Woody Hayday

Wired this month features a short story on Hans Ulrich Obrist and his archive of recorded interviews surmounting 2000 hours. It is proposed that the digitalisation and distribution of these is a challenge. This is a challenge I would like to take up and in almost open letter format I would like to discuss some ideas I believe may suit this wonderful, unique and invaluable collection of video’s.

Initially I would break down the challenge into two components. Technically getting the video from DV tapes and DVD’s onto home computer screens, iPhones, iPods, projection screens or whichever selected medium would be one. The second would be the contextual additional options. The scope for digital interpretation, communication, connections and display are countless.

The technicality of presenting video online is now an established practice. Example websites need not be named, for their popularity is well known, however despite the huge number of video’s present on these websites, quality is not a depth assured, this has allowed website’s such as TED to secure a wonderful foothold for providing higher quality (depth of thought not number of pixels) video. While TED is worth noting in this case (the people behind TED would perhaps be an excellent collaborative option) it is also important that it be distinguished from the possibility of this archive, of which there is no other comparative example on-line.

Not to dwell on the finite steps within the technological process I would concentrate this example into a summarised plan of 3 steps. 1. Digitalise all video in as high a resolution/audio quality as is feasible, taking care to maintain a proper method of file storage. 2. Process video – depending on the theory of display this could entail community based transcribing and translation, face tracking, tone mapping, automated or again community sourced segmentation or several other processes. 3. Encode all video for its intended output, depending on the intention this could be 1080p full hd video’s for download/streaming/youtube, higher resolution video for elsewhere or mp3’s for audio versions.

I think it would be important in this instance to not get caught up in the technological solution before deciding on the second component, the theory behind this archives display. Hans Ulrich Obrist has probably already considered many options, within the article it states a “system of tags” and a wonderful phrase “The archive is a polyphonic novel of more than 1,000 protagonists”. Tagging could work quite well but the question of who and how remains. These could be generated from the transcription of all the videos, something that could be tasked to a community to do digitally or by Hans himself or a third party such as end user’s viewing the video interviews. I think tagging broadly speaking will be a necessity in this project, however with such a unique collection I believe there are a variety of options to differentiate the final outcome of the archive online.

Whichever other process’s were performed on the video archive to provide a link or communication between interviews later on I would without a doubt get the video’s transcribed. This would add a depth to the archive enhancing features such as searching as well as providing opportunity for interlinking, communication and a larger array of display choices (the archive could then effectively be available on e-book reader’s, phone’s, subtitles for the deaf etc.) As aforementioned I would propose that this be done via crowd sourcing. Providing the video’s openly on-line and early-on in the project while asking viewers to transcribe what is said within the video they are watching (or even a small clip to maintain a higher attention span and therefore overall accuracy) could allow for a financial saving in transcribing coupled with a positive discussion about the proposed on-line archive by the contributors. This has been proven to work in a number of different scenario’s online and based on the depth of quality of the interview videos there would be more than sufficient interest to power this intention, in fact it may even provide a path to a much quicker transcribing (if a large enough community is built around the task) of the archive as a whole, especially into a number of languages.

From transcribed data a whole host of options are available, providing “similar interviews”, creating discussion groups from interviews, tagging and combining video’s. The transcribed archives could be coupled with other processed data such as face tracking. The interviews could be processed to record the facial expressions of the interviewee’s as they address different topics, acting as an “Expression tag” you could then similarly create different connections, suggestions and comparison’s using these. A web of similarities and dissimilarities could form an interesting navigation and experiment into the archive.

Similarly the video audio could be utilised to the benefit of the project, mapping sound-bites to text or even facial expressions. Perhaps visualised in-line with the video or just tone mapped to provide similar talks or a more interactive navigation. The interviews could also be available as mp3′ files for download or online listening, for the enjoyment of the blind or for people to listen too on their mp3 player’s or mobile phones.

Another possible pursuit would be to break the video’s down, segmenting the archives into smaller chunks might provide more options for viewing the material, communication between interviews and more. This again could be done through crowd sourcing of the task or by using audio mapping to distinguish breaks. This would completely depend on the length of the interview video.

There are, of course, numerous considerations and possibilities for a project like this, providing a way for users to access the archive that is both relative to the archive itself and conceptionally creative would be a great, fulfilling challenge. In just thinking about the idea for around an hour I have come up with a lot more ideas than I can fit in here, how about a new type of “rating” system that identifies the least viewed interviews and shows them on top of the page, therefore creating a cycling effect ensuring that the viewers on the whole as a crowd see the huge variety present? perhaps a single page with several segment’s of different interviews all pertaining to one topic all displayed and ready to click, draggable video frames that immerse the viewer in the interview? I would love to work on such a project.