The Metadata Report

I've written in the past about photographic metadata on web sites. There had been some grumbling in copyright circles about how metadata was routinely stripped from photographs, especially on news sites, denying a credit to the photographer.

News outlets are excused from a requirement to credit in the UK's moral rights legislation. This omission, which dates from the 1988 act, clearly predates the web and comes from an era where to give a credit would mean taking up precious page real estate (although this wasn't the only reason). In an online electronic world this should no longer be a valid reason since metadata can always be included in an image. I should note that the Guardian (if not others) do find space to credit photographers ... but what about web sites?

I thought it would be interesting to check out a few of the major news web sites to see whether any creator information was encoded in the images. I realise this is a straw poll and won't apply to absolutely every image on a site.

• The Times are including their URL as source metadata but nothing else. They also include a text credit including one for the photographer.
• The BBC news web site includes a creator metadata item as well as a burned-in credit.
• The Guardian gives both text credits and includes a fair amount of useful metadata, although it's in a Photoshop namespace rather than the expected places so you may have to look at the raw XML to find it.
• The Express occasionally has burned-in credits but I couldn't find any metadata.
• I couldn't find any metadata in the images I checked on The Mirror web site.
• Briefly venturing overseas ... Reuters and the New York Times do give good text credits as, usually, does the South China Morning Post but I could find no metadata in the photos I checked.
• Finally the Daily Mail. Many of their online images not only include creator metadata but can also include a substantial amount more. This is on top of always having a visible credit 'burned' into one corner of the image. One image, illustrating a story about a child being taken away by social services, included a legal warning that the family must not be named and that the case was subject to court proceedings ... in the metadata. A front page story today about the raising of the Costa Concordia featured a large image from Reuters, with again a substantial amount of metadata which I assume originated with Reuters. Unfortunately not all the images on the web site (that I looked at) include metadata but I have to give the Mail credit for including such extensive metadata when they do include it.

So on the basis of this straw poll I believe that, in most cases, news web sites are finding space to credit the sources of their images on the page but few seem to realise the usefulness and importance of metadata. The Times makes use of it to point out that the image came from their web site, which is a variation on a theme, and I have elsewhere seen source or creator metadata that only said something like 'other' or 'record company' and agencies are mentioned far more than the photographers. So on the whole there is room for improvement.

We should all try to include metadata in images published on the web. I certainly try (and usually remember). I've looked at news here simply because of the moral rights exemption (which I'm glad to see most are ignoring) but if your web site isn't news then you may well have no excuse.

Should we worry about metadata?

What is metadata? One way to think about it is as data about data. Metadata is to your data file as the contents, index and 'back of the title page' are to a book.

On the web there is metadata provision in every web page, in the form of the meta tags in the header of the html file. Here we can put key words, a description of the page, and a variety of other pieces of information that, to be honest, we often don't include. It is also possible to include metadata in image files, including JPEGs. When you take a photo with a digital camera, the camera will automatically include metadata about the settings when the shot was taken, and in some cameras you can set that to include the photographer's name and copyright information.

The use of metadata in online images is becoming increasingly contentious. One of the problems being that in some cases any metadata, including the data that might identify the copyright in the image, is being removed, by accident or by design. Many in the photographic community are worried that widespread posting of images on websites, especially social media sites, risks creating so-called orphan works; where the author is either not known or cannot be traced. This is because sometimes the internet seems to be viewed as a copyright-free zone. I particularly like the copyright notice on the middle photograph.

So, what should we be doing if we publish images on our web sites?

1. Make sure published metadata includes attribution data (ie who took the photo) and the URL of your web site
2. If you allow user uploading of images be sure to include any metadata in the original file
3. If you resize images make sure any metadata is carried over
4. Don't use images without permission unless you know they are out of copyright

There's a great online tool which can show you metadata for a bewildering range of file types, including the common image formats. It was written by Jeffrey Friedl in 2006 and you can find out about it on his blog and try it out for yourself.

[PS: Jeffrey Friedl wrote the O'Reilly book Mastering Regular Expressions. An essential programmer's read.]