 Hello. I'm Dave. Thanks for coming along and having a listen to what we've got to say. I'd like to sort of start just by saying nothing we're going to say here is the rule. We don't have the answers. Just really seeking to try and provoke a few thoughts and ideas amongst all of you so you can go back to your own respective institutions and maybe go with a few fresh outlooks. To treat this with that in mind. What I'd like to talk about is our collections and by that I mean not just archives in New Zealand's collections but the collections in your own institutions. So take what we're about to talk about back to your own desks and your own theories and your own ways that you do things. What I want to kind of break into is a little bit about how we traditionally build our archives or our collections. We have our physical things that we preserve. You know we stick stuff on the boxes on the shelves the film reels and the cans that kind of thing. We have our digital preservation which we've been hearing about all morning about how we take either the born digital objects or the things that we've digitised and we stick them in a nice big system that looks after them forever and ever and ever and then we all have to provide digital access to things as well. OK. And this is how we currently think the archives and perhaps all the collecting institutions this is how we build things. This is how we work. So I'm going to break those down for you a little bit as well. I'm going to go into physical preservation and can I just point out by the way this is the menu from the dinner when the All Blacks came back in 1905 and we're christened the All Blacks. This is the actual thing. It's quite cool. Physical preservation. It's secure right. Basically it's pretty difficult for somebody to walk in the front door of our buildings Potter through however many doors and security guards there are and start having a nosey round at all the stuff on the shelves. Pretty difficult to get out. So they're quite secure sat there and they're comparatively cheap. You know because the cost per sheet of paper to stick a file on a box on a shelf is pretty cheap really. It's provenance implicit and by that what I mean is when you come into archives New Zealand you say can I have the thing? We give you the thing if you're allowed to see it and therefore it's the real thing. It's not one of 300 versions. It's it. A real archive with provenance implicit. They're pretty easy to care for. You can stick them in a nice stable environment, pop them on the shelf and they'll be there for a while. They kind of last a while. But they're pretty difficult for mass access. So if I said to you guys that we wanted a million people to see one of our archives physically today somebody in our reading room is going to have kittens. It's pretty difficult to do. Digital preservation. Again secure. We build these awesome systems where archives go in. We look after them. We think about stuff like migration and backups and redundancy and all that kind of stuff. Really really awesome secure systems that we build. So it's pretty costly because all that infrastructure or the wine glass as we happen to be able to refer this morning it's pretty expensive to do. It's provenance crucial though because we strip out the metadata say from an email. We tell you who it's from or who it went to or when it was sent and we take the title heading out. Just words doesn't really mean much. So the provenance of a digital archive is quite critical because it's not really up to much without those context around it. And these systems that we build are realistically engineered for migration and they're designed to preserve things first and make access to them second. Which is the odds really with the concept of digital access. As we live in this world where everybody wants stuff right now, thanks in whatever format I want it right now, thanks. And that gives us issues. We talk about security as being bipolar. By what that I mean is if we have to make something digitally accessible that we are worried about the security of that archive or that file handing it out there to the world is a bit of a scary thing to get our heads around. But if it's stuff that we actually want people to have it's pretty cool right? You just kind of go there you go, have it, off you go. It's cheaper than digital preservation generally. We don't necessarily have to have the zillions of copies of it and the infrastructure. We just stick it up there on the web away we go. And provenance irrelevant which might be a bit of a scary sort of statement to say but it is provenance irrelevant I think. As soon as you publish something on the web, the ability for somebody to control what happens to it and how it's reused and its context of provenance and everything else like that pretty much disappears on site. So realistically you can just let it go. And all these systems that we try to build for digital access we kind of want to make them so that access is the main design feature. That's number one on the list of things we want to work at. Because the public have this expectation of anytime, anyhow. And then we get this, okay? A big dollar sign. Which is the problem every single one of us in the room have got. We've already heard it three or four different times this morning. It doesn't matter where we are, anywhere in the world, our institutions have got this kind of problem about kind of where is the money coming from? If it's there, how do we spend it? It's not quite enough to kind of go around and get the job done properly. So in our traditional model of how we build an archive of these three tiers of kind of collections. What do we do about this? Is really what I'm trying to get at today. So I'm going to pose a question. We're going to pose a question very shortly. I'm actually going to refer to a quote that myself and a colleague, I think he's in the room actually, Sean Mosley works with me here at the Archives of New Zealand. Kind of half finished each of this sentence over a coffee table. You know, kind of looking long and lean to each of his eyes and we came up with this quote. So for those of you tweeting or taking down notes, feel free to rob this one because we're quite proud of it ultimately. The only thing easy about