Hammer Channel is a living archive of public programming at the Hammer Museum. Using open source principles, access to the Hammer Museum's exceptional public programming is now available for the world to enjoy online at https://channel.hammer.ucla.edu/. In this panel we'll give an intro the to project and describe how your museum can benefit from the work to date. Track:Interpretation & Storytelling & Education
Unknown Speaker 24:47
So hi, thanks for coming everybody. And it's nice to see some people's names I know, good to see that everyone's everyone's joining us today. And today, we're going to be talking about hammer channel, and the video platform that Hammer Museum and hosts, and their, their events program on. It's an open source video platform. And so we're going to give you a rough introduction to who we are and the project itself, you know, how was conceived was built, and sort of where it's going, and then we'll ask you guys to ask us questions so there'll be a session. Probably most of the second half of this session, discussion, which will be fueled partly by your questions hopefully so please do add questions in the chat and we'll keep an eye on that. And in response to what we're talking about, or general general things as they come up in the discussion. Okay, so I think we'll just introduce ourselves. So, you want to go first.
Unknown Speaker 25:51
Yeah, I'm Philip Leers I'm the project manager for Digital Initiatives at the Hammer Museum in LA, and I was the project manager for this for this project for hemichannel.
Unknown Speaker 26:03
Unknown Speaker 26:04
Alright. Hi everyone, I'm Neil. Senior Developer at Cogapp down in Brighton, on the south coast of the UK. And I was technical lead on this project, working with Phil at the Hammer Museum.
Unknown Speaker 26:21
And Mandy, Andy Cummins, I'm one of the directors at Cogapp, and Cogapp the digital agencies in Brighton as Neil says, we've been working with museums, galleries, libraries, etc for about 30 years, a long time. And my role on this project was as a kind of directing the project from Cogapp side, and governance, those kind of things. And yeah, so that I can 100 this project.
Unknown Speaker 26:54
Yeah, I'll give a little bit of context for from our side this first quote that you see is our mission statement, The Hammer Museum believes in the promise of art and ideas to illuminate our lives and build a more just world. And we do use that mission statement, as much as possible to guide what we're doing. Hammer channel itself came out of the second phase of what eventually became a six almost seven year Digital Initiatives grant from the Mellon Foundation. In phase one, we focused on the hammers exhibitions and collections. And when we had the chance to, to renew the grants that add another three years to it. We, we knew that the public programs were kind of the next big column of what the Museum does what we present to our audience and our visitors. So we know we wanted to focus this project on on our history of public programs which has gone back over, over a decade now. The videos and hammer channel go back to 2005 So, more than 15 years old. We, we saw this opportunity as a, as a, as a chance to further our mission through the through artificial estate that's the Cogapp term for it. I love that. But we really see the public public programs, as, as the main, the main conduit for that ideas side of art and ideas. And we wanted to create. We proposed creating a tool that could basically take all of our history of videos and put them in one place, so I'm going to talk about the planning and goals from our standpoint. This is a three year project that we've got 15 minutes or so to talk about. So, I will. I'm going to try and give a broad sweep and then maybe hopefully get into some of the nitty gritty in the questions but I pulled out some landmarks from the process of the three years so the project goals here are from the actual grant proposal, and it's fairly simple we want to ensure the long longevity of our video assets both digital and we had some that were on on tape. Want to develop internally an infrastructure for dealing with this massive collection of video. We wanted to create a public platform, which is obviously what hammer channel is and then lastly to disseminate and share the project. So, entering the project. The issue that we had wasn't that the programs weren't documented they were very well documented. We had hundreds and hundreds of videos on our YouTube page on our Vimeo page on our website. The, the issue for us was the goal for us was to, to, to basically gather all of this together in a way that we can present the full scope of the history of our programs. Even though we're a contemporary art museum. Our public programs run the gamut from contemporary art to politics and poetry readings and performances. Talks on the environment, on technology. And though we were, we've had a great we have a great working system to documents, everything we felt that we weren't really taking full advantage of this giant collection that we were building of topical, you know, hour to two hour long videos with people who are at the top of their field. And we wanted to present them in a way that we couldn't through, you know through YouTube or Vimeo or on our way to have it all together so that people can search within and really discover everything that is that we've been working on for the last 15 years or so. So, our first, first step really in this process was having a content strategy strategy meeting in August of 2018. The project started in, I think, May 2018 where we pulled in the main stakeholders and we talked about I of course like all great
