 Thank you all, Kiyota Tato. I'm not sure how the time's going to go. What I'm about to show you is very broadly based on a presentation I gave about three weeks ago at the International Federation of Television Archives conference in Amsterdam. And the process we're going to go through today is more or less unique. There are similarly scaled or larger-scaled digitization and collection development and preservation projects around the world in the audiovisual area. But not ones of anything like the scale, which are like the scope, in other words, ones which encompass an entire national screen and audio culture. Altogether, there are large television collections. There are large audio collections being digitized. But very few instances where those things are coming together with the aim of making some kind of coherent whole. And I was speaking to the next presenter in discussing my dress today. This whole process starts with a political imperative, as you'll see in a moment. And as you can see, by the way, I'm dressed. I myself am subject from time to time to political imperatives. And right after this presentation, I'm actually going to go and meet some politicians just to keep it real. But where I'd like to begin is on launch day, in fact, launch eve of Na Tawanga Sound Division, when the then-Minister of Broadcasting, Craig Foss, gave a brief but impassioned speech that, amongst other things, laid down this challenge. And it's easy to say. But it encompasses an enormous amount of detail within this very simple overall command. So our question was, could we do it? How could we do it? Could we do it better? In order to make some kind of sudden, we had to break that down to some degree. And I've broken today's presentation down into two parts, which is more or less how our rationalization went. First part is, in order to do that, we have to integrate. We have to actually bring together three organizations, three collections, three very distinct ways of doing things in order to make a national collection. And as, let's say, a given, although others may contest it, we're going to do that in the digital world. It is not really meaningful anymore to try to do that in the analog world. And I will say also that with this presentation and with the task, it seems to me that the further ahead we look in these kind of projects, the harder it is to make up the detail. And I think you'll find the detail gets fuzzier and fuzzier as we go along. I make no apology for that. So the second part of the task really is at Rhymes, which is handy, but that's to curate. In other words, having passed that first milestone, that really just gives us an undifferentiated mass of data. And that in itself is not an accessible collection. And what are we going to do to make it accessible, both in terms of process and an outcome? It has to be made accessible. It can't just simply exist on servers or on tapes. So we start with the most obvious thing. We have three organizations, three cultures, as I said. Very brief, you all know, well, probably know all three of these, almost certainly know two of them, quite well. Ready in New Zealand. Established as a state broadcaster many years ago. Established an archive along the way for its own purposes and had got to the point in the lifespan of that archive where it was no longer intrinsic in the broadcaster's business. And you could say something similar about TVNZ, rather different path, a rather more commercial focus, but nonetheless had reached the point where it was no longer considering an analog tape-based archive as part of its core business. In both cases, their archives were outliers, physically removed from the mothership and at some distance, you have to say, from the center of power. The film archive, on the other hand, smaller by far than either the Radio New Zealand or the television New Zealand, but on the other hand, solely focused on archiving. So for us, archiving is at the center, for the other two, at the periphery. Now here's a quick rather cartoony look at the collections that we're dealing with. And you can see the profiles are quite different. The film archive, quite a part of the different media, as a focused archival institution has made very substantial inroads into cataloging and digitizing its collection. The sound archives of Radio New Zealand had done a reasonable job of those things, but digitization has only just got underway in the last year or so, following the transfer to the film archive's control. Television New Zealand, on the other hand, massive collection, quite a lot of cataloging, maybe in a rather eccentric and broadcast, a focused way, but in effect, no digitization. So we have a collection there, hard to count, but around 650,000 titles, items, it's a slippery word, but none of those exist in an accessible file form. Now just for the purpose of this exercise, there is a debate to be had about whether preservation and digitization are strictly equivalent, and we're not going to have it. Going through this process today, I'm going to use the digitization as more or less interchangeable with preservation for practical reasons. It really does represent the best shot in 2014. These things reach a public. You can argue how accessible some of them are. They reach the public through broadcasting, obviously enough. That's uncounted and quite difficult to count at this stage. And then there's a whole lot of non-broadcast outlets as well, which are essentially the film archives model. So we have, there are websites involved that are managed by us. We have a series through the country, I think currently 17 off-site access points and museums and libraries called Medianet, server-based. We provide links to a number of other online providers, such as Tiara and so on. We have physical venues, so we do screenings and other sorts of presentations, exhibitions and so on around the country. Including community-level activity of all kinds. And we loan material, so that a number of either real space or online