 Aiziraiwans just pointed out, Accelerator Collections Digitisation Programme is a bit of a mouthful. So we call ourselves ACDP for short, nice and short acronym. ACDP has been running for a year and I'm going to talk through the processes of actual, the technical aspects of digitisation, the equipment we use and sort of successes we've come across in that time. First of all, sort of why digitise, I guess. I won't stay on this for very long because I think we all know why we're digitising. I mean, it's very common practice these days in Klam institutions. There's a lot of demand for it. The public are looking for more digital content and institutions want to get their collections out beyond their wall. So that's a pretty common theme. It's probably two and a half million collection items. That's an estimate. We have one million registered items. The other one and a half million are either bulk registered natural history specimens or unregistered at this point. So bearing that in mind, we estimate only 12 per cent have images. So there's an awful lot of work for us to do. But the last bit, and Flora actually mentioned this, I think it's actually an important part of why digitisation is pretty popular at the moment and she used a nice term, which was robust technology. I thought that was really important because it is about the technology that's available now, means that the assets that we create really will have longevity and I don't think you could have said that five years ago in the digital realm, but I really feel like we're there now. So the team, we have two imaging technicians. Obviously they're running the equipment, taking the photos. We have one data technician. It's quite interesting, again, with Flora, what she was saying about the ratio of the team that you need. It's not just people at the coal face taking the photos. You need quite a large support team as well. And in fact, in larger institutions and digitisation projects, the ratio is normally about three staff to each technician. It depends on the state of the collections at the time. But we have one data technician and they would load all of the photos into our database which we use MU here at Te Papa. And we have a 0.5 collection manager who delivers all the collections to our imaging technicians, helps with handling and any other collection object related issues. And a 0.5 rights officer and this is quite important as well because when you're digitising, you want to get your stuff out beyond the walls. So it's got to be cleared for copyright. And so we want to make sure that everything that we're digitising is cleared ideally before we get it. That's not always the case, but we have somebody working your way to make sure that we can share this beyond our walls quite readily. So here's one of our workstations. It's a flat copy, as I say. So the ACDB project really is 2D. That's what we've worked on for the past year. And this is our flat copy workstation. As well as investing in quality people Te Papa also invested in really good quality equipment. We use Phase 1 equipment, which is a medium format digital camera. Really good daylight lights you can see there. So everything's incredibly locked down and there's a very tight workflow that we work to and I'm going to talk about that in a minute to make sure that the files that we create are really the best quality files. So it's about creating digital assets that are going to last and there's a process behind that. We use Capture One software, which you can see there, and run it all on max, which is quite nice. So when you're looking at a lockdown workflow, you have to sort of establish some standards and when you're trying to establish those standards, it can be a bit of a rabbit hole, basically. We export a lot of different areas. We try a lot of different things. Fortunately there's two sets of standards out there in the world that one's out of America. It's called FADGI and the other one is out of Europe and that's called Metamorphose. Now these guidelines, they are only guidelines because there's no actual standards you can apply, but there's very good guidelines and these two sets of guidelines are reassuringly similar, which is great because you don't have to choose one or the other. If you choose go with one way or the other, then you know that you're going to be working on a similar level. And we chose FADGI standards out of America and it's got a ranking system of stars so it's from one to four. A four star capture is the highest and the best you can achieve. And we set up right from the beginning with us to be working at the very highest international standards or above. That was our goal. So the first part of that recipe is resolution. Resolution for a FADGI five's four star is 100 ppi or above per object. Now with a digital back that we have, which is 100 megapixel back, that's relatively easy to achieve for average size objects. In fact, we're generally working well above that four star rating and we don't limit ourselves to that ppi either. If we can capture above that, we will. There's no point in just using half a sensor, for example. The other part of the workflow is, and this is critical, is colour calibration. You'll see that this is a capture one window at the top left hand there. We use digital colour checker charts so we would photograph that chart in our setup. We would take that image and send it up to a website called DELT AE there. That website is designed to analyse that chart so it knows that chart, recognise it, knows what the numerical values each one of those colours should be and will give us a report either it's a pass or a fail mark. If it's a fail, it'll give us a profile which we download. We put it into capture one, we take another photo and then we reload that new photo of the checker card back up into DELT AE and at that point that should give us a pass mark. This part of the process is really critical, I think, especially when doing flat copy work, to make sure that you're 100% accurate with the captures you're getting, that your colour accuracy is bang on. Another check we have within capture one software, and this is available in a cultural heritage only version of it, is an overlay. This one on the bottom here has got a white overlay. It's a digital reproduction of the chart. Each one of these people actually has a little hole in it which will show us what the chart looks like from behind. Visually we can check straight away to see if our colours are in line or not. This is part of