digital is the end user experience. And the more you think about that, the more it's true. It's an absolute pool of wisdom I think. So, what if? Here's a question you see, I missed a slide. What if? What if we physically preserve our archives or our collections, or we digitally preserve them and make them digitally accessible? I'm going to labour that point just to kind of make the question really obvious. What if we physically preserve our archives or digitally preserve our archives or our collections? Not both, pick one and then make them digitally accessible. What if we just have our physical archives on the shelf and we worry just about digital access? So our preservation needs as an organisation is dealt with by having it there physically. Or we go down the route of having digital preservation, be it an archive that's deteriorated, gone away, we've digitised it, we're looking after those files, or a born digital archive or a record, and then we make it digitally accessible. Choose one, okay? So I'm going to give you a couple of examples really about how some of this thinking has been permeating the work we've been doing at Archives New Zealand recently. The first one we're going to whip through relatively quickly, and the second one, the paper based one, there's a bit of geeky magic in there for all those nerds in the audience, so we'll labour on that one a little bit longer. So the first example about film, okay? We've had the film processing lab out at Park Road Post Production, a nice facility here in Wellington. Archives New Zealand has about 21,500 reels of film, roughly. Give or take a little bit. A lot of those have got some fairly serious preservation needs that need to be looked after, okay? So this partnership that we've had with Park Road Post has been really critical to our archive for quite some time now in terms of managing how that work's done. I'd like to point out that they're not just reels of film, they're not just reels of film to anybody, to be honest, but amongst them are the weekly review films, which are made by the National Film Unit, and are one of our collections of archives that are on the UNESCO Memory of the World Register. So in terms of film reels, it's quite a big deal, you know? So as it was, we used to go through a digital-intermediate process. We take our realer film, scan it, do some nerdy digital magic to it, do lots of preservation work, kind of then transfer it onto another roll of film, and off we go. A fairly big, complicated job to get stuck into. So we used to do this work basically as budget would allow us to kind of get them through and out the door. We keep both the film scans, so the digital stuff, and the reels of films that we've got, sometimes even keep in the original as well, I might add. So then we've got two physicals, as well as the digital master. And we've been making our digital preservation master file work to give us our digital access version as well. So that's how it's been working. And then this happened, which is that Park Road Post, as many of you probably already know, said that their film lab was no longer commercially viable. It was a done deal. They weren't going to bother with it anymore, and by the 28th of June this year, it was over. So for Arcos New Zealand and many, many other people, that's a really big deal, because what the hell were we going to do about continuing to preserve our film? Well, we bought it. Which has been out in the media, so a lot of the kind of local Kiwis around here probably already know this. This is basically how it looked about three weeks ago, as we're part installing it into Repository Area Archives New Zealand. We've actually brought a lot of the staff over, used to work out in that lab. They've come and joined us in this sort of team archives. So we've brought not just the facility, but the expertise to run it as well. And that's awesome. It's something quite ironic in that, in the fact that Park Road Post got hold of it originally from the National Film Unit in the first place. So it's kind of gone full circle and rejoined the government as part of the way that we do things, which is quite cool. So what are we doing from here on in? Well, film transfer will, once installation has been completed, be done in-house onto polyester film stock. So we're making new physical preservation masters that are going to last quite a long time. Digitisation is done on a kind of as needed basis now. Sometimes we will do that big preservation work on it, but a lot of the time it's going to be access standards available to you, and that's it. It is one of the only film labs, the colour film labs, anywhere in the Australasian sort of area. For a little while earlier on in the year, we actually thought it was anywhere in the southern hemisphere, but there are one or two other little dotted around. And it's nice to see our colleagues over in Australia bringing one in right about now. So they're kind of joining that sort of thought path as well. And of course, the best thing is, it allows us to keep preserving our film collection for New Zealand and beyond. So we take you to our second example in paper. These are the World War I New Zealand Defence Force files, or the NZDFs of people might be hearing it around town at the moment. They're getting to a higher and higher profile at the moment because of the importance of them to the WW 100 celebrations and commemorations. They're all digitised in-house now. We did have a little period where a third party company did some for us, but essentially everything is now being digitised in-house. We have a team of about 30 staff, four on the workstations at once, and we aim to try and get them working about 16 hours a day. So realistically, we kind of started about 5am in the morning and go through to the early hours-ish again, every day, sudden days a week. As you can see from the picture, that's kind of because the files themselves sheets of paper, glorified post-it notes on them, a little bit torn, a little bit dog-eared. They also have microfilm elements in them as well. So they're very, very difficult to do any automation to them. They're just a very manual process. About 161,000 records, and our projection is over 3.5 million images, possibly heading up towards 4 million