503 EDITING FOR ONLINE
September 19, 2002
The DC/South chapter of the American Society of Picture Professionals (ASPP) conducted a
program in Washington on September 12th entitled "Big Archives, Big Decisions: a Digital
Dilemma". Panelists included: Merry Foresta from the Smithsonian
Institution; Eileen Flanagan from Corbis, Maura Mulvihill from National Geographic, Kit
Peterson from the Library of Congress, Steve Puglia with the U.S. National Archives, Amy
Pastan and independent picture researcher and Louis Plummer from PhotoAssist, a leading
picture research agency.
It was clear from the discussion that major efforts are being made to build some huge
digital archives. But, the prime inhibiting factor is cost. Paglia said his research has
shown that on average one should budget $20 per image to cover the scanning, keywording
and other administrative costs of getting the image where it can be digitally searched
along with a usable file for online delivery. He emphasized that this is an average.
Depending on the size of the deliverable file and other factors the cost is often more.
Another disturbing fact in his presentation is that digital archives we are creating
today may be good for no more than ten years. At that point, due to future technology
advances it is likely that we will have to re-do much of the work done previously.
It was clear from the discussion that the first resource most image buyers use when
searching for images is an online database. In many cases they use online resources
Considering that most of buyers attending this meeting either were directly employed by
book publishers, researchers who work for book publishers, or those who work at
developing various types of photographic exhibits it surprised some that they do so much
of their research online.
Traditionally, these people have had the most time to choose images for their projects.
They have been willing to wait for wait for photographers or stock agencies to pull
images from their analog files and submit film for consideration. Now it seems that
these image buyers no longer have the time to do the kind of exhaustive research they
did in the past. As the lead time on projects has shortened digital search has become
their primary solution.
But, the buyer expect that every image that was available to them in analog format in
the past should now be available for immediate search and delivery online. While all the
image suppliers listed above are trying to rapidly increase the number of images they
offer online, no one seems to think that they will eventually scan all the images they
have in their collections.
The most difficult decision is how to determine which images to scan and how many.
Foresta of the Smithsonian pointed out that they have a collection of historic bridge
photos and at one point they were trying to figure out which images to digitize. As a
visual person Foresta chose certain images for their artistic merits. The curator of the
collection, an engineer, looked at Foresta's selection and said, "Those are terrible
bridges," and then chose images that from his point of view were much more important.
The images the engineer selected were not very attractive and most had little artistic
merit from Foresta's point of view. Ideally both selections should be scanned but costs
may limit what can be done.
This little story clearly points up one of the major dilemmas of the industry. With any
set of images, on almost any subject matter, there is no universal agreement as to which
images are most useful or valuable. Editors must clearly understand the needs of the
eventual users. But the users change. Their tastes change. Their needs change. History
has shown that it is often very difficult to anticipate many of the uses that will
eventually be made of images placed in a stock photo archive. In the past, this problem
was solved to a degree by storing a very broad selection of everything produced.
Later, when someone would have a unique need not anticipated when the image was
originally shot, it was usually possible to go back into the file and with some effort
find an image that would fit the new need.
Now the digitization cost -- the $20 or more per image -- is forcing everyone to limit
the number of images they digitize. These editing decisions will limit the possibility
of finding the right image in the future. Images will not be available to fill the
unanticipated uses, or uses of subject that are not in high demand.
At present, if an image is not scanned it has very little chance of ever being seen by
an editor. In a very few years the chances of finding non-scanned images will be even
Many customers would like for the image suppliers to scan everything, but that will not
happen because it is cost prohibitive. None of the speakers seemed to believe that they
would ever have the resources to scan all the images in their historic collections,
despite the fact that many on the panel work for organizations that are supported by
federal funding. At one point in the meeting there was a round of applause for the U.S.
taxpayer who are currently paying for much of this work.
In discussions after the meeting no one on the panel seems to think it likely that there
would ever be a way to generate enough revenue from the images that are being digitized
to offset the cost of digitization.
Eileen Flanagan of Corbis showed pictures of the Corbis Film Preservation Facility in
Pennsylvania. It is clear that heroic efforts have been made to protect and preserve the
prints, negatives and glass plates that make up the Bettman Archives and the UPI
collection. The images were very carefully moved from their previous location in New
York to the new temperature and humidity controlled facility. In the near future most of
these files will be stored at -4 degrees Fahrenheit which should greatly extend their
life and provide time to digitize many of them.
About 225,000 of the close to 11 million images are now available online. Scanning of
the others will continue, but it is not clear how many of the total will be scanned.
There are researchers at the site who will pull images on request, digitize them and
make them available to photo buyers. Thus, it will still be possible to get access to
any of the images in this collection, but that may not happen as quickly as some would
If the image a researcher wants happens to be in the -4 degree storage area it must
first be brought to room temperature before it can be scanned. That takes approximately
Considering the money already put into the building of this facility it is hard to
imagine that there will ever be enough sales of the images not yet scanned to cover the
costs to date, let alone future operating costs. Bill Gates should be commended for his
contribution to protecting the historical record, but it seems highly unlikely that this
facility will ever be a profit center for Corbis.
The challenge ahead is daunting for photographers and stock agencies that are trying to
profit from the licensing of images. In one sense they want to get as many images as
possible online so customers can see them. But at the same time they must tightly edit
the images in an effort to make sure that every one they select will be something a
customer wants to buy. This is very difficult, but if they invest $20 an image in large
numbers of images that never sell they may soon eat up all the profits from those that
Traditionally, no more than 2% to 3% of the images in any analog file sell in any given
year. Thus, for every image that sells there were 30 to 50 in the file that didn't sell.
If you spent $20 to digitize each image you would have a $600 to $1000 investment for
each image you licensed. The average license fee for advertising images is only around
$500. Clearly, you can't stay in business long at this rate. Once digitized an image may
have a useful life of many years and that can improve the economics a great deal.
However, the solution lies in careful editing that limit your digitization of
This is easy to say, but very difficult to do. If you stick to high demand subject
matter, and if you don't have much competition that might work. But in most of the high
demand subject areas there is heavy competition and no way to predict whether your image
or that of a competitor will sell. In addition, no single editor no matter how
experienced has a perfect understanding of the entire market. For that reason it is best
to have many editors looking at your work and to listen carefully to the judgement of
Some companies have a very tight and focused editing strategy. Others take a broader and
more eclectic approach. The only sure thing is that never again will we be able to
afford to store the very broad cross section of imagery that stock agencies maintained
in plastic sheets in file drawers in the 70's and 80's. All editing will be
significantly tighter than it used to be, but at this stage I'm not convinced that
anyone has discovered the perfect balance.