Next steps for Moving Image Archives

(This blog post coincides with the publication of the recommendations of the JISC Film and Sound Think Tank – pdf report)

JISC has invested signficant amounts in two fine archives of of moving images.

NewsFilm Online contains over 3,000 hours of film featuring events such as the Crystal Palace fire, an early interview with Nelson Mandela and the reaction to the death of Princess Diana.

The archive is soon to be built into a larger platform for images and sounds, called JISC MediaHub.

Meanwhile, The British Film Institute’s InView has 2,000 non-fiction UK film and television titles (covering topics such as health, education and the enivornment) from the 20th century to the early 21st.

Some high-level usage statistics are available for the NewsFilm Online resource (section 2.9 of this pdf report). It’s encouraging to see many universities have signed the licence to allow them to access and download the content. At least 200 institutions have signed up to use the resource.

However, usage of the resource remains unbalanced, with the number of searchers and user sessions being high in the first half of 2010 and then dropping away; overall it there are just over 7,000 user sessions per month.

Unfortunately, the number of times a video is accessed or downloaded is not indicated. I suspect, however, that the numbers would not be as high as one might want. Certainly compared to the numbers you might see for textural resources, usage for moving images archives is not in the same league.

Evidence from JISC’s recent Impact and Embedding projects provides a rough point of comparison. Web statistics always need to be taken with a pinch of salt, but the Vision of Britain site attracts 96,759 visitors a month, whereas the Old Bailey Proceedings attract 83,201 monthly visitors (see page 18 of this pdf for the statistics)

Why is this? What stops moving image archives from becoming more embedded in educational practice? I don’t think there is any one reason; but layers of reasons that build up to create a formidable obstacle.

1) Authentication

Because of the difficulties of copyright and licencing, many archives can only present their digitised film behind authentication.

Sometimes this does not prove a problem. The wonders of Shibboleth means access occurs with the minimum of friction.

But too often, there are barriers standing in the way. The university librarian needs to sign a licence; the researcher has forgotten his password; the teacher cannot trust that all his 20 students will remember their passwords, and foregoes using the resource.

It would be worth comparing statistics from an open archive (such as that at the Internet Archive or the Wellcome Film collection) to those behind access. And while most content on YouTube is not scholarly, it’s phenomenal user numbers say it must be getting something right.

If film is to be used more, much more of it needs to be made available under generous licencing conditions. Hopefully, the recent suggestions of the Hargreaves Review will create a framework for this to happen.

2) Moving image films are just too rich in content

Projects such as InView get funding because of scholarly richness of the content. But this actually turns out to be one of things that hold it back.

Television programmes that form part of the InView archive (such as legendary Channel 4 discussion programme After Dark) are full of meanings, interpretations and histocial and social signposts that provide fertile ground for scholarly inquiry.

When presented with such a film, sometimes several hours in length, the researcher and teacher still has to do an awful lot of work to go through it and pick out the points of interest that will engage her, her peers and her students. This might just be four or five clips of a couple minutes each; finding those ten minutes within a 2 hour file is a difficult job.

Projects such as NewsFilm Online try and get round this by ample description of each of its films. This partially helps, but creates another additional problem – now there is a huge wave of text to comprehend.

YouTube works well because the content that is submitted is often pre-edited into bize size pieces. One can instantly understand a 4 minute clip indexed on Google of the “Collapse of the Tacoma Narrows Bridge“. But the moving image archives present too much.

3) Lack of clear methods of citation

Once a scholar has located the slice of the film that she wishes to use, how does she share that clip with other researchers? Or with her students? Or with interested members of the general public? URLs at the moment point to whole film clips; using a URL to fix a time ‘slice’, or even an area of an image is still far from being part of the mainstream.

It’s considerably more difficult to share a clip of a film than it is to share an image or a portion of text.

4) Difficulties in editing and republishing

This is related to citation: downloading, editing and republishing movie content is time consuming and takes the right kind of software, hardware and a healthy amount of expertise to make it happen. Even with online video editors, it’s going to be a barrier to those who do not have the time to fiddle around.

For those republishing film on the web it requires the authentication to work and a workable platform to present the reedited film.

And even if you are not republishing on the web, there are a number of problems. It’s not easy to embed videos in PowerPoint; many have tried and failed.

And for much the same reason, when students present essays, it’s still much easier to do it with text / prose rather than try something novel like embedding a video. Presenting video is a dedicated task.

Beyond the tehcnical issues, the rights issues multiply (is it clear to the user if downloading and remixing the original material is permitted? can the remixed matieral be republished on the open web?)

5) Types of usage

Academics from the University of Hull produced a rich pedagogic framework that illustrated the many different ways in which moving image archives could be embedded into educational practice.