images, which we think makes it the biggest digitisation project in New Zealand at the moment. Our projected cost about $1.6 million to actually do the project, the original projected 105 terabytes of data. The scary one being the original quote of $1.5 million a year to store the files. And there's a gasp round the room, so it's not just us that did that. So what do we do about it? This is still from a little movie that two of my colleagues Miranda and Barnaby have made in the last week, which is playing on our stand out there in the hall. So I'm going to have a look at this because it's quite cool. We introduced JPEG 2000s into the mix, which for those of you who don't know, is another image format, as opposed to TIFFs, which have been the traditional imaging preservation format, and as opposed to JPEGs. So we brought this into the mix and started playing around with it. So this is a bit where we get slightly geeky, right? So stick your geek hut on and see if we can keep up. And if you don't, you can ask me afterwards. JPEG 2000 files are based up out of these compression layers, and you can specify how many you want. So think of them like a sandwich looking down from the top. Each of these layers are specifiable. What that means is when you look at a picture, a top layer is a little bit low resolution, looks a little bit blurry, and as you work down to the full resolution file through the compression layers, we get to the full crisp version of the image. And in practice, you can think of that a little bit like you see Google Maps. So when you zoom in and out, you get this kind of blurry stuff going on, and it eventually loads in until you get a nice crisp picture. And that's that kind of technology doing its thing in your browser. So this is what we're trying to use. What does it look like if we take our little film grab? The top layer may well just give you just enough information to be able to see what's happening. You get a gist of what the picture looks like, but you're not serving masses of data into there. And, of course, eventually you get down to the full resolution file, looking nice and crisp and sharp. The other cool thing about them is if we take a guess that this is, let's say, a 10,000 pixel square image, so I'm huge, you can tile it, and these are specifiable as well. What that means is that not only do we make layers of compression, but we can divide it up into little segments. It gives you a little bit of Rubik's Cube kind of look to the way your images are done. It makes it really, really scalable for resolutions and different displays and compression needs and all that kind of stuff. It's really quite clever. So when we come to serve our images up onto a phone or our different size computer screens that everybody uses out there, we've got a flexible product to work with at that stage. It's quite cool. So we might choose to just serve up, for instance, if we've zoomed into the top corner of an image, just serve that top corner, because that's just one tile, right? Or if we just need a very small browser window, we can just serve up the resolution that's necessary to serve that resolution on your screen of your phone, for instance. Very, very flexible. So we decided we were going to go with this and have a good play and set things up, which meant that myself and an ex-archive's New Zealand colleague of mine, Matt Painter, who deserves a big shout-out and a guy that's working with his in digital continuity archives at the moment called Jan Hutter. Both of these guys are myself. Big thumbs up. We worked on this for quite some time and saw a peer review all over the world. We went to Britain, to Canada, into Europe as well and said, look, we're thinking about taking all our tips for our NZDFs and turning them into these compressed JPEG 2000 files. What do you reckon? This is actually a grab from a bit of a tool that the British Library have done, and this is looking at the signature of one of the soldiers at 500%. So this is taken from an uncompressed tiff, a normal preservation format file. On the left-hand side there, you can see an uncompressed, so a lossless JPEG 2000 file. And here on the right-hand side, we've got a 10-to-1 compressed file. So for all of those guys sat down in the back of the room, we were sat going, I can't tell any difference. I can tell you if you get access to the full files as well, you can't tell any difference looking up really close, which is kind of the point. So we're taking an uncompressed tiff file and being able to make a compressed JPEG 2000 and they look the same. So, once we've done this, pros, scalable files in pretty much every sense you can think of, huge story savings. It's a very, very stable file format to be able to do preservation actions on. It's backed by a big ISO standard. You actually see this technology used in a lot of different places. Believe it or not, it's actually what they use in movies when they show them on projector screens at the movie cinema. So it's used all over the place and it's starting to gain some traction. The cons, though, it's not native to most browsers. You can't just bang a JPEG 2000 thing into an explorer and expect it to work. It doesn't. Usually it requires imaging software to be able to open them up, which is not something that everybody has, right? And the specs that you set up in how many compression layers and how much compression and the tiles that you use, they kind of work for a particular batch of records. So when you move on to, let's say, a bunch of maps, you're going to need to set it up slightly different. So there's some set-up work to do up front. But those limitations you can work around. The common way of doing that is to kind of use the JPEG 2000 files and let your image server just pump a JPEG of the size that you need into the browser live from your files all the time. So bringing this back to what Archives New Zealand has done with our new JPEG 2000 files, we're looking at somewhere in about 75% to 80% saving in file sizes over those original tiffs. And we're working with our digital continuity friends to be able to kind of put system viewers in place which really make use of those JPEG 2000 files, serving up the JPEGs that we need live into the browsers. But remembering, of course, that they