Taking Stock of the Web

'The researcher's life has changed a lot,' explains Kenn Rabin, a researcher in the San Francisco area. 'I used to be hired for, say, six weeks, to work on a film and see it from start to finish. Now my work...
July 1, 2000

‘The researcher’s life has changed a lot,’ explains Kenn Rabin, a researcher in the San Francisco area. ‘I used to be hired for, say, six weeks, to work on a film and see it from start to finish. Now my work is much more fragmented. I will very often get phone calls from a project that will say, ‘We want to hire you just for a couple of hours just to fill in the gaps.’ That’s sort of a frustrating thing for researchers.’

Rabin isn’t the only researcher who’s frustrated. Online access to footage houses has producers questioning the dollars they spend on research. Filmmakers now have a direct link to archives through their desktop, and if they can find the footage themselves it can save hundreds, even thousands, of dollars. But it’s a trade-off. By doing so, they give up the expertise of the researcher. RealScreen decided to examine what producers were getting in exchange.

We surveyed 75 websites to find out how simple the process was for producers who decide to take on the online research themselves. Although each website is as different as the company it represents, there are some key factors which determine success or failure for a producer on the hunt.

***Contact information***

There’s no substitute for human contact. At some point in the process, producers will have to talk to the people they’re getting the footage from. Says Graham Shirley, a researcher in Balgowlah, Australia: ‘If what I’m looking for can’t be found on an Internet database – or if an organization doesn’t yet have their collection on the Internet – I’ll either e-mail, fax or post my list of requirements to the archive or library that I know might be able to help me. Sometimes, I’ll rely on my older files and notes to re-visit a film or video. Frequently, I’ll rely on an Internet catalog listing. Quite often, I rely on the staff of an organization to suggest an item for me to view.’

>> Of the websites surveyed:

92% Had phone and fax numbers

84% Had e-mail address or links but of those 84% – 1 in 10 didn’t work

79% Had snail-mail address

37% Had a contact name (the name of a human being, rather than a general contact)

A majority, almost 91%, had a company synopsis or background, so that visitors get a better idea of who they are dealing with.

***Finding your footage***

A stock footage website is only as good as its search engine. It’s the interface that allows you to find the footage you’re looking for – everything else is just window dressing. A good search engine will allow a producer to do three things: Look for the footage they want; find it; and return the information in a useful form.

>> Of the websites examined:

8% Had neither search engine nor index

29% Only had a text listing of the footage on the site

63% Had a text search engine


On the sites that would allow text searches, we tested both for completeness of cataloguing and fuzzy logic.

Fuzzy logic allows a search engine to bend requests in order to return matches that might not necessarily fit the exact wording of the search. For example, would a search for ‘sunrise’ and ‘sun’ ‘rise’ return related results? Another example: what would happen if you looked for ‘u-boat’, ‘uboat’ and ‘u boat’? Fuzzy logic allows producers to find footage they’re looking for, even when they aren’t certain of spelling or proper syntax.

>> Of the search engines tested:

54% Used fuzzy logic consistently (i.e. Would find the same footage regardless of text.)

9% Used it occasionally

47% Did not use it (i.e. If you used the wrong word, you got different/no footage matches.)

To test the completeness of the cataloguing, we performed searches such as ‘united nations’, ‘u.n.’ and ‘u n’ – searches where there was an accepted synonym, or a short form in common use.

>> Of the search engines tested:

32% Returned consistent results for any of the word searches

17% Returned partial results, depending on which text was used

51% Found only the text searched for

Results returned:

A search engine is only as good as the information it uncovers. About two-thirds of the websites examined returned useful information.

>> Of the searches performed:

71% Returned some form of text description of the clip

60% Returned the date of the clip (for historical material)

68% Had a reference number in order to make ordering simple

63% Gave the length of the film clip

46% Noted what format the clip was archived on (video, 35mm, 16mm…)

>> Searches that returned images:

25% Returned sample images for the text search consistently

5% Returned sample images occasionally

70% Returned no sample images

Researchers offer their suggestions:

The research professionals had opinions on how to improve the capability of the footage search engines.

Lewanne Jones of Autonomedia in New York lists her biggest annoyance as the: ‘Inability to search by date or date range, and the lack of clarification on the site about what the search engine is actually searching.’

Steve Bergson of Prismrays in Kent, U.K., says he is slowed down because footage search engines don’t leave you ‘a trail to show you hits you’ve already made when trying different search approaches.’

Amy Lennie of The Rights Company in Toronto, pointed to the Producer’s Library Service as a good example of how not to inundate your visitors with information: ‘When you enter a search term, it gives you a list of one-line descriptors for each clip. If the descriptor interests you, you click on it. It’s very easy to pick out which clips you want more info on without drowning you in words right off the bat. When you want the detailed descriptor, it gives you a full description of every shot on the reel, the year produced, what format the master is on…’

***Film Clips***

As bandwidth increases, more and more footage archives are offering visitors the chance to actually see what they are buying through full-motion film clips. Although the technology is relatively new, many archives are already getting on board.

