us by our own technology.
NOW...ALL WE NEED IS SOMEONE TO READ THIS BLOG AND DO SOMETHING...!
Interested? Here are some relevant links:
Adobe XMP metadata format
Stock Artists Alliance (SAA)
International Press Telecommunications Council (IPTC)
Dublin Core
Universal Photographic Digital Imaging Guidelines (UPDIG)
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment