 Keoratato, maenai James yna, Ellen ʻalon ganunni athau ba gadebiu saple-jack te rei te silegau pa mērana, maenai mērana te te dōgais kaipteru awa i anulatrizaňa kitionau pa gadebiu piairu beth na maipa caipteru ʻalon kai maiannou athau pa mērana teалu. Athau pa kai mērana niei gado ynni kaipteru. So, for those of you that don't know Ngātonga, which I hope most of you in here do, we're a new-ish archive, we were formed from the merger of the New Zealand Film Archive and the Radio New Zealand Sound Archive in 2012 and the merger of the Television New Zealand Archive in 2014. Public access, of course, was an important reason for bringing these three organisations together, and of course part of this comes from making collections findable and accessible online. We began a new refreshed combined website development project last year, and as I said, an important part of this project was working out how to present the new combined archive with one coherent access point. And I should just note that today we are going to be talking about bringing the film archive and the sound archive collections together. We'll be bringing the TVNZ collection on board over the next year or so. So, this whole process has made difficult because we had two quite different institutions coming together. Both had very different collection management systems and different metadata standards that the two organisations had developed. As well as this, both had quite different online catalogs, neither of which could easily incorporate the other one. This is a solution that we realised fairly early on wouldn't work. We couldn't just merge the sound archive into the film archive's online catalog nor could we do the same vice versa. So, when the organisations first merged, what you can see here is we brought both under the Natonga domain, but kept each online catalog operational until we could find a suitable solution. So, there's the former film archive and the former sound archive websites. The Natonga website for a long time was basically just a rebanded version of the film archive one. So, for about a year or so, we maintained two separate and quite different online catalogs. So, this is a look at the old film archive online catalog. There's a number of flaws with this. There was no Boolean searching. It was very difficult to refine searches. We had limited ability to embed video. We could only include very short clips on the website, and also it was difficult for us to update and to enhance. And this is the former sound archive's online catalog. As you can see, it wasn't a particularly friendly user interface. Again, it was difficult to refine searches and embedding audio with items was difficult as well. So, in terms of the back-ends that were driving this, there was also two quite separate collection management systems. So, the sound archives collection database is based on Vernon. They still work with Vernon. We've got various metadata standards here, PB Core, Library of Congress, things like that. All the sound collection, all the radio New Zealand sound collection items are accessioned into Vernon across the varying formats. That's that part of the collection. They have a documentation collection, but that's not yet part of that. And then the film archive was based on a bespoke and highly customized FileMaker Pro database. It was very loosely based around PB Core. But as I said, there was a lot of development of bespoke fields and bespoke metadata over the years that this was being used. So, what we had was a situation where there were multiple collection types being brought together from both CMSs. The film archive had a much broader range of material within that FileMaker Pro. So, of course, we had the Core collections of film, audio, television and video collections, books and periodicals, posters and stills. So, as you can see, there were, at least from the start, there are more other types of collection materials. So, there's also manuscripts and archives which you haven't brought into this yet. So, initially, there were seven different types of collection objects, types that we wanted to bring together. There were different metadata standards, different fields being displayed across both archives, which also complicated things. So, really the big problem facing us was how do we bring these together and how do we combine them and present the metadata from these two quite separate, quite unique collections and two different CMSs into one combined, coherent public face. And a really important consideration for us right from the beginning was that users coming to the new Natonga online catalog shouldn't know that the information being provided to them was coming from different sources. As far as possible, we wanted them to be presented with basically just a seamless experience for the end user. And, of course, there's also the changing digital context that the Glam sector operates in. Both of those online catalogs were five or six years old and, of course, a lot of things have changed in that time. And certainly, people's expectations about usability and user-friendly interfaces have changed a lot over that time as well. So, on the one hand, what we have is what you could call the Google effect or one search bar to rule them all. That's generally for the casual visitors or for those looking for particular items. But we also needed enough granularity for in-depth research and researchers who are looking to explore both collections