providers access the collection and use it in their programs as well. And altogether, that adds up to about 800,000 non-broadcast users a year. So a pretty substantial audience already. But only accessing, in that big picture, a very small part of the collection. If we simply bring those three collections together, just count them as one big pile, those three graphs turn into that graph. And you can see there pretty clearly the size of the gap between the total collection number and the digital portion of it. However, it's easier to draw that, that isn't the same as combining the collections. They are still very distinctly separate. They're in separate physical spaces on separate media and so on. So looking again at the collections in a different way, how do we know what we know about them? Well, essentially through databases. And once again, there are three different cultures, three very different platforms here that we're working with. And significant amounts of information over the whole collection, it's not on databases at all. If we're lucky it's on spreadsheets and in some cases it is still on good old fashioned index cards. The content on those records, call them, is also very different from institution to institution. What a broadcaster has thought of is valuable to capture about their holdings is quite different from what a public archive might or a public user might. So there's a lot of information, for example, in the television New Zealand database, shot by shot around news and current affairs. If you want a picture of John Key holding onto his tie and looking to the left, you can find it. This is not necessarily value for money in trying to catalog a collection this large overall. Another thing to bear in mind is that this is still largely an analog collection. And while librarians may struggle with some of the differences between bound books and large scale printed objects and photographs and so on, the landscape is far more challenging in the audiovisual area. This is just a short list, believe it or not, of the different kinds of technologies and formats that exist in the collections, some of them in very large numbers. And I think the thing that is particularly pressing about this is that each one of those formats needs an entirely different and unique form of player technology. So it's all very well to count up how many high band-umatic beings you have, but high band-umatics are only worth more than a bean if you can actually get a high band-umatic machine to play it. And we're talking about huge numbers here. So this is just an estimate, but for example, if we go to those umatics, we need to have, roughly speaking, 10,000 hours of play time to get through that pile of umatic tapes. Now, we have about four or five, depending on whether the wind's blowing the right way, high functioning umatic machines left. Each of those machines uses contact magnetic heads on tape, those wear out. No one in the world has manufactured those heads for the last 10 to 15 years. There is a distinctly finite resource of umatic play heads. We might have, hard to say at this point, two, 3,000 hours, if we're lucky, of play time left on our current stock of machines. This is reproduced across virtually all of those formats, these kinds of equations. So you can see that the job doesn't only involve dealing with a lot of videotapes, it actually involves a very deep investment in totally obsolete technology. And one thing I'll say about that is that that creates pressures and drivers on a preservation process which are quite separate from content. If you have a pile of tapes which are themselves beginning to degenerate, you have an urgency about preserving them which might override, it certainly will compete with the public interest in the content of those tapes. If you're faced with the thought of either providing an audience member with what they want, or, and as a result, consigning some unique material to perpetual oblivion, it's a difficult choice. I won't bother actually doing the sums behind this, particular piece of arithmetic. We have to do these sums involving those machines, their time, the human time. Basically nothing in the collection can be transferred any faster than its original playing time. In most cases, it takes a great deal longer to prepare, repair and deal with that transfer so the productivity rates can be a lot lower. Some things can be done simultaneously but just taking this particular algorithm or something like that really, equation, it looks like that pile of work I described before is about 23 calendar years for eight people working full time. Now, really nobody has 23 years to wait, at least of all the minister broadcasting. However, it's not just about the content, we go back to this again and I made some comments about how different parts of this have been compiled in different ways. So the object here will be to actually investigate all the data, the metadata that's been captured, select the useful and compatible parts of it and migrate it. So if we do that and we take those elements from each one and we add to it the media from the digitization process and the metadata which is created in the digitization process and makes it possible for us to meaningfully use that material afterwards to find it, reproduce it, play it and so on. We have a new, is it an object? New element, which we're calling a tuple. Now, tuple, I'm not a mathematician but a tuple is a real word and it's applied, I'm told, by mathematicians, particularly to relational databases where it represents a set of value attributes. So it's more or less equivalent to a record in a standard non-relational database or catalogue. We have to create a new tuple for every element that we put into this new merged collection. We have to migrate not just material data, video, whatever, out of old analog forms, we have to migrate data along with it. Then what? Well, we have to make some choices. Now, if you're a politician, that's pretty straightforward. You just take the