Week 33: Pad.ma
Scott: Hello there
Hey
Scott: Well, for the most part, the audio is surprisingly not horrible
It looks like a number of people are contacting right now to be added to the chat, but I’ll wait just a second…
Yeah so welcome guys, thanks for joining our chat at this ungodly hour for you, either staying up super late or waking up just for it in order to talk about Pad.ma, we're really excited to talk about it with you, especially in the context of what we're calling plausible art worlds, these examples of creative cultural eco-systems, the kinds of things that support art practice and help art to be understood as such maybe in a different way because it's structured differently, and it seems like Pad.ma is an interesting example of this, so we're really happy to have you.
Steven: Do you want to take it away or
?: I need one minute, sorry
Scott: Yeah, if you guys just wouldn't mind giving us a short intro maybe for the people here that don't know anything about Pad.ma, that would be great.
?: Well, I have to say that we're missing a few people here and the reasons for that is that somebody who works close here in Pad.ma is sick at the moment and is in Bombay, and Shaina who is my partner is 9 months pregnant and she is in bed and I'm unable to wake her up at the moment…
Scott: Congratulations
Ashok: yeah it would be nice to have them but there's also a number of other people involved I think you have the list on your website, and it started off as a meeting of people who in different ways have an interest in [inaudible 0:02:44.9] and I could describe just briefly what that was; about maybe two and half years ago, several of us representing four or five different mini--some small, some large, large barely existing--institutions met and kind of centered around a body of material that was already existing; that was video material that was already existing in Bombay and centered around a project that already in a way predated Pad.ma which is the 0XDB which I'll let Jan and Sebastian speak about, kind of centered around these two existing potentials and things that already were there began to build what is now called Pad.ma and I think that it was really a diverse group, for example from Bangalore, a group of lawyers and legal researchers who had been working around issues of copyright and a lot of the local work of one [inaudible 0:04:00.5] continues to be done by this group called the Alternative Law Firm in Bangalore, there were two large Bombay-based NGOs who had a lot of material from the last ten years around cities, conflict areas mostly in India, documentation of slums in Bombay, stuff like that which was quite valuable because it had been lying around on tapes, as it quite usual in large institutions, and they were trying to get off the ground a kind of archive initiative. There was us, which was a group that was just starting to call itself CAM, but through Shaina's own work as a kind of experimental film-maker and working in video a lot over the last ten years of TV, and the ways it had been around in the Indian documentary film-scene for example, she in particular had an interest in the video archive, and I can describe a little bit more of what that was. Then of course there was Jan and Sebastian who we had met in Berlin and had seen by [inaudible 0:05:22.0] and seen what she was doing in terms of being able to look inside video material rather deeply that [inaudible 0:05:33.1] and so that was the starting point of the conversation. It has since changed shape, this was a kind of early matrix of interests if everybody go that, I'm not sure if everybody is able to hear me I may be…
Scott: Yes, we're able to hear you really well
Ashok: …developments that were in the pot so to speak, around two and a half to three years ago.
Hello Jan, are you here now?
Jan: Yes [I thought it was in two hours? 0:06:29.1]
Steven: I'm not sure what to ask first, maybe you could talk more about each of those components of the matrix, they're all extremely fascinating in their own right, but I don't know if that's the thing to ask you first, maybe the better thing is to talk about what the matrix actually is and what you build together; whichever way you'd like to go.
Ashok: Sure, but maybe, I could do that from the point of view of, well
One of the large components of it, I'll describe in detail one of those kind of situations as I saw it, and of course you have to realize that at this point I'm also speaking for a lot of other people so I could be saying kinds of things that they may not agree with, for example, one of the organizations that was involved with the early inception was [Modulus 0:07:44.2] which is a large NGO in Bombay which has a cultural kind of legal wing, it does very serious legal work around women's rights, all sorts of things around conflicts, class-based, caste-based work and so on, and have been very prominent on the city kind of legal scene, they also have, as part of a two-phase structure in some organizations in India, they have a cultural wing which actually uses funds from a variety sources, this is also typical in the Indian context from typically Modulus' direct would say they use animal husbandry money to do films, you know there was a large amount of cultural work, film production, and so on done with this kind of development money at a certain point from the middle of the late eighties to the late nineties and so on.
Scott: If you don't mind, sorry because we're projecting this and taking a look while we go, we're trying to find a website for Modulus and it seems to be a general term
Ashok: Modulusbombay.org would be the website
So in their case, they had a lot of material, they were trying to get off the ground an archival projects called Godam which, in this particular case had a lot of material around the city of Bombay and a lot of material on Kashmir, and it had been collected in the course of a career as an organization which actually have done rather well. So they had a lot of tapes, they had a lot of material in TV over the past ten odd years and they had done a project on Kashmir they had done several projects on slum and or informal housing situations in the city of Bombay and a lot of this material was then being put together in an archival they were calling Godam which translated into English means warehouse, now they were attempting also to make this material publically available and they were experimenting with various ways to do it, Shaina had been a fellow of Modulus, and was invited to be part of the exercise of imagining what Godam could be and how they could actually make it a more public archive. At that point [inaudible 0:10:48.2] organization which I was involved in, and we were giving birth to this thing which is now called CAM, a kind of organization that is now a non-profit, [inaudible 0:11:04.0]had its own archives which overlapped with Modulus' in terms of the Bombay material and Shaina had been working around [inaudible 0:11:15.2] of distribution of video in her own ways through things like [inaudible 0:11:19.2] television and so on, and so there was a group of people in Bombay who were involved in this discussion around the public archive, and this kind of video memory of the last say ten-odd years since the DV kind of thing had happened in India and had caused an explosion in the amount of production which was not the theme, or was not followed by an equal number of streaming venues or an equal number of even films being shown or seen around screenings and so on, so there was lots of people shooting, lots of material, but not necessarily many venues to screen it and not necessarily a sense that there were platforms that this material could be shared and so on. So this was the context that existed around two and a half to three years ago where Godam was already a existing archive project which needed a kick in its backside in order to--and they admitted it themselves-- which needed a boost to get it out there, all this material sitting in cupboards and actually other people were contributing to and so on was not accessible, was not being productive. Into that conversation also stepped in the software that Jan and Seb had already worked on and maybe it's a good time to ask Jan and Seb to talk a bit about that from their point of view.
Steven: Jan and Sebastian, maybe you can also give a kind of a bit of a history the way that Ashok did of where you guys came from.
Scott: Steven do you mean where geographically speaking? or where politically or
Steven: No, I was thinking of [texts? 0:13:26.3] basically
Sebastian: I mean historically speaking we were coming from a practice of copyright [inaudible 0:13:37.1] mostly and when we met with Shaina and [name 0:13:41.2] when we kind of had our initial discussions and pretty soon gave birth to Pad.ma, the project, Jan and me we had just been running a thing for a couple of years which was called Pirate Cinema. Pirate Cinema was a series of weekly screening we did in Berlin, we started in 2004, and the idea was that we just noticed that we were downloading so many films that we couldn't just watch them ourselves, we needed some help with that, so at the, yeah, it's there, but it's not there there, but Pirate Cinema was in a way they follow a project to [inaudible 0:14:30.9] which was just a huge online [inaudible 0:14:32.8] but Pirate Cinema had more of a physical component in the sense that we had a space for it, and every Sunday we would do a screening of films we had downloaded and you could screen it for free, you would get a copy of the film etc. Pirate Cinema [inaudible 0:14:50.0] has a schedule of what we did, I can paste that too, yes, no that was the problem I mean we noticed that once we'd seriously started to get in the business of downloading either we would become like lonely archivists or we would have to do something that involved more people so one of the things we did to involve more people was of course running this as a cinema, and I mean there were many pirate cinemas everywhere at the time, and all over Europe, in London, in Denmark, and Sweden there were many people we also knew who ran similar operations, but then we also noticed that in order to not become lonely archivists, we had to do something about archiving in a more technical sense which was to do some archiving software. I mean anyone who deals with a large number of files knows that it sometimes is not so much fun organizing stuff that sits on all kinds of different hard drives and so from that Pirate Cinema at the moment, in 2007 we were running this project called [All 21 0:16:05.9] the all of the 21st Century. So this had funding for a one-year long project conference, meetings, workshops etc on intellectual property, and part of the things, I mean when we did the budget we thought we also want to do something practical even though by the beginning of the project we had not really much of a sense of what this would be, we wanted to produce something that could be a blueprint for an archive, and since we were sitting on so many films, we thought ok, let's do something that takes film as a body of… as a medium and let's see that we can do with all the data we have that we normally don't find on the internet. I think 0XDB takes some time to explore, but what we basically thought was A)now that we have so many movies and so many subterms and files for them, actually do one full text search in movies, so for example, something like, I don't know if I have a good link..
and so one thing is, as you can imagine if you have subterms and movie [inaudible 0:17:37.6] or then later with Pad.ma text [inaudible 0:17:39.6] time-based, you can easily reference and retrieve information that is at a particular point of the film and then the other idea was as we had a bit of knowledge about video codecs and some idea about graphical representation of film, what we thought is that film is always so hard to browse on, if you have a huge digital library of films, it's almost impossible to handle it if you don't have the right tools, so one of the tools is maybe something like this
so we try to [extract visual information.. inaudible 0:18:22.8] blueprint-like overview of what's happening. I mean that last link you can just browse over see the subtitles and click on…
So in the end use this data gathered from all kinds of dark networks, make some public website out of it, and try to give a hint at what all the funded archive video projects that were already existing online failing at. Which we thought was mostly about search and video and mostly about making accessible, representing graphically [inaudible 0:19:12.9]
Of course it was also strategic kind of investment of time because we thought that a project like Pirate Cinema would actually win if it had as a side project or as a co-project a resource that had all the looks of being serious about software and serious about technology, so we always thought this was a nice combination. Then from this point on it became relatively clear, relatively soon what we could do with Shaina and [names 0:19:48.5] in the future, because all these archive tools just called for an application with actual material contributed by network [inaudible 0:20:06.0] and also called for actually user-contributed annotation.
Scott: Yes, we're here, Steven got dropped for a second, we're just adding him back.
?: I don't know how much of the [inaudible 0:20:48.3] … on the screen there but we have this timelines and ability to reference video at a specific point on the video so it's not only that you have the entire video, but you can say at this point in the video-- which is then something that Pad.ma plays a larger role with so we don't just have information about videos that are collected from various sources online, but is actually entered by people that mostly have a relation to the video that they work with the video, so that they have [inaudible 0:21:35.7]environment.
People can directly link to a segment in the video.
[pause]
Besides linking to a particular point in time, you can also
Scott: Ok, so we're looking at this link that you just sent, and you were just describing that a little bit? We haven't used the tool here on our end so, the actual workings of it are a little abstract for us.
Jan: [inaudible 0:22:47.2] … a bit like a video editor, so we have these different views of the video on the top on the left, while the most left one is a virtual player and the central one is the [inaudible 0:23:04.8] and the right one is the out one and below it you have this time-line which is a representation of the video we are looking at one pixel is one second, and we have a bit of information about what you can actually see there, so you get an idea of the videos.
Scott: Wow, this is amazing
Jan: If you navigate now to, you can click on anywhere on this timeline and then you see the player view will change to this position and if you for example press "I" and "o" and you can set the in point and press "o" to select the outpoint, so you could mark your own field, and if you would log in you could now add a new description, or keyword, or location, or transcript, you can also press the play button below the player which works with browsers that support html5 video playback [inaudible 0:24:15.1]
Ashok: If you want to explore the site later on for both sites, it's nicer if you get a free account on this site because then you cannot just read, but also write…
Scott: Yeah, absolutely, we definitely will
I was just browsing through using; not to get too technical about this, but this is a technical project on some level, I know it's both theoretical and practical, but just sort of scrolling back with my keys I can scroll through second by second through the entire clip that you posted, and see the description change as the clips change, and I guess presumably we could annotate this?
Ashok: If you press "0" then you jump to the next point where something changes so you don't have to jump through it second by second, if you press "h" which will bring up a small help screen with all the keywords that you can use.
Steven: How much video footage do you actually have on there now and how much [is voluntary? inaudible 0:25:48.7]
?: I think it's something like 600 events now that are different videos, and they have many thousand layers to that I think, for example I have some stats here; 7000 descriptions, so most of the videos have a description and a transcript layer, so there are many more transcripts though, 14000, sometimes they are a bit finer grain, but also they are of larger blocks
Ashok: Since there was already some transcript, many of them are also done specifically for the site.
Scott: Guys, if you don't mind me asking, just in terms of scalability, 8000 movies is not very much compared to other online, free online video repositories, I realize this is a bit different, I guess I was just wondering, when browsing through these, we haven't touched on all of them at all, but we get a sense that a lot of these are not necessarily all coming from the same political vantage point, but they seem to have social element to them. I was just wondering if this was something that was opened up for literally anyone to upload video and it became highly popular, and there were thousands or millions of videos uploading where you had to look for the documentaries amongst beer -fart jokes and you know, sort of frat-humor videos if you know what I mean; I wonder what that would do to the project.
Ashok: Yeah, I mean there's definitely the art of growing slowly and I think while Pad.ma has upload functionality and I know that anyone can just upload their video, there's still currently a moderation process so everything uploaded lands in a queue, and there's also not; I mean I think this question of scalability I would rather address it once is occurs because like so many projects don't make it even to the hundreds of hours because they're already thinking what they're going to do with millions of hours, and I think once Pad.ma has ten times more material than now, we can think about it again and result of this would always be a bit different. I don't see in terms of infrastructure that is technically why I wouldn't scale of course hosting more video material online costs more money, but otherwise I think the system itself is quite sound.
?: If we grow it slow enough then maybe the cost will not really rise, because all this is getting drastically cheaper, as hard drives and bandwidth goes down
Ashok: I also think that with Pad.ma that by establishing a certain level of description and transcript, and work that is putting a video on there, it doesn't fit for everything, it's not a place where you would just put your video if you are not interested in this level of annotation, so by that it also filters out a certain category of videos. Sure it can happen that the most interesting things people want to write about is some funny video.
Steven: But it is possible not only to comment and to comment on the comments of the visuals, but it is also possible to upload new video material if you're a registered user.
?: Yes absolutely it's possible, and one of the things we want to add because so many people want to do it; I mean what you already can do now, is you can download any video on Pad.ma in relatively nice resolution, download
Ashok: Not only can you download the entire video but you can also download a second of the video, if you mark an in and out-point, you can in the actions menu select just to download this clip, which allows you also to extract just a second of the video from the archive.
?: And obviously what people really seem to like to do in addition to that, and what we hope we will implement in the next version is that you can also remix the stuff right on the site, so now that you have this nice timeline, this nice video editor-like thing that allows you to set in point and out points than you hit "d" and add a description for example, people really do want copy and new empty timeline and paste, so you can basically create a huge list of bookmarks of clips and that will be a video in itself, so that's one of the things I think we would love to add in the near future.
Jan: I wanted to add regarding the upload functionality that when you have an account you can upload a video and you can do anything with the video, you can also send around the link to the video, the only thing that is not happening right now because it will not show up in the search unless it went through moderation processes where we decide if we want to have this as a public video, so that is the level on which it is currently open or not.
Sebastian: And then maybe one more work regarding scalability, I think that the thing that's least likely to scale very well is thorough annotation because for many videos, let's say for example we work with filmmakers that will work with people that have larger archives, now many of them are really willing to give them material or provide the DV originals and so on, but what seems to be really hard to do is good annotation because in some cases people have already logged the material, have their sheets so this can be easily imported into Pad.ma, but in many of the cases, lots of people work hard for weeks, if not months to get proper annotation done, which is descriptions, which is transcript, which is mapping out location, referenced in the video etc, so unlike 0XDB with which all we have to do is put movie files and subtitles on a server and be done so that can grow relatively rapidly, with Pad.ma there's another vector of growth if you want, which is not so much volume, but depth of annotation, and that we'd love to see grow just as we'd love to see the site grow in terms of volume, you just don't see it that easily, but if you dive deep into Pad.ma you will notice how deep it is in terms of what people annotate, what people contribute, in which ways people use the annotation function to actually incorporate their own writing on films that my span different items in Pad.ma so there's a lot of depth also.
Ashok: I was going to say that what we do also think about is there is a kind of ecology or a kind of process which relates to the term footage which was very important to the way Pad.ma was imagined and I think the way I think it has grown. This is a crucial term because it is distinct from film and deals with what you might call the remnants of all the differences in the processes of taking video, or shooting video, or being a kind of camera person and making films, there is a vast gulf between those two things, and it's something that we felt and physically felt in our context of piles of thousands of tapes gathering around up, but very few films to show for it exactly, but in general the idea that filmmaking economy, especially with things like documentary or art projects, creates its own [brutal? 0:35:32.0] selection process and while we can say that there is merit in that selection process, there are also things that get left behind, and there are things that potentially other people could use, and there is productive remainder which was one of the first things that Pad.ma was address, that there is the filmmaking process but then there is a thing called footage. Whether it's found footage, or footage that was shot for a project but ended up being used somewhere else, or not used ever, or has been lying around waiting for its film, waiting for the film; things like that which are quite common in I'm sure in all of our experiences. But what footage does is it gives you a long view, or a very different kind of landscape and now when Shaina and me for example watch films, we're watching footage, we're always that the edit would have been a bit different, or if we couldn't see a more of that; so there is a way in which the filmmaker is trapped into this ideal making of film, which is not necessarity the only result of that practice.
So to give an example maybe, I don't know if I have… ok so Pad.ma has for example lists, this is a project that refers to the previous link I sent out… so that's a list of events; events are individual pieces of video on Pad.ma; a list is a collection of such individual pieces, and is in this case a single project done by Shaina and myself in which we…so it's in the context of an art project we were invited to work with a group and a space in city of Manchester, and it was called CCTV Social. We had access to large CCTV rooms in the city of Manchester, and we invited people to come into them and to these kind of one-hour clinic sessions, or sessions in this CCTV room with the police, with other kinds of provocateurs and so on, and these are documented in their entire length in this list, and that has a very different kind of sensibility to perhaps what was presented within the context of the art work, or the exhibition that was created out of this process. And already there have been many people who have referenced this material in other ways; some people have used parts of it [inaudible 0:38:53.1] of Manchester for example, and it enters potentially a wider economy.
The idea was the Pad.ma give it as much context as possible to really think archivally in the sense that you're not only putting this stuff for viewing pleasure, or it's not a kind of YouTube situation, but it's really about a context for usability, a context for a kind of productive context that could be generated out of this material that mean the ability to download is very explicit, there is a lot of textual material, there is many many ways to search it, it's very deeply annotated, and there is a kind of generosity in the idea that we like which is that footage; not necessarily everything that you shot, not the stuff where you were testing the camera and so on, but also just not what you produce in a certain context, so there's a space between those two things.
And I think that extends then to this idea of the footage [inaudible 0:40:13.9] like film inherently extends perhaps to things like actually writing about films which is… something like that. So if you click on that that's an example of Lawrence, who is part of the collective as well who is writing about the viewing citizen as constructed through in the cinema; writing about a particular clip in the film, which shows an audience, a so-called [inaudible 0:40:52.2] and the ability to see the material, so if you imagine cinema scholarship or writing about video material which lets you actually see it, which is actually a primary condition of talking about it, it's something that Pad.ma lets you do, and in a way there's been already a move locally within our context to try and use Pad.ma for this kind of, not only for footage that is left over, but then to turn film themselves into a kind of footage-nest, an idea that's a bit looser than the in and out points that were determined by the length of the film at that point or by the BBC or whatever, was a constraint, but to be able to see it at other levels, and to then of course, as Sebastian said, you would be able to do in the next build and so on, and to be able to use this material in a way to see it, which we haven't really seen, I mean there's all this scholarship around things and the writing around it for example and viewing of films have been in to a degree a separate experiences. At one level what very simply happens is that you can see them both together and they don't necessarily need to be indexical related to each other, I mean they can be neighbors, they can be sitting next to each other in a space, and that's I think is a pretty powerful thing that happens in Pad.ma already.
Scott: Ashok I just wanted to ask you a quick question, about the CCTV Social project, I understand now what the features lists are about I think; those are project-based lists it sounds like, and I guess I just wanted to point out that in a sense it seems that this tool can be used in many ways, both to build meaning visually and literally too to edit, I mean for instance, the CCTV project it looks almost exactly like one of those monitor rooms, just that page link that you sent earlier, whereas other pieces of the site have a different kind of feeling; the editor for example and other sections, but yeah I guess… you were bringing back the disambiguating the efforts of Pad.ma from other pseudo-archives like YouTube or something like that, because it is moderated, and I guess I was curious because we are talking about this in the context of supportive systems for creative cultural practice, or critical cultural practice, Pad.ma is not only a tool, but is also a group of people; for instance it's moderated, so who moderates it? It sounds like you guys, or a few of you, or something like that?
Ashok: The thing about moderation is that Pad.ma has had a slow growth as Seb mentioned before. so what we started off doing was we started out with actually a group of people who had [inaudible 0:45:36.1] putting in a lot of effort, and I mean this quite seriously, there was a lot of effort that does go into the writing process, the annotation process, so a group of people including us, including the whole initiators of the project, and a slowly expanding group that was provisionally invited to try and create a kind of starting mass for the project. Because of the sheet amount of labor that it takes compared to an economy like YouTube inters of the kind of context that all the videos provide, it's not [inaudible 0:46:15.7] for random people to just come in and be part of that ecosystem and I think we've just got to a point where we kind of publically opened it up to people upload on fattier own, and then this has been happening but the volume of it is not huge amounts so the moderation process is still quite casual, it is done by a group of people who both moderate and invite chunks of material, I mean it's much more that we're interested in these mini archives that people have around a certain sense of idea, sense of projects or specific geographies like Kashmir, or specific city contexts or specific times like [inaudible 0:47:10.8] or 1992, or [inaudible 0:47:14.0] and so on, these things are [inaudible 0:47:18.3] and then people who look at these things on Pad.ma then write to us and say "Hey I have material around this as well…" so it's still at a state where it's a humanly possible set of interactions. This could change but I think also we're at the moment, still very much in building the archive mode where it is not necessarily advertised as this "come and upload" your own video site" it's advertised as a rather serious archival slightly nerdy kind of site where people are spending a lot of time over little bits of video and only a limited number of people are interested in that, so at the moment, and also I think the time that situation may change with the new software situation, so I a lot of this is very much the beginning of what we help to build densely, deeper rather than in purely in terms of participation of contributors of video [inaudible 0:48:37.9] to encourage to keep a relationship with it. So at the moment it's a combination of inviting people and moderating people who might be interested, and that's worked reasonably well, we demand a lot ..
Steven: [inaudible 0:49:01.5] I noticed in Beirut, in the context of Lebanon in the Middle East, when you did a workshop there, one of the sounds that I think that you encountered was that people in that area where very much involved in their own conflicted situation and lacked a kind of fluency in connecting with others; I think that is kind of [inaudible 0:49:44.0] to what Matthew wrote just a moment ago, is that it's very difficult to make all the links and then work that scale of [inaudible 0:49:54.9] between two local conflicting situations and a global framework which would allow them to be connected.
?: I mean because one of the challenges that you face in, not just in Lebanon, but in Lebanon it's very obviously, is that a lot of the people not being able to network stems from the fact that they have virtually no internet, so while you always find a specific archival situation if you want; both in terms of archival of politics and in terms of just technology and what you have in terms of infrastructure and for me it has been very interesting to do a lot of work on Pad.ma and work with a lot of communities around Pad.ma in conditions where both the politics of archiving and the technologies of archiving are not as clear cut as you would think maybe in the US or in Germany where you have this kind of idea that things are just as they should be because how could it be otherwise--no, I think in terms of technology we've been pretty, since we're used to it, we can run Pad.ma locally as a server, we could put it up in a room, it takes s few hours, it's not such a big deal, so you can play with Pad.ma locally if your internet doesn't allow for other things, and we've had the same thing with many contributors and not just in India, they said "ok, I put my video online, but just annotating online, I don't have the bandwidth, I can't do it" so you also have to account for that. Then, there's also of course the archival politics which means by now many people have noticed that archives--and this can be archives of art, this can be archives of [inaudible 0:51:51.4] sensitive material that a resource than can be monetized, that can be sensitivity can be exploited for all kinds of means, you can have the author who said that ok think is my archive, but if I share it online everyone will remix it, you can have the museum director who said "here is all my stuff" but there's some politically sensitive bits, or you can have like the huge companies who go after the many individual archive owners not just in western Asia, who actually reinforce that notion that there would be a monitory value in this and you can't just put it up on the internet and open source is not going to help etc. and given that situation both technologically and politically, I think you have to work with what you can work with; which in the case of a Beirut workshop where a couple of people we knew and many more people we maybe didn't know before, and tried to… I mean part of my [inaudible 0:52:53.3] is a tool for research. First of all, it can be the part of research that people could do when they look at their own things' you look at your own footage, you edit and you annotate your own footage, and then by opening it up, you allow other people who may or may not be familiar with the topics raised to annotate more and to continue that discussion, for us it was a logical starting point to involve people, to work with people whose research we could actually follow and had an interest in.
Steven: I was just wondering if you had encountered [inaudible 0:53:48.3] from people in getting them to add their footage for either copy right or intellectual property right for whatever fear that they have.
Ashok: It always happens, it always happens and to me dealing with copyright online for a long time, I think I can recognize in this.. we've dealt with filmmakers who have said we have this anxiety once their footage is out there, it can be misused. Whatever that means, I mean hardly anyone will walk up and say I'm [inaudible 0:54:37.2] copyright so I'm not going to put it on a platform where you have a creative commons [inaudible 0:54:43.2] and then everyone can download. Everyone individually is a big fan of liberal copyright and open source software etc, but when it becomes personal and concrete he will often be like "no this material is politically sensitive, someone can misused it" someone will say "ok, other people can annotate, but what about other people adding false information" so the figure of this author, they're very very strong, this idea that the internet is out there to screw you over and is common among many producers and today archivists are authors too so you feel that you're the author of the archive so the problem perpetuates. But the reasons people come up with I think are mostly, I mean I read a funny story today in Germany there's a huge discussion these days about Google Street View, it's going to be launched by the end of the year and now the government is critical about it and people are invited to privately, to personally have their own homes taken off the Street View for privacy reasons, but this is a --and today in one online magazine I saw people were posing, there was an article about these people taking off their houses off Google Street View and there was actually a photo of these people standing in front of their own house with a banner, and these people are going to have their house removed from Google Street View, but these people are actually in a publication online standing in front of the house they want to remove for privacy reasons. So I think there's always something else that motivates people to have what they think is their home taken off the internet, and in many cases it's so strange, I mean we can talk about this for longer if you life, we've-- one of the by-products of Pad.ma is deep research into author's anxieties to publish. Even though we would always way you don't have to put your material on Pad.ma, if it's juts about making content available, put it on YouTube, that is fine, there are many many options, but if you want it to have context, and I think Pad.ma provides context, even more than it provides content, if you want it to be on a specific context and be part of a network and be in a region of the internet that has a certain density, then I think some people then realize that this is actually a good argument, that it may be of value to contribute.
?: Yeah, and I think it's been important to make the suggestion--I mean to say on the one hand yes it is the decision including of the ethical decision of the filmmakers contributing their material or the authors otherwise contributing their material to Pad.ma it remains their decision to contribute whatever they wish to, but to push that always to the questions quite fundamentally, in order to debate what it is we're afraid of we have to see it and if we don't let us see it then the debate around what is the ethical problem can't really be had, so in a way you can black out entire sections of video and so on, but in terms of the discussion and in terms of being able to enter the discussion, it's impossible to do that without having anything to talk about, and part of our approach to filmmakers has been that yes, include the problem the archive is about talking through your immediate fear or its misuse and so on may tell us something about the nature of, not only of fear, but also if something is misused in a certain context then that may also tell you something about the footage itself that maybe more valuable than your fear at the moment. And I think Sebastian said this once that if they rightwing uses your material then maybe that said something about the footage itself- -is there such a thing as right-wing footage that the rightwing can use? and in which ways?
So in general, the provocation and idea has always been that the archive is a place for these conflicts, the archive must tell us more than just what we know and what is available anyway in other formats, or there is a space for this discussion here, and I think there's an ongoing kind of attempt addressing your previous point Steven, there is an ongoing attempt to now make this more of a reasonable question than rather just question stemming from the original groups kind of material. and it's just a point we haven't reached yet in Beirut, as you know, the internet is functionally quite bad, and nevertheless there are conversations that are going to continues; we're doing a workshop next month in Cairo, and so around this area, and the Middle East that we as [inaudible 1:00:33.7] work in as well, there is a bit of discussion now around what material can be in common, what ideas can be exchanged, there's a bit of a shared history around the documentary film tradition and with Egypt with the film tradition as well, so there are several possible points of [inaudible 1:00:56.2] and I don't think we're at that point yet.
Steven: I have a question to do with the type of search through the archives, you can do I know searches through the annotated texts, but is it possible, or is it conceivable to do searches that are based on for example visual criteria, or sound criteria, or criteria other than annotations based on the keywords of the annotation?
Sebastian: We'd love to do a bit more of this in the future, but I think if you wanted to search for the guy with the glass; technologically most of the interest in this field, I mean there's face recognition and there's also all these security related challenges that, of which a lot of effort has been put into at least getting face recognition right, but we're talking here not about a photo album, or limited like always fixed angles, CCTV situation but to be able to search for visual objects in films, is still a bit ahead in the future, nevertheless, what we can do now, what we can do today, and what we will do in the next version of Pad.ma is to retrieve a bit more visual information from the clips for example you can easily detect cuts in the material, you can easily, for example with something like 0XDB, because it's a different body of material, It's not footage, it's cinema, it's almost 10000 films of cinema, it would be really nice if you can at least sort your results by color, or by brightness, or by saturation, or by cut frequency, or play a bit more with these visual materials, but if you really want to find visual objects in moving material, I think you still have to wait, for a bit.
Scott: What about other ways of looking or working with the material such as audio? I was curious just for practical reasons, but also I'm curious, I know this is meant to be a visual index, but I also understand that this is video too, not just silent video, I was curious about the audio component, if that's something that's come up for you guys either theoretically or just technically.
Jan?: I mean the things that Sebastian just mentioned for more visual indicators there are some things that we can do for sound… most of these are right now more of abilities to sort through the ideas, not so much limit it. I mean you can sort it by the average volume, or if it's a loud film, if there are a huge variety of sounds, but in order to search you have to use some form of [inaudible 1:04:48.4] to formulate something, and then usually you do it with text, and so to transform that again into something that is not text… like you sing something and then it finds it, or so that you put up a picture and you draw something, It requires completely new ways of inputting a search, and then you could find it. But that is still quite far from anything that we've actually done.
There are some projects that try to do where you draw something and then it finds pictures there are similar, if anyone could do something like that as part of video…
?: and you could say that..
Jan?yeah I mean by systems where you sing a song and then it tells you which song you were signing, mostly used to detects sounds that are played in clubs, you phone out and then it records a bit of the sounds and then it tells you which song you were listening to.