>> Of the websites visited:

40% Had some form of playable film clips

6% Had film clips for the majority of searches conducted

>> Of the clips discovered:

31% Loaded quickly (less than a 20:1 ratio – length of film clip to loading time)

35% Played in the web browser, or offered a link to the proper player

***Completing the transaction***

>> Costs: Of the websites tested

34% Had a listing of footage costs

31% Listed service fees (dubbing, sample reels, etc.)

34% Explained extra fees (shipping and handling, etc.)

What makes the Internet efficient is the fact that it allows surfers to begin and end transactions from the comfort of their office or workstation.

Once the searches are complete, and you’ve found the images you were after, the unfortunate part follows – paying the bill. While the vast majority of transactions will still involve a follow-up to the site visit, few footage sites offered producers the chance at one-stop shopping.

Remarkably, only 6% of the websites examined gave producers any indication of the rights information attached to the clips they were interested in.

>> Transactions: On the sites examined

2% Allowed you to do the entire transaction online

17% Allowed you to submit your order online, after which you would be contacted

15% Only allowed you to send your personal information, so that you could be contacted

***A little help along the way***

Navigation tools are important on any website. On archive sites, which offer both complex information searches and online retail, they are especially critical.

Site maps – stand-alone pages which allow visitors to jump to any other page on the website – are especially useful, and are standard on most retail websites. If you can’t find it any other way, a site map allows users to instantly jump to whatever they need.

>> Of the footage sites visited:

9% Had some form of site map

5% Had a complete, easy-to-navigate site map

86% Had no site map

Along with a site map, most websites will offer visitors online help, or FAQ (frequently asked questions) files.

>> Of the sites Surveyed:

23% Had a complete online help section or faq

18% Had some help/faq

About 34% of the sites examined provided links to other websites (film forums, footage/government agencies, etc.) Only 6% of all the websites visited offered information in more than one language.

***Response Time***

In order to test how fast the online companies returned e-mails, or if they returned them at all, RealScreen sent each online library an e-mail asking them how much of their archive is accessible through their website.

>> Of those sites surveyed:

84% Had e-mail address or link, but 9% bounced back

14% Responded same day

32% Responded the next day

10% Responded in two days

2% Responded in three days

2% Responded in 7 days

40% Did not respond after 10 days

>> The Fastest responses came from:

Associated Press Television News: Under 1 hour

National Film Board of Canada: 2 hours

CNN Imagesource: 3 hours

Vanderbilt Television News archive: 3 hours

Oddball Film and Video: 4 hours

Hardy Jones/Julia Whitty Productions Library: 6 hours

Of those that did not bounce back:

No websites offered a definitive indication as to the extent of the archive’s online content. This rankles researchers like Rosemary Rotondi, who operates out of New York. ‘I believe each site must inform the researchers of how much of their collection is on their website, i.e. 25%, 20%. So far, I don’t know of one site that states clearly how much of the collection is accessible via their website.’

Kenn Rabin agrees. ‘It’s always just a small sampling on the Net. Sometimes the scarier sites are the ones with the higher percentage online, because it’s one thing if you go onto a website and it’s obvious that they’re only giving a little sampling – it’s another thing if you go onto a website where it looks so complete, but you’ve got to read the fine print and see, ‘Oh yeah, it only goes up through 1991.’ As a researcher, you’ve got to be careful because some of these things look like they’re complete, and you really have to know that you’ve got to call and ask if they’ve got anything else – you’ve got to be suspicious if you look for something and nothing comes up and you say: ‘Wait a minute, they’ve got to have something on the Gulf War.”

Only 46% of those archives responded when asked about the completeness of their online sites. While responses pegged the amount of their archive available online at anywhere from 0% to 100%, the average worked out to 36.5%. It’s important to keep in mind that this figure is based on estimates collected on the ‘honor system’, and only a few of the total footage archives responded.

***It’s a wrap

Whether or not footage research on the Internet has achieved the level where it could be considered a completely reliable and useful tool to producers is still a matter of opinion. It’s more likely – at this juncture anyway – that online is a good complement to a personal relationship with either a stock footage provider or a researcher.

Maggi Cook, a researcher based in London, offers her opinion. ‘A film researcher uses their experience to know where material may likely be. That’s the starting point. Talking to subject experts, interest groups and yellow pages leads to other sources. Basically you are trying to find a ‘rich vein’ of footage that you can then go to, research and view in more depth.

‘The Internet is vital, I still find it amazing that I can sit in London and search the National Archives in Washington. [But] the most productive sort of research is visiting the archives, being able to go through paper records related to their collections and talking to a dedicated librarian who loves their collection. It will never fail to yield a gem.’

>> Possibly the best search information:

Independent Television News (ITN)

>> What a good FAQ looks like

Intervideo – Russian Stock Footage Library

>> A nice, simple site

Index Stock Shots

>> Good design & multilingual

BBC Library Sales

About The Author