Assessing a Collection for Preservation

Presentor: Sarah Stauderman (Smithsonian Institution)


Sarah Stauderman

SARAH STAUDERMAN: I'm Sarah Stauderman, and I work at an archive at the Smithsonian Institution. It has a lot of paper and old photographs and glass-based materials, but it also has a lot of audio materials, video materials, and film. I'm going to introduce something I've developed over the past several years--working off of what the Society of American Archivists had done to prioritize their collections within archives. I recognize there are many individual owners and institutions here that do not strictly fit into an archives type of collection. Please keep your mind open and see how this might relate to your collections, or to your interaction with collections of videotape. Help me figure out whether or not this is going to be a useful document for people who are trying to prioritize their collections within an archive, or for that matter, when trying to advise other people about how to approach their collections.

The Society of American Archivists has something that they call a matrix. Basically, they're trying to combine issues of value, issues of the use, and issues of the risk of any given material. I adapted this for videotape.

Obviously, the first thing you are going to do when you walk through a collection is you are going to identify it. It is sometimes just useful to know simply what the age of it is. The first thing we designate is the series or collection name, the date range, where its location is, and how big it is. Then you come to a box that says “survey results from part four.” I put this on the front of the document because I wanted people to be able, once they had gone through the entire collection, to put the results of their search at the very front. This could pose a major formatting problem, and I'm willing to put it at the end. There is a place for your name, a description of the person who is doing the survey, and the survey date.

The first question asks: are the majority of items identified with labels or descriptive materials, yes or no? Boy, do I see lots of collections that don't have any descriptive materials or labels! This is our first big impediment when we go into a collection and try to prioritize it; if there are no labels or descriptive materials, it is going to be really hard to justify to my managers that I want to reformat this material. If [there are descriptions] we can proceed to part two and collect one hundred dollars; if not, we continue with this section, which simply asks one more question: can the majority of items be played back? Are there playback machines available? Yes, no, do not know. Now, for a lot of people filling out this form, they won't even know what they have. They'll be looking at something and they'll see that it looks like a VHS tape, but it's actually thicker than a VHS tape or it's actually smaller. What they probably have is a U-matic that's twenty minutes long, but they don't know that, so they won't know whether or not they have the playback machine. Or maybe they do have the playback machine and they feel confident about it and for the purposes of illustration, they can go ahead and look at some of the materials. If you can do this, randomly select 10% of your collections to preview and determine content. If you can not or do not know how, seek conservation or audiovisual assistance - which I love, because as a conservator, when I read that I'm thinking: I do not know what to do next, and I'm the conservator! They are supposed to tell me what to do! But anyway, the idea is to go find the experts. Do not bother proceeding with the survey; don't waste your time with the rest of this unless you know what you have, what you are going to be looking at.

But let's say we know what we have and we have descriptive material; we can go to the value assessment, which is part two. This is sort of “jargon-y”; it's the stuff archivists sat around a long time and thought about. Trying to stuff this into, say, an aesthetic collection, artistic or museum collection, might be difficult but let's work with it. We'll ask the following questions and give them points: Do the videotapes within this collection relate to the mission statement and collections policy of the institution, in terms of the topics that are being documented?

I'm going to give an example from my own archives of early videotapes of pandas. The Smithsonian Institution has pandas and many panda videotapes from 1972, half-inch reel-to-reel. These materials were watching the pandas every day, and used by the researchers to determine panda habits. How does this relate to our mission statement and collections policy? We collect materials that are related to the Smithsonian, and our mission statement says that we will preserve them. The topic that I'm looking at is on Smithsonian pandas, so it gets a two.

The purpose of our archives is to preserve materials and make them available to researchers. This is a major issue for us; if it is not available, it doesn't exist. Actually, that's like a line out of “Star Wars: Attack of the Clones” - it's actually a really bad line, but… If the material isn't here, it doesn't exist, which is not true. But it suits my stated purpose, so the answer is yes.

The next section is regarding institution users. We know who our institutions users are, and we know who our clients are. Knowing who your users are is an extremely important part of whatever you do, or whoever your customers are. We have lots of developing interest in pandas, so it is assigned a two there as well.