depth. So, this, of course, reflects the broad users of our collections. So, we have, as I've mentioned, what we call casual users. So, they're people who are doing things like family history, they're looking for relatives, they're looking for things, programs that they've been in or appeared in, or just basically old TV or radio broadcasts that they've enjoyed over the years. There's also people doing, as mentioned, in-depth research for a whole range of different reasons. So, we've got people using our collections for production purposes. So, television and radio broadcasters and film and documentary makers who want footage or want audio to use in the programs that they're making. And, of course, you also have academic and other forms of research drawing on visual evidence and visual sources and the sound collection material as well. So, there's a lot of different users that we had to try to balance when we were coming up with this new catalogue. So, over the course of this work, we worked closely with the team at Boost who we had worked with to create the Anzac Sound website. And it was actually staff at Boost who also worked with Digital New Zealand that suggested that Supplejack, that suggested Supplejack could be a solution to the problem that we faced. And after a bit of discussion, we realised that that was a very elegant solution. So, what is Supplejack? So, Supplejack has been designed and developed by Digital New Zealand and it's a tool for aggregating, for searching and for chewing metadata records. So, it's used to collect metadata across millions of items from hundreds of data sources across many different data formats. It's an API which transforms messy data, creates a unified search index and makes consistent metadata widely available via an open API data service and built using open source technology by Digital New Zealand. So, this is what we came up with. So, the really important thing here for us was the ability to be able to export two sets of metadata from both CMSs as XML and then to be able to unify them within this new search engine and it was basically Supplejack doing the magic of unifying these disparate sources. An added bonus was that Supplejack was built to handle and to display digital items, which is an increasing part of what we provide online in the form of images, video and sound. And so, Alan will talk a bit more about embedding media and how that works shortly. And another factor that sold us on using Supplejack was that we could see its potential application of bringing together a diverse collection material for other projects, for other digital projects that we run, such as education and a refreshed media net, for instance. So, we could reuse the work that we were doing with the online catalog and these other digital projects. So, that was a big bonus as well. And of course, Supplejack was free to use for its open source licences and we could avoid getting tied into often expensive proprietary solutions. And I should just mention that recently Supplejack and Digital New Zealand won an award for best use of open source in Government this year's open source award. So, that contribution is being recognised. So, we knew that we wanted to use Supplejack. We knew that we had the different CMSs. We knew that we had the different metadata. So, to start the process of aggregating the collection records, we had to begin with a series of meetings, consulting with staff, of course, from across the archive and also started the process of metadata field mapping. So, this is an example of what we did as part of this field mapping process, basically just in a basic Excel spreadsheet. So, we identified the base fields that we wanted to display, then the common fields between both collections. We also combined things, titles and other titles and secondary titles and things like that. And we also had to figure out which fields we're going to be showing from collections. For instance, the sound arc I've had a much fuller range of subjects that we show whereas we don't tend to show topics and subjects for the film collection material. So, overall involved 19 metadata fields and seven different types of items across both collections. From here, we exported XML sets from both CMSs in preparation for harvesting by Supplejack and now I will hand over to Alan to discuss this further. Hi, everybody. So, I'm just going to give a very rapid fire tour of the back end of our online catalogue and the various different sources that we're asking Supplejack to harvest from and bring together online as a single-point search. So, this is a Supplejack dashboard. It's the back end of our online catalogue. It's all, as James mentioned, built in open-source software so we were able to take a product that was an existing thing that Digital New Zealand had developed and then work with both developers to do some customisation of the codes, the harvest codes to go out and connect with our various sorts of collection metadata. So, I'm just going to show a quick video of the Supplejack back end. So, within the Supplejack back end, there's a bunch of different types of harvest scripts. So, these scripts go out and look for videos in Vimeo, audio files, Vernon information, FileMaker database information. They're on schedule, so they look for this information at regular intervals without us having to do very much, which is very handy. And just to look in a bit more detail at some of those harvest scripts. So, there's a bunch