Unknown Speaker 31:32
web projects that starts with a whiteboard and a bunch of sticky notes, but we did some, some basically group thinking and some some exercises to identify what our audience is that we want to target and what our goals are for the for the visitors, you can see here we ask people to answer the three, you know the three prompts I want users to feel this I want users to know this and I don't want users to feel or know this. It was an interesting process of doing this, the audience we found for that we kind of could identify for this project is much broader than, then our projects tend to be usually our audiences are for kind of eyes when I say our I mean, on the work on the Mellon projects working for towards a more academic audience but with this we saw a much broader interest in from, you know, people who are not just educators, or, or scholars or things like that but people who are just really interested in the news, people who are just generally curious, you know, journalists filmmakers so we, we kind of took this as our first step we used it to draw together a document of of goals and use that to to power an RFP. We don't have a staff who can, who can handle a web development project like this we don't have developers, and we were looking for a new developer development partner to work not just on this archive, but also a full replatforming of our website onto a new CMS. So we wrote a proposed an RFP in 2019, and went through the process of that. And in May 2019 Finally we, we came to an agreement with Cogapp to begin and that really was a huge moment in in opening up the possibilities of what hammer channel can be.
Unknown Speaker 33:46
And, and that's where, you know we we came in in earnest, but we'd known people from the hammer for years, three MC n, you know all the lovely conversations in the corridors and drinks and karaoke and all that kind of stuff. But what, so we knew about the hammer and we knew the good work that was done there. But even in knowing that, As a company we have our own principles quits to a mission, I suppose. And we're here to build a better online world, you can read this yourself, but roughly speaking, when we get the opportunity to go for a project like this project, we try and see if we're the right fit, and if it's something that we are the best place people to, to help with. And so we, we hold the various things up, and the various facets of the project, up to our own principles. And then on the next slide, I've just pulled out briefly where we saw the kind of where the overlaps were. So we knew that the hammer wanted. And this robust video storage and management I've written there but, you know, some way of making sure that the disparate sort of organic growth of the programs and that, you know, rapid growth of gigabytes and gigabytes of video are rationalized and stored, and sort of provide a good platform for keeping it that way that the future. So that's kind of better for the people who use our work so that's better for our clients. So, that's what the hammer are getting out of this in the back end. And then, for the people who use the work and what Phil said, you know, you could see a lot of this content in various places, but it was through the YouTube interface or algorithm you know you could watch a video and then be sent somewhere else, And you couldn't really explore the wide range of things that Phil's just described. And it didn't look hammer and you know look YouTube or Vimeo, or whatever. And then another facet of our principles is this leading our industry forward so when we do our work. We hope to not just deliver for our clients but to kind of come to places like this talk about our work and hopefully, and inspire others and make connections where we can learn from other people that kind of thing. And on this project. One of the key things with the Mellon funding is it has to be open source so we like that idea. And we like that it has to be accessible, we're big proponents of that since. For a long time, in particular, and 2012 when we did the Olympics website. And then extensible, we always try to make our work extensible so we're trying to lead the industry forward in this open source project and we're hoping that you guys will tell other people about it today. And then we also, you know the kind of. Last but not least part of our brain, to try and you took part in our community. And it's obviously that's what the hell you're doing as well. So we knew that they'd be a good and a good fit, and when we could bounce off each other well. And so next, we've talked a bit about the context and how we, how we ended up working together but so it's just going to talk about the actual work so we you know, we know we're engaged with one another. And, um, you know, quite far apart different sides of the Atlantic. But Phil is going to talk about the content and all the work that he and the team did there.