The general and specific functions and/or topics; this is a very archival term. The question is: do the materials relate to the mission statement or collections policy, with regard to your general/specific function, or your general/specific topic or area? If you're a rare books library and you have materials on 20th century industry, they don't relate. That's what this is about--whether or not they relate. In my institution, our general and specific function is to collect Smithsonian history. Things that are vaguely of interest to us might be things that we would capture off line, off of a television screen, when the secretary of the Smithsonian talks on Nightline or something like that. But that would be secondary; the videotape of his actual presentation made to our associates would be of primary interest to us. In the case of the pandas, we collect the research product of our employees; therefore, it is assigned a two.

The next section is relationships and repository goals to other repositories. Yesterday, somebody said, “We have a problem where we may have a videotape that isn't as good, but for all we know it may be the only copy of this particular artist's output. But I think there might be families out there or collectors who may have a better quality copy.” Within research libraries and within more established collaboration organizations, there are systems for determining where other copies exist. In this case, our repository may be very interested in collecting panda material, but the Bronx Zoo may as well. We may decide that we are going to share information, and they will collect certain types of panda material and we will collect certain types of materials. If it is possible, and you can make liaisons with other groups, you might be able to answer that question and say, “Actually, that institution over there is responsible for maintaining this artist's work.” Or, “That institution or collection is doing a really good job of reformatting materials, and I'm not going to spend my money on this item, when I have other items over here.” So that is where that question is leading. In the case of original panda materials, it's a two; it's a yes for me.

The next section is the relationship to the known world of related documentation with functions and topics. Again, it's a very similar question to who else is doing panda research? Who else does a better job? Maybe there is a panda research library out there that we should be sending our materials to because that is all they do. They are pandas, 24/7. In this case, we are the only people I know collecting live panda documentation from '72, so for us it is also assigned a two.

That is the first part: what is its value to your institution? The second question is: what restrictions exist so your items cannot be accessed? Does the institution have legal custody of the videotapes, or the expectation of obtaining legal custody? This is a great question for those people who think they have the most valuable collection of late seventies kung fu movies, or something like that, and they want to duplicate them. They are not the legal owners of that material. It may be obvious to all of us, but it is a good question to ask. Since we have legal custody of these panda videotapes, it gets a one.

The next question: are the videotapes accessible to users without excessive restrictions or hindrances? This is a tricky question, because if it is in half-inch reel format, it is not accessible at my archive because we do not have a playback machine. I will have to reformat it. What this question is asking you is if you theoretically had a copy of it, would it be accessible to users? Yes, there would be no restrictive hindrances or copyright issues--they would be accessible now.

Finally: are there resources to preserve and maintain the videotapes, or the expectation of obtaining resources? Well, I figure once I fill out this form, take it to my manager and tell them it's the number one priority, there will be resources available. So basically, for my panda tapes from the early seventies, I'm giving this a high score. You just add up all the numbers you have assigned. When I do that with this example I get a fifteen. If you had other things that were lower assigned values you would still add up those to get your total. Down below it gives you a value score. So maybe the pandas weren't a great example because it's the highest value material I have. As an illustration, sometimes it is good to have a medium value material. You would put the information in your little form, and would know you had a value score of fifteen, or an “A”. My caveat here is if your overall value is high or moderate, proceed to part three. If it's low, it's not worth analyzing. You already know it's going to fall to the bottom of your prioritization issues.

Part three is about risk assessment; we are going to determine fragility or obsolescence of the collection, and the need to reformat. Physical condition is going to be a matrix of the instability and the physical damage. There are a couple of problems with this matrix. First of all, there is currently no system for determining damage with materials unless you actually play them back. We don't have an acid detection strip - though Linda Tadic mentioned the Image Permanence Institute is working on some sort of detection system for determining condition in the collection. You may have had an opportunity to play back one thing and you saw that there was lots of gunk on the head, or as you cleaned it there were a lot of problems that occurred in your collection. You look at the physical condition of your collection and first determine the number of highly unstable materials. It may be a type of chromium metal tape from the late seventies or early eighties, and those are notorious for being unstable. Then you'd look for the amount of dirt or other physical damage. Under the “Physical Condition” category it says: amount of highly unstable materials. My panda materials are actually not in bad condition; they've been stored fairly well. Even though they are half-inch reel-to-reel, they are in their boxes and there is no evidence that there is a problem. I'm going to guess that it has a B value, that it is moderately unstable. Then after assessing the amount of dirt or other physical damage, such as poorly wound tapes or if you saw spoking--any number of physical things when you're looking at the box, you can give it a score. The actual physical condition is pretty good too, but I'm still going to give it a moderate value--a B. The idea with this little matrix you have is to draw a line. The unstable material is a B and the physical damage is a B. Where their lines intersect, it says C, so the condition, as I gave it, is low. That has me scared, because I know it's a half-inch videotape and I want to reformat it. But that's where the obsolescence comes in and I really didn't know how to address this issue in any other way than this complicated format.

First of all, you need to know what you have. If you know what you have, you are going to look at the format and the age. This is how I broke it down, and I'm very willing to add information to this. Any reel format - two-inch, one-inch, half- inch, except for one inch SMTPE type C, any unknown format, some sort of weird consumer format from the early seventies, any tape between 1956 and 1970- is going to be what I consider the most vulnerable group. The second most vulnerable is cobalt type metal-based cassette videotapes, BetaMax tapes from 1971 to 1980. The third most vulnerable is three-quarter inch U-matics that were made that were made between 1971 and 1986 (I think that 1971 is too early for U-matic) one-inch SMPTE type C, and the 8 millimeter BetaCam, and tapes from 1981 to 1986. Fourth, three-quarter inch U-matics from 1987 to the present, VHS, BetaCam SP, tapes from 1986 to 1990. And then finally, digital formats and tapes from 1991 to present. The breakdown of the dates is because I was following John Van Bogart's “Magnetic Media Storage and Handling Guidelines” from 1995. His research indicated that videotape has a lifespan of ten to thirty years, based on research done in the National Media Laboratory. We all have tapes that are older than thirty years that playback fine, and we have lots of tapes that are only ten years old that have problems. It is a rough guideline, but it is where my dates came from.

So how do you fill this out? There is a “most, some, fewer, none” column- three columns-and you choose one of the most. Let's say you have a collection which is primarily VHS tapes with a couple of three-quarter inch U-matics thrown in. You would find VHS and circle under most, that number which is twelve. In my case, going back to my panda videotapes, all I have are half-inch reel-to-reels. So they get the most. Now, I have few or none of the rest of them, and so I have to just basically circle the rest of because I have to get the right number. I believe I will get a forty if I add up all those numbers. But the concept is that you have to have one in most, and then you fill out the rest according to some, or few, or none. That should give you this rank. When I put that in the obsolescence score, it gives me a score of forty-eight and that is high, or an A.

So my condition is C, my obsolescence is A, and now I have to do the level of risk. I'm going to do my little line; when I do that, my matrix ends up indicating the need for reformatting is still high--an A. I've gone through this entire system. Some people will love it, and some will say, “This is just too many rules, I don't like it.” I think it's very useful for people to think along the lines of the way that librarians or archivists think sometimes.

When we finished part A, which was the risk assessment, physical condition, my condition score was a C and part B, my obsolescence score, was an A. Then you take box two and box three, you put them in the matrix, and you do the same matrix thing you did before. Obsolescence score is an A; and you just draw a line at A. And the physical condition over here was a C; and you just draw a line at the C. Does that make sense?

QUESTION: So why does that come out as A?

STAUDERMAN: Because where the lines intersect is the A. Sorry. It's like telling people stuff they should have a fifty-page instruction booklet with.

Masters and originals, unique copies, are the most valuable part of a collection. Audio/Visual collections should contain user copies as well as maintaining the masters in their original format. In addition, a duplicate master in a contemporary format should be considered. Does the collection have user copies? If yes, consider whether these copies are adequate for users; and if no, consider increasing the value of the risk score. There is subjective measurement there. Does this collection have duplicate masters in a contemporary format? If there are duplicate masters, it would probably not even be a reformatting issue and you don't have to do anything; but if there are none, you need to increase the value of the overall risk score.