of different ones for video and audio. This is the Vimeo script. It's in the Ruby programming language, all based on open source coding. Here is an example of the script that goes out and looks for the information, metadata from our FileMaker database about films. And here is a list of scheduled jobs, so things that have run when Supplejack was scheduled to go out and look to updates to the collection information and new video files and audio files added. So, these Supplejack scripts pull information from five different places. There's one script that goes out and looks for FileMaker updates, metadata about the film and video collections. There's another one that goes out and looks for Vernon updates about the sound collection. One goes and looks for new Vimeo videos, which is how you're able to play videos on our website. Another one goes out and looks for MP3 files and audio, and then the final one goes out and looks for JPEGs of posters and stills. So, this is information being exported from our internal FileMaker database. Then the script in FileMaker runs a harvest to go and look for that new XML file and it's updated instantaneously on the online catalogue. So, that runs like that. We're using Vimeo to host video on our website and we did quite a bit of research into video players before we made this choice. I'm going to move on to the next slide, I think. So, we did look at creating a bespoke video player to play video on our site, but eventually realised that there are a lot of advantages to using a third-party platform rather than building and maintaining our own bespoke player. So, some of the key advantages of Vimeo were that Vimeo was designed with the Filmmaking community in mind, so it has very high playback quality. They have very good support systems, so if anything is not working the way that it should be, they're very responsive about fixing it, which takes a lot of onus of maintaining a video player off us. Vimeo videos are responsive across devices, it serves up different videos for mobile users versus desktop users. It also can tell what level, what somebody's level of internet connection is and serve a lower definition if they don't have very strong internet connection, so it's very good at serving people an appropriate video for their device and their level of internet connectivity. And another thing that was really important to us is that when you upload videos to Vimeo, the rights are retained by the original rights holders and Ngataonga Sundan Vision doesn't actually own the rights to any of the items in our collections. These are all maintained by the original depositors, be they somebody who deposited their home movies or a big production company like TVNZ or Radio New Zealand or another distribution company. So we were really cautious that we didn't want to be giving away any rights to any of the collection items and posting them in our online catalogue, but Vimeo does not subsume any rights ownership, they stay with the original rights holders and they also have very good security and levels of password control over videos and things like that. It's also significantly cheaper to host mass amounts of video on Vimeo than it is to self-host on your own site server. So if you start getting into the thousands or tens of thousands of video items, that can become very expensive to host yourself, but Vimeo is very cost-effective. Vimeo also hosts the bandwidth, so there's no limitations on number of users and players. And another thing that's really cool about Vimeo is there's a good amount of customisation of the video player, so we've managed to create a lovely branded video player, which is in the Ngatonga Style and Turquoise and Fitsyn with the rest of our branding. So this is a video of the back-end of our Vimeo account and you can see that there's an OC prefix on all of the videos. And that's how Suppleject knows in the Harbour Scripts to go out and pull those videos into the online catalogue. It goes and searches a Vimeo account for everything that has OC in it. There's an example of a video with the OC and reference number title. It matches the reference number up with the reference number in FileMaker. We host the photographic items separately to the videos. These we self-host on servers at Ngatonga. And again, these are labelled by their reference number so that Suppleject knows to match it up with the right record from FileMaker. It runs the harvest, matches the media item up with the reference number and they're all unified on the online catalogue. Same thing with audio files. We self-host these on our own servers and name them by reference number so that Suppleject can find them and recognise newly added material and link it up with the proper record. And it's the same process of running Harbour Scripts on the media items as well, which Suppleject is scheduled to run Harbour Scripts every day to pick up new videos and audios that we're adding to the catalogue because we're always growing. The number of data sources is there online. We're adding new content almost daily. And so that is what the back end of Suppleject looks like. I'd encourage other organisations to talk to Digital New Zealand about it because it is quite easily customisable and Suppleject can harvest from any number of different data sources so various different types of databases or even things like Excel spreadsheets. Very flexible. Any number of different sources, Digital New Zealand uses it to harvest from