Unknown Speaker 37:29
Yeah, again this is like seeing three years of work in eight bullet points is really terrifying for me but I just wanted to kind of make a checklist of the various things that we had to do over the course of this time over the course of the three years and each of them had their own challenges and everything took longer than we thought or shorter than what they you know like everything ended up being a bit of a surprise but first we created an inventory of all the, all the videos we could find at the hammer across all the platforms format storage locations. To see what we're working with, and we have almost 2000 I think total video assets they were able to find that later went through kind of a process of of trimming down and prioritizing. We had about 700 I think videos on mini DV, which is ops, if it's not obsolescent it's, it's obsoleting, that's definitely not a word obsolescence. So we had those migrated ascertaining rights status for, for all of these videos was a giant challenge. Fortunately, you know, we, we have contracts that we signed with the people who give the talks so generally we have releases, but going back as far as we did made it pretty tricky prep videos for upload that is like what forwards that to, you know, weeks and weeks. Some videos had to be edited, we had to do title cards for some. We had to download them from wherever they were or source them from wherever they were, to prepare to automatically upload them into our into our dams, and eventually we also had to select thumbnails for all of these videos which is something that I didn't think about that took a lot of my time. And then and then Metadata was its own. Huge, huge, huge projects that, that we ended up ended up becoming a massive team effort thanks to COVID. When COVID arrived and in our lives when it stopped things happening at the museum. We were, you know, what what we thought we were in the kind of final stages of trying to finish up hammer channel. We had a period right when this happened when we started working from home, where there was a sense of, oh, hammer channel is is perfect for this moment. Can we make it happen like right now, which we knew was impossible so instead we, we took advantage I guess of the of the time, and that we were given by COVID. We push back, or our planned release date from the fall to the spring basically. So that what we could basically take advantage of a lot of people whose jobs were made difficult to do because they couldn't be on site so people who are working in our visitor experience team, people who work in our store. People who do AV for our programs. There are a lot of people under contracts, who basically were unable to, they weren't able to do the task they were hired for, and so we were really lucky that we were able to borrow these people to help us with this massive Metadata project we had as MANY as I think 12 people on the, on the team, working and I put a little screenshot of our Google Doc, this is like 1/80 of the actual document. This is like, we created these core categories you can see if title people topics which is sort of like general, a list of 1312 general themes that we thought, kind of categorize the video as well. Tags descriptions and playlists, I actually think we have at least one person who was on this part of this effort in the on our participant list today. So this was the, I mean, I don't I can't get into it too much but it was just a gigantic, gigantic process.
Unknown Speaker 42:01
And on top of that we also, since accessibility was a big focus of this we wanted transcripts for every single video that we put online. And as we were planning, like our plan was to launch with 1000 videos, and it was very quickly became clear that the only way we're going to be able to transcribe that that scale is by using automated transcription, we did a lot of research on what what we could use and we ended up with a service called Trent, which works kind of similar to otter which were, which we're using today here. So, yeah, we had to basically run transcriptions of 1000 videos. We also had to, to develop a, a sense of being okay with things not being perfect, the transcripts we knew would not come out 100% Right. We were hoping for 90%. But that's a lot of cleanup, and we knew that we weren't going to be able to very quickly became clear that we were gonna be able to edit every single one of these. But I think in the end we ended up editing almost 300 of them, and they usually it usually takes about twice the length we found of the program to edit the thing and so our programs can be two hours plus. So you can imagine this became a huge project but again it was a really lovely COVID success story of everyone coming together it was a really wonderful team, and it kind of like doubled what we thought we could accomplish with him in general there's, I don't know what we would have done had this not happened because it would have been me and a couple other people doing this, but I can tell you it wouldn't have launched with 1000 videos. Sorry, I should I should segue Neil's gonna talk about the technical development now.
Unknown Speaker 44:01
Unknown Speaker 48:52
And so that that was you know the content and the tech stuff. But then one other, you know, I'm not a designer, nor is Neil, I don't think Phil is, but it's probably that the mid now, but there was a whole design process and UX process that was followed as well. And this was my colleague Gavin's work here at Cogapp, that just to say a quick note on it, you know, it was an iterative process and pretty low phi we kind of like to prototype things, see how they feel and we use paper and pen but we also use things like, you know keynote on Apple slide deck, things like that just to mock stuff up quickly to see how things feel. And then we tie that in with the work that Phil was doing. And then some of us have code that we're doing and populating those big spreadsheets, to see if the interfaces that we were coming up with were actually, you know, supported by the content that we had. And if it wasn't, Then we would fill in those gaps. So, like, the transcriptions, you know, that supercharged the search for example, and it's accessible or accessible I've already mentioned that we wanted it to be fun to use. So there's lots of animation in there, and you know it's it's quite slick, and then I just noticed as I was doing this slide, and it's the same color as the MSDN templates, almost like we planned it. And so Phil's going to show you how much I don't know. And then after that, we're probably going to start talking, and, Well there's one more slide and then discussion so please do that any questions because that's what we're going to talk about, so add any questions into the chat, please feel free to fill.