Now we take this A, the risk score and go back to the front. We know our value score was a fifteen, or an A, and we also know our risk score was an A. When I do my little crosshatch on this front template, it becomes a one. This is my most valuable collection and I have to reformat it now. This is a prioritization scheme. One of the reasons why I never have showed this to anybody is because one, I think it really needs a lot of instruction and it needs more than just my thoughts on it. I don't know if it would be useful for people, or if on a theoretical basis it's generally useful. On a practical basis it may not be very useful. I don't know.

There are questions that are asked in the midst of it. Maybe you need to have a survey for the actual sampling of the physical condition of your materials and that's not a part of this. This is asking broadly: Do you have sticky shed, or do you have dirt? I mean, if you needed to do an actual survey, I have a two-page survey that is very complex and shows all sorts of information. But surveys like this general prioritization survey, at the end of the day, are probably only as good as some of the analytical tools- some of the sub-survey activities that went on while doing individual item surveys and so forth.

QUESTION: I think this is a great first step. Although we were doing the panda tapes, many artists are also dealing with camera masters, rough cuts, and final edits. How would this factor in the relationship? Because if you're dealing with half-inch or three-quarter inch tape, and you have a whole box of tapes that deal with this one edited master, obviously the cost of preserving all of it is one factor--versus just the edited master. So I think this also might be a useful part of the paradigm.

STAUDERMAN: Just by saying what your collection size is, you could say “I have eight hundred tapes that comprise this collection.” That's just at the very top, right next to that it's your highest priority. Obviously, this is only part of your entire cataloguing system, your approach anyway. But I agree: how can you use this tool effectively within your collections, when you have all these issues?

QUESTION: What I find helpful is part two, with the questions about ownership. That is great. In working with some public television collections that were offsite, and many boxes, that dealt with this issue: to pull boxes that have masters first. We talked about a collection policy and then pulled the boxes according to that. We pulled masters first and high quality dubs of masters, or clones, and then we pulled camera originals, original audio materials and original film. We worked our way down that way. For producers, I would encourage that whole masters- or camera originals, audio originals and that stuff - should be close to the front.

STAUDERMAN: At the top, yeah.

QUESTION: Regarding the economic issue. There's an interesting statement in item D, where you say masters and originals, unique copies are the most valuable part of the collection; and audio/visual collections should contain user copies, as well as maintaining the original in its original format. I was haunted by what Kate was bringing up yesterday about she had sixteen hundred U-matic tapes. Now she'll have sixteen hundred U-matic tapes, and sixteen hundred sub-masters of the tapes. We have not yet talked about increased storage costs when we talk about remastering. We talked to John yesterday about EIA's storage, and how expensive that's becoming. As soon as we make duplicates, we have to put the originals somewhere. We can't throw them out, because we will invent new and better formats.

STAUDERMAN: You know that is a debate. We could have a good debate about what it means to keep your originals. Obviously, it's going to come down to an institutional ability to do that. From a conservation background, that's what we do. The notion that there's an intrinsic value to those tapes, that somehow we might, in the future, be able to get a better image off of the original tapes if we just can hold onto them long enough, is where that is coming from. Yet I understand storage. If it's not worthwhile to hold onto these original materials, if they're going to end up being put in somebody's garage - which is not going to be adequate. They need controlled storage, and it needs to be maintained, while you're maintaining everything else. I have a hard time convincing my bosses about maintaining those things.

QUESTION: I find this forum to be incredibly useful. I think it's really interesting, because I wouldn't use it the way you use it; but I would use it. I would modify this and use it to evaluate individual tapes in our collection, and I would tag all those tapes with a modified version of this. What I see that is really important is that it is a terrific way of translating, a sense of values- and information -about a specific tape to a younger generation of workers. In some cases, they don't know the value of the individual piece and in a lot of cases; they don't even know who the artist is. So if all this could be tagged and evaluated, with a prioritized list - then that can be carried through time for a group of people who don't actually understand the importance of the original material.