whole collection, you catalogue it, you digitize it, you're done. But as we've seen, that's a 23-year prospect at best. Could well be a great deal longer if things go wrong and heaven knows they probably will. So we have to look at some pragmatic outcomes. So I'm suggesting, for our planning, that we look ahead about as far as we realistically can, which is about three years, which happens to dovetail with the minister's own useful instruction. Where will we be at the minister's deadline? Well, if we work away at about that productivity rate I indicated and the collection doesn't grow too much and I hope TV3's not wrapping up the trucks right now, if it just grows at a containable rate through new production and deposits, then it'll look a bit like this in three years' time. So if we combine existing digital items, where we've already done, new born digital items that come into the collection and the migration output over those three years, we get to there, which is about 17% of the collection digitized. It's a huge number. It's 150,000 titles, items, made digital or delivered initially as digital. Huge number, far more than the minister's target to put online, far more than we would be able or I think want to put online. It will contain a whole lot of material preserved for those reasons I said before, reasons of degeneration or loss of technology or whatever rather than content. So it will be a very, very, very broad church, a critical mass and what we have to do is work within that and make some choices. This badly drawn thing is the familiar long tail curve known as the Pareto curve I'm told and the basis for Amazon.com's business model I'm also told. And this is used in all kinds of ways to basically estimate where activity will happen over a body of work. So this could be a collection of stock in an online bookshop or whatever. And one fundamental part of that is the idea that about 20% of given stock in Amazon for example or in a bookshop in Whitcalls will attract about 80% of the demand. This is sort of a truism of marketing and planning. If we take that particular ratio and apply it to our collection, we find that's about 30,000 items. So we're actually looking, if we were to somehow choose the 20% of the greatest public interest out of our critical mass, we would have a body which actually over achieves Minister Foster's target by a considerable amount, 30,000 items. It's a lot of films, television programs, radio programs, pieces of music and so on. And it's less than 4% of the collection. Let's just do an exercise of saying, well, what might you put in that to make up? There's a guess, just a guess, of what a national collection made accessible at these kinds of sampling rates, what it might contain by those sorts of measures. Once again, it's a tiny proportion of the collection, but it's a huge quantity for these kinds of projects. Just to give you one scale comparison, the entire history of New Zealand feature filmmaking has produced a tad now, I think over 300 feature films. We're talking about 7,500 films of which feature films would make up that very small proportion. There are other similar projects around the world like the BFI and others are trying to do things like this. And they are typically dealing in maybe a tenth of the quantity we're talking about here. So these are very ambitious targets, however small they are within the context of the whole collection. How are we gonna do it? And this is where we're starting to get into that slightly more fuzzy, long-range detail. We need to harness a whole lot of inputs to do that. We need our own staff expertise. There's an awful lot of knowledge in all three archives about what's in the collection and what's interesting and valuable about that. We need to actually use the feedback of our professional users. So that is, that's the production industry reusing material, it's researchers, scholars, filmmakers and so on. We need to use targeted audience research. We need to actually go out and question people about what they would want from such a service which 4% of the collection actually is of most interest. We need to use once we get rolling crowd sourcing techniques to test some of these assumptions and speculative moves of getting stuff up in the first place. And being the kind of organization we are, which is not just hemmed in by but genuinely respectful of copyrights and a whole lot of other interests held in the collection by the subjects and so on, we need to work within a framework of rights and permissions. All those things have to be brought together in order to give us confidence about our choices. The upside is our margins of error are quite good. When we're dealing with 30,000, you can be liberal about the feedback from a professional user. If some particular barrow pusher really, really, really wants to do something about trams, well, there's room in there for trams. We looked earlier at these, oops, they've gone pale on, at these existing outlets and most of those, if not all of them, I think are still relevant and will still persist into the digital future. Most of them, in fact, all of them really actually rely on digital technology anyway, even our live presentations if we go to a church hall in Haukatika, we are actually presenting a digital experience. It may still be on a hardwooden seat, on a bed sheet on the wall, but fundamentally, it's got there now because of digital technology. But we have to go beyond these things. I think the first thing we have to do is confront these numbers and be realistic that you don't deliver those kinds of very large scale access through any of the existing tools that we have in this country. There are some examples around the world the BBC's iPlayer and a French national institution called INNA where they are delivering on a very large scale and those things are built essentially to do one thing which is to deliver content as directly and as unfiltered as possible to an audience on these kinds of scales. They're essentially