?No I mean there is speech detection obviously but that works--I mean we've tried this clean room situation, English text, and it works relatively miserably, now a lot of the material in Pad.ma is not really clean-room, the audio doesn't have to be very good, and I don't know, we don't have a display that says how many languages are on Pad.ma, but in many cases, it's about finding someone who can at least transcribe this to English, it's not so straight forward and it's not so easy.
I mean one of the things, I think so far and I think we'll stick to that for a while that is to build on large amounts of good annotation, so if you wanted to have sunsets in Philadelphia, we cannot help you with the sunsets someone will have to mark sunsets in the films, but for example [inaudible 1:07:07.9] with the subtitles what we want to do is extract automatically location information which is place names, or calendar information in the broader sense which is the mention of Events or anything that occurs in time, and map them out automatically on a calendar or on the map so that you get that part for free and with a huge body of annotation which is like 10,000 films per subtitles on 0XDB or which is really deep annotation on Pad.ma on which you could also extract all kinds of things semantically, I think there are a lot of ways to get what you want.
Scott: Not to focus too much on the technical because I think we are also covering a lot of other ground, but have you guys considered audio to text methods at all? Do you see any practical advantage?
Ashok: …I've tried it, but it doesn't work, I was trying it with several films and it was really interesting because there was a scene where a woman walks down the street with high heels, and the text that was recognized was "Iraq, Iraq, Iraq" and I think that was more an indication as to who pays for this technology that is developed that has a military interest. But it is not really helpful for films, one problem is that a lot of the films are cut, and sound is edited with transition so that you don't have clear separation between two people speaking, and that way it was also really hard to recognize when they change, so when a new person starts to speak, and generally, at least with the software that I tried, it was not acceptable, and it is mostly limited to American English, so whenever you have anything that has many languages, or local dialects it becomes [inaudible 1:09:14.4].
Scott: I can definitely see that, I know with some of the more expensive audio-text software solutions you really have to train them, and training them for different dialects and different languages that sounds like a nightmare, but I was curious because it seems like you're interesting in what you called "misuses" and I was thinking perhaps you might be equally interested in the sort of miss-transcription or whatever they are, they would be called, interspersed with then they get it right so to speak.
Sebastian?: Yeah, if they were really original all of the times then yes.
Jan: Beside "Iraq" more of them were not even words, they were just made up things, so I was quite disappointed with what I …and if you train certain things it becomes better. One areas which we might explore at some point is to take a technology like that and only extract certain things again from these texts, so you don't take the entire transcript that is automatically generated, but within that you can, with a certain certainly decide on that certain terms are usually recognized right… maybe not "Iraq" …. it was recognized too often.
Steven: You've collectively authored and [inaudible 1:11:05.0] I've just posted a link to that on the Pad.ma site, maybe [inaudible 1:11:14.8] critical culture, archiving, political cultures [inaudible 1:11:21.7] in these series of discussion s of plausible art worlds, and I think [inaudible 1:11:29.3]
Ashok: I think one of the things about that text is that it was presented as a, in the context of a presentation, so it's not really meant as a text form, it's meant as a series of pretty much polemical short statements, and in that sense you know, and should be read as a kind of spoken text if you like. The "Don't Wait" thing I think was actually the title of the workshop of Beirut and I think the kinds of ways that the waiting around or for the archive for some kind of higher force of the archive or for some higher force of the archive to come and rescue the situation is the common, is at think that we experience a lot in our context, and I think in Beirut it's combined with the kind of waiting that's created by a very commercially [inaudible 1:12:42.6] specially among the media now increasingly about you know the cultural issues with…
So to give an example, there's a large archiving of [inaudible 1:12:59.7] which for ten years has been promising us the publication of its archive, or the access to its archive in some form, but we are supposed to wait until the archive launches, until a certain text has been written about it, till the institution gets its act together, which is never does because actually it's selling off parts of the archive using that money to trade on the stock market so that it can raise funds to make more archive which is still also not going to be visible to anyone, you know so this has been happening for about ten years, and it's really a situation where the archive becomes this promised land that [inaudible 1:13:46.0] the idea of something finished, but also something that is promised in various ways but never quite arrives. And especially in our context at least if you're going to wait for the state or you know, something like that, or a larger kind of commercial power to come and do it for you then obviously you would be living in that kind of regime. So the "Don't Wait" was a kind of practical call to think about these conditions in a way that was using the word archive in a sense politically was saying that it's not only about elections, but yes let's talk about the archive, let's talk about something that used to be the prerogative of large institutions, the state, and so on, and let's talk about that as a collective proposition.
Scott: I like that you have different ways of talking about it besides just talking about it as well; it seems like the Pad.ma project not only some of the content that foreground that I've noticed deals with copyright issues, but also the entire way that it's structured that you guys are actually doing this stuff and then inviting other people to do it as well is another way of talking about it.
I only mention that because well oftentimes there are criticism leveraged at people involved in these discussions where they kind of end on the discussion level and a project like this definitely does not.
Ashok: I think it works that there are many people involved, various people who have different things to take from it, and [inaudible 1:15:57.6] they're written by three or four different people so you know it's not like Pad.ma itself, it's not a canonical kind of statement, it's more like, it's not even going to words a set of canonical truths, but stacking up, layering up, adding to ideas that are already existing in some way or that really different people have different things to take and give to Pad.ma even at the moment, and this has been true I think for its short history so far.
[Alarm clock] It's five hours
Ashok: It's five am guys!
Steven: One more questions, I guess it feel like in the comments that have come up [inaudible 1:16:54.4] a few times already it's been mentioned [inaudible 1:16:58.9]
…a kind of an archiving as you see is like some sort of radical pedagogy and obviously [inaudible 1:17:08.6] teachers or people involved in alternative information production to want to use Pad.ma as a tool. What do you think about that, I mean that's, what if I said I want to do seminar with my students and I want to get them involved in this, I want them to be [inaudible 1:17:37.9] and annotating stuff, [inaudible 1:17:44.5] or what?
Ashok: Steven, I just have to go and get my power cord, but I think it's a great idea, and there's a couple of things that are happening, Pad.ma gives out these small fellowships at the moment and one of them is creating a pedagogic unit around the video archive of this organization that works which basically slum rehabilitation and kind of politics around that, around the situation of housing in the city, their archive is being turned into a kind of class room unit, which can then be used by different--so it's an experimented creating a kind of pedagogy unit, so I have to run off and I'll be back in a moment.
