Meta Data

What’s in a word? 2005

“ Sometimes the credit is included in the caption field, sometimes there is no credit, sometimes there is no caption. The ID number is also sometimes changed on delivery of the high res. Really frustrating!

This cry from the heart is not a one-off. Although picture information can now be carried within the digital image file, a vast gulf has opened up between our technological capabilities and the reality on the ground.

The problem, put bluntly, is that no-one knows what information to put, where it  should go, or whether anyone is going to read it. The use of metadata, for that is what we are talking about, affects us all, but as Neil Turner, photographer and education coordinator for Editorial Photographers UK says “ If you ask most photographers about metadata they’ll say its something to do with web sites. Its something other people do. “

Talk about IPTC fields, and you will get a more positive response. The IPTC fields in Photoshop were originally set up by the newspaper industry to standardise the way information was carried in the digital file. Some of the IPTC fields are included in Photoshop under File/info, a fact which leads many people to refer to metadata as IPTC.

The XML standard format (Extensible Mark up Language) now enables a much wider range of information to be carried within digital files, and read by compatible software. Adobe has created the XMP platform (Extensible Metadata Platform) which allows users of Photoshop and other software to create custom data fields for their own use if they so choose. This has huge potential, but the standard fields in Photoshop File/info are the source of metadata for most people at the moment.

Open an image in Photoshop, go to File/ File info and you will find information fields describing the image and its provenance. It is confusing that different versions of Photoshop label some fields in different ways and group them on different panels, but the essential information like picture number, caption, credit and copyright notice is there in all versions. Most image viewing and management software now includes the ability to write information to the image file – and if it doesn’t then it should. And despite the confusion of all the different sets of data standards – IPTC, Dublin Core, Adobe XML fields, and other more specialised data sets – our key information is more or less mappable between them all.

But the metadata is not being consistently used, and researchers like Debbie Weatherby, quoted above, have to work their way around a messy situation in the best way they can. Picture researchers often add their own information to the image file, but as one says “ it is very time consuming and boring and I prefer to take images from sites that have the information already embedded in the low res files”

Maureen Elliott, Picture Editor at National Magazines title Good Housekeeping rarely looks at File/info. “ It’s often blank. We’re short of time, and if it means going to a field that’s mostly blank we don’t do it.”  But she would like to see the metadata used consistently “ Ideally we would open a page in File/info and everything would be on one page. It would be very useful if everyone were using the fields. At the moment we ask for credit information by phone or e-mail. If it were in the image, the subs would receive it with the layout.”

And here’s another point. Some picture libraries are meticulous about adding information to the image file for high res images, but leave it off the low res versions on the internet. There may have been an earlier reason for this- the metadata does add a little to the image file size -  but most people agree that with broadband  this is no longer a problem. Lack of information in the low res file is a big drawback for researchers who search across a number of libraries and download low res onto their desktops or onto a system like Capturepix. Either way, metadata plays an important part. Picture libraries are only beginning to catch on to this. Most do not populate metadata fields in low res images, and those that do often leave out the caption. The picture number quoted on the site is not always the same as the filename of the downloaded low res image. A recipe for confusion.

Many picture libraries dump all the metadata into the caption field. The reason for this is that many clients, particularly in the newspaper industry, ignore all other fields. Steffen Wedepohl, responsible for German rights at The Bridgeman Art Library says “ As clients want the key information at a glance and many image viewer programs offer only caption and keyword metadadata fields, we make sure that all important information is in the caption field. It’s too risky to put essential information in a field that a client can’t access for any reason.” Another problem for the Bridgeman is that the caption field, with only 32 characters, is too short for the multiple credits it needs.

Putting all relevant information in the caption field is sensible, but what about those clients who want to automate? The beauty of metadata is that it can be picked up by client software without the need to re-key or to cut and paste. A credit line embedded in the caption field will always have to be separated manually. Picture users agree that the information should be in the separated fields as well, for those that can make use of it.

It’s easy to enter data into image files, but who should do it?  Picture libraries and publishers have their own way of presenting data, so cannot fully rely on photographer input. What they require will depend on their workflow. But photographers have their own ideas about metadata too. Some of them see metadata as an important way of preserving their rights under copyright law.

The photographer can be credited in author, copyright and credit fields. Copyright can be stated in copyright and copyright status fields. It’s not immediately clear where best to put the information, and images with multiple copright are even more difficult to manage.

Do any of the metadata fields have any legal standing as copyright protection? Copyright lawyer Charlie Swan explains: “ Whether or not you put a credit doesn’t affect your protection, but from practical point of view it’s a warning which will make people less likely to help themselves. If you have a copyright notice on an image, it is more difficult for people to claim ignorance and it will affect the damages you can claim for infringement.”

The benefits of using metadata can only be realised if publishers have software systems which can read it. Jon Moore, Digital workflow Manager for Conde Nast indicates that things are at an early stage: “ We are trying to educate people that metadata even exists. We want editorial teams to understand they may have to put some data in upfront, but they will get benefits later.”

Without metadata, Moore says, a workflow system simply cannot work. So are all publishers investigating the options? “ Definitely. They are all looking at managing assets and that always has metadata implications.” He hopes that picture libraries will be able to supply images in a set of standard fields, which need to be decided upon by industry associations and their clients.

Mark Winterton, Intellectual Property manager at IPC believes automation is on its way. “We don’t use info embedded in images currently but we’re looking at updating our software. We’ll be looking at automatic processes in a few years time. “

Winterton is aware of the difficulties in reaching common standards, but agrees that it is best to keep the format simple. He suggests that a good place to start would be for picture agencies make a proposal to publishers about standard fields, saying  “ the best thing as its essentially new is to keep it simple.”

Metadata standards are being discussed by BAPLA, and will be part of the Pic4Press guidelines in the coming year. There are also moves afoot, particularly in the US, to agree on standard licensing terms which could be carried in the image file. This is an endeavour which goes beyond the immediate need for basic agreement about fields for the three C’s - copyright credit and caption. But in the long run there is no doubt that this is the way the industry will go. Roger Ressmeyer, director of picture library Science Faction in the US has been considering the issue for PACA over the last year or so. He says:

“From PACA’s perspective it makes total sense for much more information to be stored within individual image files.  It would be nice if they would self-destruct when the license period has expired, or have the image fade over a few days, and phone home to licensee and licensor with a notification of license expiration.  All this will be possible in the fullness of time.”

The PLUS Coalition in the US is working on systems for automated licensing information within the image file. What do the users think of the idea? Given the current state of play researchers are understandably cautious.

“ I think it’s a wonderful idea” said one of them “but I think it’s a long way off.”

Published in Visuell 2005
(c) Sarah Saunders, Electric Lane