Unknown Speaker 50:34
Yeah, okay. So, without further ado I'm going to just do a quick kind of run through of hammer channel, and obviously this is accessible to all of you, so I won't, I won't try not to belabor it. But yeah, going, going off of what what Andy just said. We design what we, we ended up, I think, going with the simple, as simple as possible because one of our concerns was making such a big and broad collection of videos feel not overwhelming. So this is our homepage, which we basically got three types of pages we've got the home page the search page and the object page the video page. The home page, we wanted to build on a, on a kind of a reasoning of of browsing and discovery. So, along the bottom of the screen you'll see we've got, I think we call this the topic trail. These were the the 12, the 12 categories like I said that we that we drew out of our, our collection of videos and we thought it would work well as kind of as a signpost but also as a quick bird's eye view of like, this is the stuff that we talk about it's not just art. If what you're interested in is social justice, you can jump down to this area and see what we've got here. We've got these horizontal bands that are very familiar I think from using Netflix and, and YouTube and things like that. It allows us again to like to show broadly and deeply so we're able to put a lot of content onto the homepage without it feeling hopefully too overwhelming. We've also got this nice slideshow at the top where we can pull out particularly relevant titles, so that's great from a kind of comms communication standpoint. The. Just a note on the design so the pink color is what was like we wanted these kind of really subtle references to the to the hammer and the color was meant to match the curtains of the Billy Wilder theater where these most of these programs take place. And in our search this overlay with the kind of polka dots, actually is echoing a design on our bridge in our courtyard so some cute little some cute little nods to the building. As Andy mentioned the search we want it to be really really powerful, and using the automated transcripts really allowed us to allow for that super deep, searching for the people who have a better sense or they're not just browsing they're looking for something specific. So I'm just gonna show you how it works. So if you're looking for censorship. Actually you know what, let me take out that search result and just show you we've got 1100 videos in here. We're able to filter those through a giant list of topics and tags, through a list of people, which all of these lists can be searched all the lists are gigantic so we had to make them searchable we have playlists about that are like certain themed series, and by date by year. But let's say you're interested in censorship specifically you get 49 results, and as you can see it's not just searching in the title or description, it's searching through the, through the transcription and it's also displaying that in the results which we love it. You see, the italicized text, it means that it's drawing this from the, from insert within the talk, so it gives you a sense immediately of kind of what you're getting into, obviously if it's not the title of the in the title, it might not be the central thing that they're talking about but maybe you're like, oh, Suzanne. Suzanne arlena and Roxane Gay talked about it. Let's see what that's about, you wouldn't maybe suspect that they would be talking about censorship, so this is our video page. We wanted to make it as interlinked as possible so you can find anything that these two have have appeared in by clicking these you can find other videos on the same topics by clicking these. Here's the transcript which works really really beautifully. It will scroll along with the video. But you can also search through the transcript which is really exciting because if you're only interested in censorship and they mentioned at 40 minutes into an hour and 20 video, we don't. We don't expect our viewers to you know spend that amount of time going through so we have this beautiful search even will suggest your search query. And then you can find the section that you want, you can play the video, I'm just gonna play on.