Steven: A question for maybe Sebastian and Jan, what [inaudible 1:18:43.4] language other than English, because if English is in India is the language which belongs to nobody so it's everybody's, that's what [inaudible 1:18:54.4] but what about people who wouldn't feel comfortable expressing themselves in written English, are there other languages on there? Is it possible? Is it conceivable? Is it desirable?
Jan/Sebastian: I mean the interface itself is English only right now, as for writing annotations, you can just write any language you want, and you can also use-- there were some issues with right-to-left languages, but in Beirut people also use Arabic and there [inaudible 1:19:33.4] so other scripts are used on Pad.ma ; the problem that arises with it is that if people search for something they will not find it because they might not be able to search with tat character every easily, or they don't know the language so it creates a sub community in that language
Scott: Would you guys be interested in creating a translation interface, or something like that?
Jan/Sebastian: I mean right now you could just add another annotation with the same in and out point and use a different language and they would show up, and once there are [inaudible 1:20:10.4]that doesn't really work anymore and becomes messy [inaudible 1:20:15.1]
So far I think it would be just that you take a text that exists, an existing layer, and add a new one within different layers and it's always possible if you are on the side to switch off one user, like different users, so you can only see the annotation of one person, or not of one person in the contributor' [01:20:46.2 from the top. So if there is someone editing annotations or some way of getting a translation effort then use you always switch that off.
So that is what you could do right now, and I think at two levels too different multilingual things, one is translation and the other is the people are working indifferent languages, and then I think it makes sense to do that and if it is a translation then this is a another layer of work that has to be put into the side and as people already have problems [inaudible 1:21:38.6] … is a lot of work, I don't know what they would think if [inaudible 1:21:46.2] has to be translated into.
Ashok: For example currently there is a --we're doing a film by a guy called [name 1:22:15.2] a kind of interesting figure in the Bombay documentary scene, and there is a Hindi and an English transcript of it happening at the same time, because it's a very specific king which I think merited that kind of attention and in Beirut for example, there was a few things that I think were translated, and there's already a bunch of Arabic stuff in there which ; this is a random selection but with something like this you could see…
Kind of descriptions in Arabic which I think are going to be followed up by transcripts in English, which are translations of what's going on, so there's different levels of [inaudible 1:23:23.9]
[pause]
Scott: Definitely
Jan/Sebastian: I think it's from a local TV station that [inaudible 1:23:56.8] type of footage
Ashok: Yeah so I mean there's a bunch of foreign material in different mixed languages which I think have been treated, this one has Kashmiri, Urdu, English as well and so on, so yep I mean there's an interesting set of questions around translation I think.
There is this nice feature on the info page that you can, it's like a little flip book of the frames, and the I don't know if you guys can see this, but on the last link I sent out, you can flip through the entire video via rolling over, by scrolling over basically the image you see so and when you click on a certain point, you're taken to that point on the timeline.
Scott: Did, was this covered earlier? I know we're talking with Jan, Ashok and Sebastian, Ashok I know that we had talked very briefly as part of this panel with Temple University here in Philadelphia, maybe a year and a half ago or something like that, or more; I was curious if I guess someone here just asked about the technical side of this, who was responsible for the different components of that night, my understanding was that all of you are, but I'm not sure if that's true, and are there many more people? Even though it's a freely accessible API it's not exactly an open source community right? It's a pretty close-knit group of people even if you are geographically dispersed?
?The software of Pad.ma is open source so it's available and can be used by other people, the development so far is what the people involved with Pad.ma do on it, there hasn't been anyone really from outside that's making for the development. We are right now in this next phase of work which is also open source and can be followed on these links. Both systems also have a public API that can be used to interface with the instance that is Pad.ma, and this will also be extended in the next version that there are more possibilities from embedding segments of the video to having ways of displaying annotations in different ways and using it on your side as you want to do.
?: I think what we want allow for in the future is like, many people like Pad.ma [inaudible 1:28:02.5] but then maybe they sit on larger archives that if they would just contribute the whole thing into Pad.ma would shift the focus of Pad.ma to much of people who want their independence, so people want to be able to run [inaudible 1:28:16.1] and while a couple of people have actually managed to do so with the existing software for Pad.ma, we want to make that much easier in the future by actually developing a video archive framework which then allows us to then 0XTB and Pad.ma on it, but which is then [inaudible 1:28:35.4] but which would allow anyone to run 0XTB or Pad.ma's style online archive with their own content, because in some cases it's really interesting if people contribute to Pad.ma, for some they have existing resources that are more a thing of their own. There would be many things you can do in the future for example, for now I find the uploading process is, just the concept of uploading a video to a site, it's nice if you upload every now and then, but if you have huge amounts of material, wouldn't it be nicer if Pad.ma in the future of this archiving system behaved a bit more like I-tunes maybe on your computer like here on my files, you are on a website just use a [inaudible 1:29:19.9] extension have this read in your own [inaudible 1:29:23.2] .
So what we want to make easier in the future is these workflows for users that are not so much individual who maybe add a bit of annotation here and there, or upload a video every now and then, but who actually want to manage within Pad.ma, or beyond it, larger collections of [inaudible 1:29:48.5]
Steven: I think that the, it's just that it's pretty overwhelming what Pad.ma is in a certain sense, I'm tempted to think, not so much to say about it, but you really have to have a hands on relation to it, I think [inaudible 1:30:40.5] because otherwise it's just an overwhelming amount of information potentially [inaudible 1:30:48.3]
Scott: Steven, could you just repeat that last part? For us here, it got garbled.
Steven: I'm feeling a little overwhelmed by the wonderful potential but that's not necessarily a feeling which I… I'm impressed but I don't want to be impressed, I would rather be a user, and I think it's in a sense what Pad.ma does best as well, is beckons us into a usership relationship rather than being spectators of this incredible device with which we can… we don't even know what questions to ask about it.