are technically lossy. They're not the preservation format. We always thought they might be because we're preserving them physically afterwards. But they're visually lossless. So you get in all the benefits without the costs. But to take things further, we are still actually thinking about how we go further again and use the web to really liberate this approach and really hammer home the access side of things. So although it's kind of very much a work in progress, one of the things that we'd like to be looking at doing over the kind of next 12, 18 months or so is really getting to the stage where we can put digital access systems in place. And that's kind of a bit of work that we're working on here with that. So we grab a working progress from there. So I want to throw a statement at you guys now to get those cogs turning again. We're building digital systems that replicate a physical archive and it isn't working. Digital systems are being constructed with this ethos that things come in and they get filed nicely on a shelf and then the door closes behind them and somebody has to knock on that door and say, can I please have that archive back because I'd like to look at it. The rest of the world, the Instagram world where a thousand pixel image is more than good enough for everybody, the digital world isn't working like that. And I think at the moment we are very much in a sort of state of flux where we're trying to figure out how to do all that preservation stuff and access. And I think we need to have a look at this because the current model that we're working to isn't working. And I'll leave you with a couple of questions to take back to your desks, okay? So when you're done and dusted and we're all sat back in our own office if nothing else, give these thoughts a little churn over in your head to see if it can prompt things in your own work. Why are we paying twice to preserve something? Why do we pay to have a physical preservation of an object on a shelf? Then we scan it and pay to digitally preserve it as well. Why? Pick one. Have we lost sight of who our customers are? And by that, what I mean is that we as professionals archivists, librarians, curators and so on are we doing the work that we do at work because it's what librarians and archivists and curators do? Or are we actually looking at our customers, the people of New Zealand and the world and beyond and going, is this like what they want? Or are we just doing it to be self-serving in the way that we as professionals work? And a quote for you as well. This came from one of our senior managers. I don't know who, because it was anonymous, but it's from a bit of digital futures work that's been going on with the Department of Internal Affairs recently. What digital really means is that we become part of the global market. So when you guys go back in and you log on and you start doing your work again next week towards the end of the week, is that digital stuff you're doing honouring this? Because if it isn't, we probably need to have a rethink about it. My name is Dave, I'm from Archives New Zealand. Hopefully we've thrown a few little sort of off the wall little ideas at you to prompt a few thoughts. Are there any questions at all? Stun silence. So for the benefits of our video colleague at the back, the question was, is the film lab, once we get hold of it, going to be taking any commercial working? At this stage, I think so. That's the plan, is that we will be able to offer it kind of secondary to the work that we've got to do, but be able to provide a commercial operation to non-archive stuff, basically, yes. So again, for the purpose of a video down the back. So we've got about three questions in it, I think. So basically, the first one was, why not put in a digital preservation system so that we don't have to re-scan it all over again? The reality is that at the moment Archives New Zealand is going to be scanning these files, originally as TIFFs, and then migrating them into JPA 2000 files. And there's so many of them, and the scale of the project is so big that we will be putting them in a digital preservation system. The reason why we're kind of looking at this digital access system as opposed to digital preservation and posing that big question is more to do with the fact that we are actually already paying for them to continue living in a box on a shelf. So if we're paying for them to go into a digital preservation system as well, that's kind of duplicating that preservation side of things. And that's the kind of question that we're really posing, I guess. At this stage, because of the scale of it and because it's unlikely we'll ever scan them again, they're going to be going into that digital preservation system. But I think in terms of all the projects, the question I would go back is if your projects aren't quite as big and as scary as that, is the access system enough when we're already preserving it somewhere else? Jay. What do you reckon the lifespan is of the JPA 2000 format? Oh, thank you. What do you think the lifespan is of the JPA 2000 format? So we know it's uptake isn't huge, but we know it's got some massive benefits, as you say. The problem will be if we don't put them in a preservation system within 20 years. Are they likely to be dead and we can't do anything with them? We will have to migrate them, which means they're in a preservation system. Yeah, it's a good question. Realistically, I think TIFS, the kind of known entity for images is a preservation format. Personally, I'm kind of of that belief that any of this kind of work needs to be kind of looked at as in how to migrate from one format to another. Be that in a full-blown preservation system or a more access-orientated one. I don't know if you can actually say that TIFS or JPA 2000s or JPAGs or PNGs or GIFS or anything like that are going to be around longer than any other. These will be going into a preservation system. Yeah, yours. And we'd like to look after them. I think the take-up of the JPA 2000 thing is growing in strength because I think people are starting to recognise that it's as viable, if not more so, than something such as TIFS. So the signs are encouraging, I think. Cool. And that runs us to time, I think. So, thank you. And we'll see you soon.