Unknown Speaker 55:15
One of the people she can see we have this beautiful a library now with beautiful
Unknown Speaker 55:20
video player with the pink accents. So once you find this area that you're interested we also wanted to make it usable in a different in certain ways so we wanted to create one of the big challenges are, our content poses is just the length of that, you know, it's like you can't really, you can't really expect people to share a two hour long video, or to be watching video after video so we wanted to create a clipping tool that allows you to take the person clip. Set the start time, you can let it play on for however long that you want until you get to the end of the at the end of time and then you've got a URL that will take you to a clip of that video that you can then use on social media or put in a slideshow if you're teaching or something like that we, we wanted to make it as open as possible without being without having the permissions to download the videos that's not a capacity that we were able to, to provide so we thought this was a nice, nice option for people to take the bits that they want, and then share it how they want. So, let me just I just want to, before we open it up to questions just talk really quickly about sustainability in the future. So from my from my perspective like one of the challenges here is getting from project based thinking into acknowledging that what we have now is a living, breathing website with its own dams, and a standard of content that is pretty high. So, thinking about how we sustain that it's something I think that museums generally are not always have at the front of their mind, especially when we're getting funding. But what's been involved since we launched which I probably should have written down when we launched I completely forgot. Time is melted from my brain. But we've had to integrate hammer channel as a new asset a completely new storage, and publication system into our workflow that we already have that goes from planning, and recording the program to getting it online onto our website and then into the archive. The time and labor involved in that is not nothing, it's quite a bit and we obviously we're not we don't keep this team of 12 people, it's, it's now up to me and one or two other people to keep it moving. So, and we do a lot of programs in usual times we might record like three or four a week. The cost of ongoing fees and services, like I said we have a dams now we have trends that we need to keep using because, like I said we set A, We set a bar for the for the content, the amount that we're doing with it, that we now need to keep up, and then there's like the fact that we just created 1000 new video, webpages, storage for the video and everything like that. The, the ongoing fees when it actually hits the museum's budget is, is always a touchy area, thinking about how we can improve and update on the system is something that we'd like hardly even had time to think about but we need to user testing and analytics we've literally had no time to think about it, you know, those of you who work also work in museums know it can be tough, we're kind of like sharks always moving forward. So, the looking back and iterating. It's just, it's a tough thing to do in the course of our, of our, of our daily lives. And then other kind of challenge opportunities just disseminating the project which is what we're here to do, and we want to continue, we, this is something we would love to see other museums, take on and adapt for themselves. And so, we're here. I'm here. Cogapp is here to help you understand what's involved in that. And with that I'm gonna, I'm going to stop sharing and I'm going to open this up for questions, or for the reading of questions.
Unknown Speaker 59:31
Cool. I can see a few questions, and we started talking, I guess, we've got a terabyte 10 minutes so we'll go quite quickly but does the content taxonomies on AMA channel extend to the collection taxonomies in the museum website.
Unknown Speaker 59:48
No, it doesn't. It's a completely a completely separate system for the video like it's a, it's a different dams that we use, it's a different all the, the Metadata that we're gathering is, we were basically not totally starting from scratch, but we were developing with this project in mind so the, the collection and taxonomies were developed separately and, but the, what what you're suggesting is like, that's what we should be doing right like our hope is always for these things to encourage a more holistic like, you know, moving forward at the institution, and that's something we always struggle with is just like we build these things as projects, and it's not they're not always speaking the same language or sharing the same technologies and things like that, so. Yeah, no, they don't extend.
Unknown Speaker 1:00:48
But I think what Neil was saying about the technical approach to the API's means that none of that stuff is cut off. So, there isn't there there are collection digital archives and things like that on the main website right now, where there would probably be scope to do things like map is the balance of the mission to the hammer is art, and ideas are two sides of a coin, but it doesn't mean that they can kind of link to one another, even if it's not directly,
Unknown Speaker 1:01:21
we could we could blend those together if we, if we wanted to, I suppose, yeah.
Unknown Speaker 1:01:28
And so the next ones around, I think you kind of answered this one a little bit already feel about. Have we find any unexpected audiences we, there is quite extensive analytics on the website. And, but we haven't. I mean we were on a bit of a, we launched.
Unknown Speaker 1:01:53
And we haven't done much, so that's been building up but it would be something interesting to look at for sure. It's probably, I would definitely like to look before the end of the year, and maybe we could get some time will be interesting.
Unknown Speaker 1:02:05
Yeah, it's one of those things where it's like, oh let's just give it a year so that we can really let the analyze all the data build up as opposed to like it's been six months and we have not gotten our, our stuff together to do. But yeah, we don't know what yeah we haven't found much I mean I think there's been a bit more attention on this since we've gone virtual basically of like people realizing oh people will come to our programs from Europe or you know from New York. And I think we would probably find some of the same. Same thing with obviously with with hammer channel but yeah, I'm eager to jump in
Unknown Speaker 1:02:42
on launch day, we call that we kind of sit and look at the analytics, and there were people that were not in LA, and they weren't in Brighton. So there were people that were, that were seeing this content, probably for the first time, and there were people in Asia and people in Australia, places like that. So, even if it's just those people, I'm sure there'll be more, but it was interesting to see them, maybe not the scale but the reach, it was happening.