databases full of media. There are some very specific audiences that we already engage with in various ways that I think that who could be incorporated into this much bigger delivery. Education is just one of those, but it's a key one I think and one likely to be supported financially and politically. We have the world of curated websites, for example, the Natonga themselves. In April next year, we will launch in conjunction with the National Film and Sound Archive in Australia an ANZAC website to commemorate World War One. A much smaller, tighter-focused device entirely than a $30,000, $30,000, sorry, I wish, $30,000 item database. There are existing curated websites such as Zealand On Air, such as Zealand On Screen and so on who would also benefit through this, through getting the ability to have more tightly focused but richer content. And then there's the question of other platforms and I'm gonna leave this here now. It's a challenge that everyone in the archival community is wrestling with. What is the point of building these things when we already have YouTube? Are we actually just empire building or should we really go with the flow, follow the audience to the existing platforms and put our collections through those platforms to the wider public? And that's where I'm gonna stop, thank you. Thank you very much. Quick questions from the floor, we've got one over, two over there. Probably have to leave it at those two questions, I think. Oh, hi, Frank. Test from NZ On Screen. My question is around the collaboration, especially on the last one around using the existing channels. You're even going to do interesting that there wasn't a mention of collaboration with existing platforms. I was a little bit distressed to say that TVNZ, you said zero digitized one, we have hundreds of thousands digitized one. It's just sort of great to sort of get your feeling about how you can work with other organizations such as NZ On Screen, Audio Culture, National Library, et cetera. And not just on the content, but also on the technology as well. This is something else that we can share or collaborate on for you. Just winding back, the comments on the collection and the digitization, that is what is held within the TVNZ collection. So while there are digital iterations of many TVNZ programs hither and thither, none of them are actually in the TVNZ archive. They have not retained one file through all the work that's been done. So that's a big challenge for us. Moving on to collaboration. I mean, nothing in this presentation specifies how that is going to play out. It doesn't in any way rule it out. It simply focuses on our immediate housekeeping, if I can call it that, task of dealing with these absolutely enormous numbers and how are we going to do that and how are we going to get up to industrial scale, which is what the political expectation is. What do you collaborate on? What's needed to be, I think, through additional things to get content out now because we're always getting a competition situation or we can't get content out. And I find that disgusting when there's so much to be digitized that it's not an open where we can create a situation where we can collaborate together on that. I think some of it, I mean, this is a very big question. I can't really do much about it today, but part of that comes down to the observation I made pretty briefly at the beginning about conflating preservation and digitization just for the purpose of this. Our focus is a preservation one, so that, for example, if we're not going to be too technical, when we have a high band of humanic tape, what we need to do is to turn that into a 10 bit uncompressed video file. So it is preserved. And in order to do that, we have to set up a digitization factory, essentially. Now, the current activity around things like web use and so on tends to focus on much more convenient pocket-sized codecs which work well for that kind of use, but which do not provide preservation. And there's a real huge hazard in all of this that if we do not immediately get on to proper preservation by digital means, then the means to future proof even those digital files will go. So we have a big focus at the moment on preservation by digital means. There will undoubtedly be, for anybody, major access outfalls from that. Once you have your 10 bit uncompressed file, then it's quite possible to make any iteration you choose for H264 for this year's web fashion or for something else next year. So that, I agree, the driver behind all the arithmetic here. Just none of that, to emphasize, is in any way a comment on willingness to collaborate. It's just the reality of the preservation job and nobody else in any way is equipped to do that. We came into TV and said thinking they had been doing it to discover that they had. They'd been copying analog material to digital data, which doesn't actually get us any further on towards true digitization. So we've got a really, really serious issue around preservation to address to make that clear. Okay. If Simon's got a question that he thinks has a quick answer throughout our art, otherwise that would be the last one. It doesn't mean we're going to dispose of anything, but I think we have to have given that the legacy databases had very, very specific in the case of TV and Z, particularly commercially sensitive information on them. They would not form part of a public database in the future. That data would be thrown away, apart from anything else TV and Z themselves would be taking it for their own purposes. Okay. Thank you very much. So thanks to Frank for that presentation and also thank you for throwing us a party for after the event. Don't forget that the official after party is down the road on Tauna history at Ngataunga headquarters. And there will be drinks and nibbles followed by a screening of sleeping dogs from, so after party from five, screening from six thirty down there, make sure you, if you're going to attend that, take your conference badge to get entry to the film.