STAUDERMAN: I remember you said yesterday that you also had to change or mix into some of your grant proposals some things you knew were important, but that you didn't feel contemporary grant reviewers would understand as important. The thing that something like this tool does for you is it says “we have a legitimate system of going through our collections and determining for ourselves what's important. You don't need to know who these artists are; we've documented why they're important. This is our legitimate system of review.” Then you don't have to worry about whether or not they've heard of your most important artist from that generation doing whatever they're doing. That is something that might be useful.

DARA MEYERS-KINGSLEY: What about adding the issue of appropriate storage conditions to this form? In other words, where are the materials presently stored, and once you've made copies, will they be appropriately stored? I would assume that becomes a risk, or physical condition part of the survey.

STAUDERMAN: Well, this is where this needs a narrative, you know?

QUESTION: Is the collection catalogued?


PIP LAURENSON: I love this way of thinking about it. I think what conservation adds at this point is that we have this obsession about risk assessment, which is giving us a way of thinking about that which we should really worry, and how we should direct our anxiety. This is what conservators do, they worry a lot. One of the things that I would like to see come out of this meeting is that we take Sarah's form and think about what the areas are that we need to enhance to get more accurate assessments. One thought I had was we need to better understand what sort of failures we have in playback. Quite often we have something categorically different between digital failures in playback from the analog formats. We've got different types of failure going on, which involves different notions of risk. We have points where error correction mechanisms hit a peak, and the sort of creativity involved in getting those back is different from cleaning a tape that's got sticky shed. That was one thought in terms of trying to determine where the research needs to go, so we can make some assessments there, and with the storage issue and the nature of risks with different formats.

For example, there is information from Sony that they will be making a real effort to go backwardly compatible with their half-inch formats, which they are not doing with their three-quarter inch formats. That information is there, but at the moment, we're relying so much on the rumor mill. To collate this systematically around this type of form and system, I think, would just be fabulous.

KATE HORSFIELD: The form itself, I think is excellent. I think for those who are here who distribute the works it is a very different situation than an institution which has a collection. We may have works (this would be internationally) we may have works in common, just distribution dubs. As we move forward in time, it may be necessary for some specific places to take a custodial responsibility for a particular artist's work, I would say. That may be something that comes up more. It would be EIA looking after a specific artist's work where copies may exist in other situations, as you were talking about. But they also may be with other distributors. That may be more a Canadian problem than an American one, but it is definitely an international problem. As not only the tapes but the artists age, that may be an issue that we'll have to come to terms with. It doesn't look like the notion of exclusivity looked ten years ago - which was what we were looking at as distributors. It may look like more of a conservation responsibility.

STAUDERMAN: Great idea. The Association of Research Libraries decided about fifteen or twenty years ago - to not have everybody doing microfilm, but to have a system so they weren't duplicating efforts. There can be very formal, or there can be informal ways for knowing where this stuff is so that there is not a duplicate effort. It takes a lot of work, but it's worth doing.

ELIZABETH WEATHERFORD: I think this is prefigured by the IMAP project, which has been trying to get across this notion of registration. But obviously, the purpose of the registration of commonly held titles would be to lead to these types of proactive responses. In one case we might have an archives, in another case we might have a library collection of independent media. The libraries strike me as very close to the distributors, in the sense of taking on common purpose, and archives having a slightly different set of relationships to the uniqueness of its materials.

JIM HUBBARD: In terms of the IMAP template, there are about thirty-five to forty users now, and we're looking for funding to do a pilot union catalogue. We hope to begin work on this by the end of this year, with three to five collections, and get the concatenated records on the web. This is also part of and a model for prefiguring a much larger project, what's known as the AMIA MIG, the Association of Moving Image Archivists Moving Image Gateway, a project that AMIA and the Library of Congress are co-sponsoring. It will be a much larger group of archives and collections and conceivably, it could have every collection in the world--although practically that is not going to happen. These sorts of mechanisms can facilitate and it is one of the primary purposes of these projects.



Looking Back/Looking Forward: A Symposium on Electronic Media Preservation

May 31 - June 1, 2002

Assessing a Collection for Preservation 1