Practicum PNG Archive

Introduction

In several rooms, in the building where I work, there are row upon row of grey boxes containing Betacam SP tape recordings of everything from interviews to football matches, street scenes in Dublin to an indigenous celebration dance in Papua New Guinea.

 

An old system of archiving them, called 4th Dimension, once existed on an old floppy disk and some of its information was printed out to paper in the late 1990’s. But at the moment, the only information, the only data about what is stored on these tapes, are the slowly fading handwritten labels originally written by the cameraperson or director on each shoot.

 

Some of the shooters remain with the company, some have moved on and some are deceased. When the ink finally fades on the labels, the only way to know what is in these 20 year old tapes will be to play them, and therein lies the problem.

Over the last 10 years, the weakness of the SP format as a means of long term storage has begun to be a serious cause of concern.

The chemicals on the tape and even the tape itself have proven to be susceptible to poor storage conditions and more susceptible to time itself. Over the last few years, loss of picture signal and tape damage has been seen on more and more SP tapes.

 

This Practicum focuses on one series of tapes, shot in Papua New Guinea in the early 1990’s.

It is my intention to preserve the footage, enable access to the collection and outline simple tools and methodologies for small companies with similar collections of broadcast analogue tape – to preserve and curate their stock in ways that add value and thus security, to the footage.

I am fortunate that in this endeavour I have the backing of the MD of Kairos Communications, Finbarr Tracey and also the Divine Word Missionaries, who have had a presence in Papua New Guinea since the 1896 and many of whose members feature in the footage. Some of those interviewed are currently reviewing the footage to add contextual information to the archive.

Outline Methodology

With the risk of deterioration or total destruction of the collection always a possibility, the first step was to stabilise the collection by converting it to a digital format.

Later in the project I will outline the rationale behind my choice but after much research of methods, codecs and containers, I decided to utilise the simplest digitisation method in order to accelerate the process and ensure the transferring to digitlal was accomplished without delay or interruption.

Using a ProRes recorder had the advantage of simplicity, speed and limited setup – and so after the purchase of 2 x 8tb drives, the digitisation was begun.

Within two weeks, a collection that had sat upon the shelves for 20 years, was transferred to a useable digital medium.

This allowed testing of transcription technologies via uploading to Youtube and also Scene Cut Detection via DaVinci editing software to commence. Both of these treatments to the files were as much about generating momentum within the project as anything else – defining for me some practical starting points on the Practicum road.

One Tape Currently in Treatment!

As outlined in the image above – the steps to digitisation were simplified and kept that way in order to accomplish the actual task – once the files were in a solid, high quality digital format they could be compressed and uploaded to other sites to run tests with free and easy too use analytic software.

Talking about Text

Uploading a file to Youtube is one of the simplest means of transcribing the audio contents. 

When a video is uploaded to Youtube, the application automatically transcribes the audio content. This text can be copied from the site by opening the  and cutting and pasting the result into a document for checking.

The user is still required to sift through the resultant text to correct errors – in the case of confused audio or multiple languages, there will be many errors – but in every transcription
there will be need for checking so the Youtube model is not unnecessarily time consuming


The first sample tape transcribed featured ‘on the hoof’ interviews with Papua New Guineans. None of these interviewees are named in the footage nor are they listed in any of the scant documentation about the footage.

Fr. Pat Hogan, featured in the footage, and he along with Fr. Norman Davitt and Fr. Gary Roche are viewing the footage remotely in the hope of adding to the information about the footage.

 

The Story So Far

Due to the risk of catastrophic loss, I prioritised digitising the footage and creating a record of the metadata on the actual tape cases and tape boxes.

The footage was digitised to ProRes HQ – which was my chosen codec after examining the risks of using a slightly lossy format versus the risks of a longer and less immediate capturing workflow.

The footage was then copied to two Hard Drives – which are kept separately in the building, with plans to move one drive to the Divine Word Missionaries building in Dublin for safety.

Files were compressed using DaVinci Resolve, a free software platform and have now been uploaded to an online viewing application for remote annotation by principles involved in the filming and/or the location. This is a closed site, and while there is a risk the archive might be leaked from here, this has been discussed with the main stakeholder and accepted as a risk.

The Practicum has already achieved one of its goals in that the footage is safeguarded and the collection now features in discussions within the company and amongst stakeholders for future online offerings by the company – and instead of being a deteriorating, space hogging and dust collecting series of boxes, it is now again a viewable, accessible moving image asset.

The Next Stage

I am currently engaged in research for a White Paper on tools and methodologies to enable small companies with analogue tape collections to simply and cost effectively preserve and add value to their archive.

Annotations and adding further metadata layers to the collections are priorities and how to achieve these goals without resorting to expensive or complicated software packages form a large part of my ongoing research.

The role of Artificial Intelligence in face recognition is being increasingly recognised in the security industry, as machines learn to scan for faces in security camera feeds – whether this software can be repurposed without massive costs or learning curves is just one of the areas I am exploring as I delve further into the Practicum.

For now I am very happy that the tapes that sat for so long on shelves above my desk and near where i drank my coffee are now digitised and securely stored in three locations. They are being given new life in digital form and soon will be a valuable information rich resource for Kairos, the Divine Word, Ireland and the Papua New Guinean people.