So, the framework, demonstrates how a teacher can exhibit ‘dissonance and shock in a small teaching space‘ or how ‘students can share and build knowledge in a collaborative fashion

It’s useful, inspiring stuff, but I suspect that most teachers still only really conceive of using moving images as simple illustrations or analogies – and for this short clips on YouTube are much more successful than digitised archives.

To use the videos in the time suggested by the University of Hull framework takes time and effort – some lecturers will engage, but it takes a special amount of passioan and dedication to make it work fully in the classroom. For some, pressed by other concerns, elsewhere it will just be too much.


In looking at moving image archives, it’s worth comparing them to YouTube. Why does YouTube work? Why does it get so many hits?

Largely, because it avoids the issues cited above. No authentication issues on YouTube; complex resources are reduced to clips; clips can be easily embedded elsewhere; the illustrative or anectodal power of YouTube videos is easily understandable and reusable by anyone. plus there is critical mass of material on YouTube – even though there is plenty of guilty cats and sporting cock-ups, there are plenty of little gems which can act as brief illustrations within the class.

None of this is to say that moving images are not useful for education. They are. Moving image archives are immensely powerful in conveying historical and contemporary narratives.

But we’ve still not found the techniques to allow such archives to be used in education. And if we don’t we will never the nurture digital moving image in the way it deserves.

4 replies on “Next steps for Moving Image Archives”

All too true. Our open access Wellcome Film archive available through our catalogue is not that easy to use as you have to search the catalogue first and then drill down to film content, and we got an average of 1,200 pageviews per month in the past year (people spend an average of 4 min each visit). This is for about 500 titles. By contrast, we only have 300 titles on YouTube, and yet we’ve had over 1 million “upload views” since we launched it, averaging out at 37,000 views per month (although this peaks and troughs).

Some very valid points. This does start from a quantitative point of view however; number of hits doesn’t always equal ‘value’. From this perspective YouTube doesn’t always ‘work’ (e.g. for researchers who wish to use video as primary data and need guarantees of longevity). Perhaps we should be looking for additional ways to measure impact (citation for example)?

However, in total agreement with the argument that combined authentication/technical barriers prevent wider use.

Your blog raises some very important issues:
1. The ‘scholarly’ nature of YouTube.
Yes, it’s unreliable, but has more scholarly content than all the MediaHub collections combined. It also takes advantage of a huge amount of free user input.

2. Small download rate.
There are very few practitioners using these resources because there are very few practitioners using these resources. The collections need to continue to be inventive in how they are promoted to stakeholders – more examples of content being used, more free pedagogically sound ‘learning packs’ based on content, and more two way discussion with users.

3. Authentication
For time being authentication is essential – just as it is for journals and other subscription services. However, if more content could be freed up, it would raise the profile of the collections.

4. Moving images is ‘too rich’.
No it’s not, it’s the tools used to interrogate the collections that are too limited. The collections need to incorporate additional functionality – automated analysis (spoken word, music, edits per minute, brightness etc.) and integrated note taking, clip creation, sharing and tagging functions (see Synote, MovieBrowser, DIVER etc).


Tim (JISC Digital Media).

Key issues being raised here.
1) Finding materials
These are ‘collections’. You need to know that they exist with something that might interest you. Instead, we need to bring resources to people not people to resources. Hence attempts like the new BUFVC federated search across all its metadata collections (if you are searching for newsreels material, you discover radio material). There are also techniques of offering material to any searcher beyond their search terms.
2) Editing material
The Hull guide is excellent, but so hidden on the NewsFilm Online site (behind a link to ‘JISC’) that no-one knows its there. I’ve just found it again, so many thanks! Surely such a key resource should be highlighted on the front page?
3) Download is no measure of usage.
Download once, but use many times.
4) Absolute value
My students from China are bowled over by the NewsFilm Online material on the Tiananmen ‘Incident’. They had no idea that the events were like this. I guess they, too, may be downloading the footage…
5) The final mile
There is still a reluctance to use such streamed material in lectures. Even if you log in ahead of the start of your session, like as not the link has timed out when you get to the point to use the clip. Safer to download , but then see (3) above. Often the AV does not work or someone has messed around with the settings, so you need to get in to do a bit of prep. Not easy when we are pressured to use our plant every minute of the day…
6) Citation…
…is the next key problem to tackle, both metadata-wise and extract-wise. Within the EUscreen project ( (based on streaming) an extract tool is being designed that will allow quotations within ‘virtual exhibitions’ and also within a planned online academic journal Critical Studies in European Television, whose first issue should be ready in November. Unless…

Leave a Reply

Your email address will not be published. Required fields are marked *