Ashok?: Yeah, but you can get that feeling in front of final cut pro as well, you know, it's a question of, yeah this is new stuff, and in a way, it's a bit of a tricky tool, always trying to do things which have not necessarily been done before, and there's probably I could tell you a lot more about the politics for example, the video call that are kind of embedded in the process and so on. Yes, of course, there is a way in which it is a bit of a new things and for the things that it's trying to do I think we're trying to a sense of what that really means beyond the first impression of "ohh this is a lot", but if you search for a specific term then it suddenly becomes a bit less than a lot, and then once you open that up and you're able to look at a piece of video and kind of look at it in a way from its inside if you life. Then it can start to make sense, and of course there are ways in which it could change and of course, Yan and Sebastian will agree that this current interface is only one possible way in which to instantiate the ideas that Pad/,ma represents. It's not the end of it suddenly, but the point being that yes, there are a few functions that it's trying to do and I think it will be good to go a bit further and say why, or what kind of, what's the word you used? what kind of impression it is and what is, what is the negative word that you used? - what's the fear and …
Steven: I think overwhelmed was the word I was using but [inaudible 1:34:21.0]
but on the other hand it doesn't mystify so much as [inaudible 1:34:34.2] really inviting us to grapple with it as users.
Ashok: Yes, and I think there will be different kinds of usership that will slowly evolve. I mean there are people obviously who want to view this YouTube style view full-screen, sit back and view a bunch of material, there are others who would be very specifically looking for things, so there's a whole range of things, there are those who might be using is literally now to put together a lecture for tomorrow morning, it could be people cutting up bits of footage and making new timelines and there is a range of activity there and we slowly find out which are the ones that are dominating depending on the context of the material, the material in it will change a lot of the views around it but yeah, so in that sense if we are over whelmed, is it a question of visual thing, or is it, because obviously you could be overwhelmed in generally quite easily on the internet, I am often, and so do we need a clean front-end like Google that tries to in a way offer you a clean way in, so these are questions yet, they are valid
Steven: That's a very interesting point: the clean way in, I don't know-- you're right that that does create the illusion of not being overwhelmed, it creates the illusion of control and there's and illusions of course, it's… I'm also kind of overwhelmed by the idea of Wikipedia for example, of all wikis, of the kind of modern contemporary miracle whereby so much brain power can be linked together, so much [inaudible 1:36:44.0]collaboration to produce something that's so enormous. It's like the number one billion it's very difficult to get a representation of it, but the only way to engage it is to just engage with it.
?: Definitely, I think that how people use Pad.ma so far I think is only a small part of what [inaudible 1:37:18.7] so it's also that they are [inaudible 1:37:22.4] of interpreting more, from new ways of linking the material to the [inaudible 1:37:33.5] Pad.ma is also really at an early stage still, there's still a lot to explore, and if you are overwhelmed maybe the best is to create an account and start maybe playing with it and seeing what comes up.
Steven: I have an account actually, I have an account but I haven't used it yet, I'm not really sure, can I just sort of wade right in now and start uploading and start annotating?
Ashok: Yes, you new videos will not show up for other people initially, but your annotations are not moderated or anything, so if you annotated an existing video that annotation will not be moderated, so it's like yep, you can go ahead and do that.
I think one of the ways I've used it quite often is to download pieces of video and also download annotations, for example recently we were in Gujarat, we were doing a project related to ships going to Somalia and so on, so I downloaded a whole bunch of annotations from Pad.ma which were around that project, which were around previous visits to that area, so it's really being used as a research tool, and to give you an example of what that was, …
so yeah, in a sense, apart from the annotation work, apart from adding to the archive, the ways in which its really convenient and fast and good to explore is the partial downloads feature that is even in the current build you can just download a bunch of clips, like pieces of video that you can then line up in something like VLC and use it for presentations for example, you can already do this and it's quite straight forward to do and the other thing that one can do is for example I was in this town and I had just downloaded all the annotations what were applicable to these videos which had been done about a year ago and I had them with me during my research trip and now the new material that has been collected is going to be added to this collection, so stuff like that, it's a kind of practical way really for us, it has been to link video material which is often hard to reference, or hard to find, or hard to look and bee inside of, to find a way to just reference that so video material that was shot last year in a certain situation has been annotated according to time code, and so we can find things within it, which is quite useful in a really straight forward way.
One of the fastest ways to zip through the annotations is like Jan said before to use "0" and "9" to go to next annotation.
Steven: Ok, so "000" allows you to skip from one sequence to the next, what about "999" what does that do?
Ashok: The same thing backwards, or from one annotation to the next one, so you can read rapidly all the annotations on that video clip by just going "00000" as long as you're in the edit page, that's editor view
And on actions, there is a thing on the drop-down menu, there is a thing called download annotations, and there are different ways in which you can download these annotations, and you can download all the different types of annotations as different SRT files that's kind of subtitles fields standard, and yeah, you can download as we said before parts of [inaudible 1:43:53.1], you can link to selections which is also quite useful, what we have been doing all evening, all morning. And we have I think a newsletter that I believe the next one is in a couple of weeks, we have a newsletter that I think comes out once in a couple of months.
And the sun is rising
Steven: [inaudible 1:44:43.8] I just want to ask, what is your usership now? How many people are registered users, and how many people are actually involved in a day to day or a week-to-week basis?
Ashok: I don’t know exactly, but I would say a few hundred, it's not a very big number, and also that will change… we've been a bit [inaudible 1:45:17.7] we're moving onto a new set of structures for bringing in contents so we've been on old [inaudible 1:45:33.9] I think that's going to change fairly soon. But yeah, I would say it's a couple of hundred people maybe....[inaudible 1:45:57.0] which are different from people who swing by and have a look who don't actually contribute the [inaudible 1:46:04.9]
Steven: Yeah, that's hard work to define with any precision at this point, but yeah I was speaking of people who are not merely spectators who actually [inaudible 01:46:17.0]
It's been a really fascinating presentation, I don't want it to end, but the sun is coming up [inaudible 1:46:28.4]
Scott: Yeah guys it's really been great having you, thanks for withstanding all of our prying questions
Ashok: So what's the Easter egg idea?
Scott: I'm actually not sure about that, I was going to ask, yeah I think Greg said something about an Easter egg idea earlier I'm not sure what that was.
However, I don't know if we have time to explain it because the clock strikes eight and we always try to end on time as much as possible for everyone who's usually ahead of us in terms of time-zone, especially you guys who … it's nearing six in the morning, so thanks again for bending over in a way to our kind of folding time in space to be with us tonight. We'd love to have you for future chats as well if you'd like to stay up drinking and taking stimulants and join us in the middle of the night, or early morning, you're always welcome.
Ashok: Yes, I think we're going to have a baby soon so that's not likely to happen, but we'll be up at 7am.
But thank you all and thanks Steven, we'll be back, we'll see you online and be around and keep checking in on how Pad.ma's doing.
Scott: Awesome guys, good night everyone, or good morning