the process of really making sure that you're capturing the most accurate version of that work in front of you at that point. The last part of the equation is file types. Not everyone comes to the same conclusions on this. We have a recent survey online about colour workflows throughout other cultural heritage institutions, and that's the most common one currently. File types, we keep TIF files only. We keep a 16-bit preservation TIF, which goes into our preservation repository. So it's very deep, no-one gets to it. It's there for longevity, and that file is intact as possible. It goes through very little or basically no processing whatsoever. None of our files goes through Photoshop. It comes out of Capture One straight into KMU. We also have an 8-bit access TIF as well, which goes into KMU and is available to staff and also goes to collections online to create the JPEGs for that interface. Once again, keeping only TIFs is a common workflow for other institutions. We don't keep the rules because rules are a proprietary file, but also beyond that, because we've locked down our Capture process so tightly, I'm very confident that should we come back in five years' time if we had that rule, we'd never get any more information out than we're getting at that point. So we make sure that our 16-bit preservation TIF has all the data you could possibly ever get out of that raw file, and there's no need for us to keep the rule beyond that. So once our processes are in place, the next thing was, well, what are we going to shoot? How do we start? How do we kick off this? So we started with postcards nice and simple, easy to deliver. We had 3,000 of them, we shot front and back. It allowed us, by shooting those 3,000, to lock down our Capture processes, lock down our colour accuracy, and our workflow beyond that, so from here loading into KMU, et cetera. Interestingly enough, the copyright around these is different from the front to the back, so we were able to release the front of the postcards online quite quickly, but we couldn't release the backs until they went through a different process which is now gone, it's gone through, and that's because it's got personal handwriting on it. And the next choice, so again, I'm sort of talking about how we make our choices of which parts of the collection are we going to do. So the next part we did, we went to our works on paper collection, and we started doing all the watercolours within that collection. That's because Te Papa is a large contributor to Watercolour World, which is an online initiative out of England that has Prince Charles as a patron, and they want to tell the story of what the world used to be like from watercolours. So we were immediately looking for a second output beyond our own collections online, so we're looking, we want again to spread our assets as wide as possible, they're not just for collections online, we want them to be shared across the globe hopefully. And beyond, after we had finished the watercolours, we moved into the rest of works on paper, so our works on paper collection was actually reasonably well documented, so we had a whole lot that still didn't have images, so we decided well let's try and finish it. So we started doing the works on paper, and we finished that, we had eight and a half works on paper to finish, and as you can see there's quite a bit of variety in the works on paper. So that's the flat copy workstation. Now we're into the transmissive workstation. This is one where we shoot all our negatives and transparencies, and once again we have new phase equipment, particularly a new camera. We bought this a year ago, it's the Serial Number 52, you might be hearing a little bit about that at the moment. And when we bought it a year ago, this is Serial Number 52, so it was a 52nd one ever made, but when we bought it there was also no user manual because it hadn't been written at that point. So it was a bit of a learning curve for us, didn't really look like a camera when we got it, but we're really thrilled with the way it's behaving, super reliable, super high res, and just excellent. The lenses are also made specifically for flat copy work, so it's about again trying to work to the highest quality we possibly could. Now just to give you an idea, so again it's another 100 megapixel back of the resolution of visual cue on the resolution of this sort of camera. Here's a photo of a glass plate negative that we've done on that work stand. If you keep your eye on the top left while we go into CSI mode, and we go enhance, enhance again, enhance again. There's a tiny bug that's been squashed into the emulsion. Now we think this bug might have crawled into the emulsion as the photographer, who's James McDonald, was putting the emulsion on in the field. The theory, we're not 100% sure, but the resolution is remarkable on this camera. So how do we prioritise what we're going to shoot on this workflow? So what we do with our collections is it's broken up into photographers, makers. So we started with, we break it up into makers, and we started with James McDonald, who was working at the Dominion Museum in the early 20th century, and he travelled around New Zealand with Elston Best, and took some pretty remarkable photographs actually, really amazing content for that era, and it was a real pleasure to work on all those. We finished all those. That only took us a couple of months to do that. And then we moved on to Leslie Adkin after that. So again just another maker. We hold quite a lot of Leslie Adkin negatives. I think we hold quite a proportionally quite a higher number of hits than any other institution. I think we've got 9,000 negatives, and a lot had images, but quite a few didn't. So once again, we wanted to finish that selection. So we worked through the rest of Leslie Adkin's work. His is more obviously snapshots of more recent New Zealand life. But one thing we did notice is a bit of a love story. This is his wife Maude up here in the corner, and here's a whole selection of photos that were shot just after they've got engaged during their wedding, but before children. They're really quite lovely. And after that we moved on to Rumsey. Again, so we're trying to here mix up the selection process. So we're moving into a different era, moving into colour. So we've got some different variety in the content that we're putting out there and creating. Another part of the workflow we do is proof sheets. Well that's, for us, there's different reasons for that. So in this