hundreds of different sources and bring these together. So this is the public interface Suppleject's doing a lot of work, automated work behind the scenes as you've seen and this is what the public sees which is quite streamlined. So this is Ngatong's online catalogue. There's most 300,000 records in there. So when you enter the catalogue you see everything then you can refine your search. Question mark brings up some really detailed search tips on how to run Boolean searches, how to use fuzzy logic, how to refine your searches and things like that. But this is really designed with dedicated researchers in mind. It works equally while just running a simple search by entering a search term into the main search box. This pink available online filter enables you to refine your search to only things where you can see videos or listen to audio or see images online. So that's quite useful and here's an example of some of the videos that we've got online. We've got quite an interesting range dating from 1896 for on tool. A couple of years ago so it's really eclectic and diverse. So those are some of the videos that we've got viewable online. Again you can filter by audio that you can listen to directly online so that's what I'm doing now. We've got almost 500 audio clips that you can listen to from across New Zealand audio history adding more all the time. And here's some of our documentation items that you can look at online. Quite a few film posters and some film stills as well. So if you want to do something that's a bit beyond a basic search in the search box you can then filter your search. So here's an example that I've listed in Catherine Mansfield. So the initial search brings up everything that Natanga holds about Catherine Mansfield. A lot of them are in formats that are not directly viewable or listenable online. So then if I wanted only things I can watch and listen to directly I used the available online filter to narrow that down. And here I have a set of material that includes a few videos relating to Catherine Mansfield and oral history and the poster. And within that I could focus just on film or just on audio or just on images. Here's an example of a Vimeo branded video player and what that looks like in the public interface. If I really like this item I can star it as a favourite. Go back to my list and then I might be interested in this history about Catherine Mansfield as well. So I click into that and here's an example of our audio player. And this is the metadata brought in from the Vernon database. I also add that to my favourites list. And there's also a poster about a film relating to Catherine Mansfield's life as well. And that's one of the JPEGs which Supplejack is harvesting from our internal servers. So now I go to view favourites and I can see the list of items that I've added to my favourites list. And I can either email Natonga's sound and vision and ask about them. Or I can send it to myself or a friend and share my favourite items with them. So I know that we're about to run out of time so I'm not going to spend too much time on this. But this is about the request film that we built that's part of the online catalogue. As James mentioned, Natonga has quite a diverse range of users which include people wanting to reuse material in new productions, people just wanting to watch in the viewing library, people maybe wanting an education screening. And all of these different users have a different team within Natonga's sound and vision who assist them with their enquiry. So we had quite a complex tree of email directs for the form in terms of the enquiries needed to go within the organisation. We distilled this down to something very streamlined with two key breakouts within it. And as James mentioned, we've just launched the new online catalogue at the start of the year. It was built in a job process which means it is continually developing. We've got a feedback button on our website where we capture feedback from our users, we keep a log of these and this is a way of prioritising our ongoing development of the website. It's really useful for Natonga's sound and vision as a charitable trust to build in an agile way because having the user feedback and continuing to build the product based on that means that we are able to be responsive and use our very narrow budget on things that our users want. We'd love to hear your thoughts about the website. Please do contact us using the feedback tab on the left-hand sidebar of the website. So this is only phase one. We've got lots of things planned coming up. We've got ideas for more sophisticated means of refining your search results. We're adding a lot more video, audio documentation that can be watched online. We're also building online exhibitions themed sets of audio and video on topics relating to New Zealand's social history. The first one is going to be about film and TV and radio advertising in New Zealand. Do you want to talk about the digital library? Just very, very briefly. One of the most exciting things about Supplejack is the ability to bring in these different digital items from different sources. We're putting a new SAN in on-site, which is going to include a viewing server, and we're hoping in the next year or so that we'll be able to have Supplejack basically powering people's ability to view everything that's in our viewing servers and our digital libraries and our Wellington office, Christchurch as well. Thank you very much.