Unknown Speaker 1:03:11
So the next question from Josh is a one which been on my mind for the entirety of the project that yeah this is definitely video is complex, and big and can get expensive. So it's definitely something that has to be thought about. Now in this particular instance the dance that hammer is using actually serves the video files so that's part of the, so that's proprietary third party provider. So that actually handles it in essence, but I think if we wanted to provide something which had more variable bitrate technologies which was able to offer more of that kind of YouTube type delivery. We would have to look at, as you say something like an external CDN which specializes in video, but then obviously there are implications there about, You know, additional places to send video, working out, has it finished processing the before you can publish it live and that kind of thing. And it's interesting, the point you make about could we just post this on YouTube or Vimeo. And so, yes, I reckon with this platform, one of the ideas is that is not locked into use a particular dance, there would be a little bit of blue code to rewrite. But you could pull information from the YouTube API. Get the playback URLs from YouTube, and then you may have to make a couple of tweaks to the video player. And of course, the one thing with YouTube and, which is one of the reasons why we didn't go with that on this project is the presentation of their video player changes almost daily about whether they're showing things and putting up ads and all sorts of details can't change but for some situations, we definitely recognize that that is perfect. So we actually did a hack day. Experiment with, with hemichannel, with the, with the open source code to sort of see, okay, How quickly can we use a different video source here. So that's something which we should actually be pushing up to the, to the open source repo. Coming soon. Let's see.
Unknown Speaker 1:05:36
Yeah, so the decoupled nature of the idea is that, you know, if you were, if you use the different dams, or if you had a CSV file with some information in it, you know, it could be as simple as that. And the we're trying to get the message out today because, you know, dams providers might want to write that glue code and submit it and then their customers could use this or you know, people with their own tech teams could do it combat could do it is not. It doesn't require, and it's not rocket science. And we're kind of what we want to want to do today is, promote this, so that people ask us those types of questions and just to let people know that we're here, and the other kind of thing we've only got a couple of minutes left but and there we can put the slides up, there's links to the repos the documentation, all that stuff exists, but it kind of, if anyone wants to know how to use this and have a sort of one on one conversation with either of us any of us, then please just get in touch, and then we can set up a call and run you through it and and let you know how you can take advantage of it. And could you stick that slide up there and 30 or so people,
Unknown Speaker 1:06:46
yeah, I'm working out. Somehow I completely skipped over that one sorry
Unknown Speaker 1:06:55
have time to squeeze in one more question I
Unknown Speaker 1:06:58
think it is similar in terms of the Analytics Suite we haven't looked at the analytics too much Daniel,
Unknown Speaker 1:07:04
but But what I can say what I can say Daniel for, sorry to interrupt. I have a minute left. And I'm going to share this is like I should probably clarify, we're not, we're not taking things off of YouTube and Vimeo like we're still using, we know that we're not going to replace YouTube with hammer channel we have realistic ideas of. We know that YouTube is a huge driver of of traffic so it's really more about like, we want to keep as MANY doors open, but also have hammer channel as the place where everything is packaged together in the most, you know, ideal, and, and self contained way where you can really really explore. And in terms of how we, it also helps us I think in terms of promoting the videos because now there's like a shared a shared does that live there's a, there's a location and there's a design to it so they're basically our website now points in MANY different places anywhere that would usually take you to a webpage with a video player will now send you to hammer channel. And our hope is that once people are in there. One aspect I didn't, I didn't show in the demo is that there's like related videos and, you know, we want to move people from one thing to the next and hope that they'll stay and have a nice in depth experience of the of the site so
Unknown Speaker 1:08:32
I'm gonna encourage you to do that and I'm just gonna have to say thanks so much for coming today. Check out these slides well I'll probably try and tweet them in a moment but thanks for coming. I really appreciate you guys coming in, Kevin, thanks so much for your questions that were really
Unknown Speaker 1:08:47
nice to the questions and nice comments thank you very much.
Unknown Speaker 1:08:51
Yeah, thanks everyone have a great conference i Cheers, bye bye.