particular proof sheet, for example, it's represented by one record in our record system. So having one photo is the best way to represent that one record. Also improved in 35mm. There's a lot of repetition. There's no real need to digitise every single frame because sometimes things can be identical. Shooting it this way allows our curators to assess a role of film and then select the ones that they might come back for a higher resolution scanning at a later date. It makes it all available. So these will go on to collections online. That means a public can see all of these images at once. Here's another one, just out of interest, and I'll show you the usability of these even though they're proof sheets. This is a selection of images by Glenn Jowett and it was following a mongrel mob crew who were on a pee-piece outing in the 1980s, which is physical work. So they were going out to do some scrub cutting. This is them on the way out. So this is just a crop from that proof sheet. This is them on the way back. Obviously a bit tiring. But when I crop into one of those proof sheets, the single frame image can be printed at A4 at 150ppi so there's still really highly usable images. And through this process we've made 40,000 frames, individual frames visible through 4,500 captures. A bit of a flashback moment. This is us digitising a 10,000th collection item. This was way back in February actually and just happened to coincide with Te Papa's 20th birthday it was. That's the image there. So where are we at now? We're a little over a year so it's six months after that point. We're a little shy of 30,000 items have been digitised in that time. The ACDB project in terms of measurables was asked to increase the rate of digitisation within Te Papa by 200% will be doing well. 300% will be doing very well. We actually increased digitisation by 700% on the previous year so we're doing very, very well. So that's our process really. So I can open up to questions. That's a good question. So we shoot with that colour calibration we calibrate every month, monthly basis. We shoot the colour target three or four times a day. But if we've got a batch where the lighting doesn't change, then we shoot one target that sits to the side. We shoot each image as high a resolution as we can and we load that target alongside each of those images into KMI. We've always got a target. It's just not in the photo but it's shot in exactly the same conditions and that allows us to maximise the resolution of the images that we're capturing. No, we don't associate any profiles. No. Not in the data that goes into KMI. With that amount of profile that you're doing, have you noticed any actual drift? No. Now I would say that we aim to do our profiling every month. We haven't achieved that every month. It might slip off. So we did have a gap there. I think there was three months went by. We went back and re-profiled. No change. And it is down to the light, the equipment you use. So those lights were used, they did cost a little bit but they don't move. Nothing moves and that's heartening for us actually. So it's worth it to invest in that proper gear. File naming. People here want to know that. Well file name comes from the collection object first and foremost. Everything that we shoot is registered. So it has a registration number. We call it that registration number. And then we've got AM for access master, PM for preservation master. For us it's a little bit different from your workflow in that normally there's only one or two images that we have and so AM and PM is sufficient to differentiate between those two that are loaded. Are you shooting a colour chart ruler accession number shot with everyone or is it just purely that? No it's just purely the work. Because we can link the colour chart that is associated with each of those sessions to that image in our database, we don't need, we don't feel the need to include it in every single image. And then our registration number is enough for us to know where that belongs so we don't need to see it in the actual image as well. And no one needs to see a ruler beside them? No, again it's down to the data that's in KMU so all of this stuff is already measured. It actually alludes to another question when we were looking at doing slides, so mounted slides often have information on the outside of them. And so our question to the curators was should we capture this? So should we do two captures? One of the image and one of the slide mount, try and put them two together so we're not slowing down our workflow too much. They came in, they assessed it and they said no don't do it because we already have that information in KMU and in KMU it's entered so it's searchable. A photo with text is not searchable. Anyone else got a question? I've worked in the dark room with real negatives and you know about interpretation. Have you thought about keeping the raw files of the negatives so that in the future they can be reinterpretated or even open for public to reinterpretate as they wish? So the preservation master of the negative is in negative form. So once again it has all the information that's in the raw file at that point and that's one that goes into our preservation repository. That's the one so you create a TIF or in our cultural heritage capture one software you create a derivative and you then invert that derivative and you make your changes. That's the same process that so you're essentially using that raw capture at that point or the raw TIF and so we can go back and do that at any stage later on because that process is incredibly intuitive and flexible still. Cultural heritage software that we have has inbuilt profiles but they don't work because film ages, it's not the way it was when they wrote the profiles and it's a good starting point but then so is just a simple recipe of invert, flip and then do some levels. Sorry. That's alright, you're just putting me under the hammer. The capture on software you're running, is that running on a standalone computers or are they on the enterprise network? No, they are on the network. Are they Windows PCs? No, it was great actually. That capture one software is only made for Mac so we were able to tell IT we must have Macs for this workflow and so it's working. I know that you guys have tried this before and found it possibly unstable but we have no troubles whatsoever with it so really stable, amazing piece of software for capturing. Thank you very much Michael, I hope we should show our appreciation.