>> From the Library of Congress in Washington, DC. ^M00:00:04 [ Silence ] ^M00:00:17 >> Fenella France: So welcome everyone to today's topic and preservation series, presentation. Delighted to have you here and welcome also to everyone we have participating online. The presentation today is by Dr. Eric Monroe. And this will be on Broken Cylinders: Uncovering the nature of damage to early wax cylinder audio recordings during storage. So Eric is a supervisory physical scientist here in the preservation, research and testing division. He received a Ph.D. in chemistry from the University of Illinois in 2008. Following a postdoctoral work, following postdoctoral work at the University of Alabama, Birmingham and the University of Arizona he was a chemist in the corporate characterization lab and came to the Library of Congress in 2015. I was delighted when I was talking with him when we started talking about degradation mechanisms and reverse engineering. I thought hum, I think we have a project that's waiting. So without any ado I will hand over to Eric and we will learn more about fun things we'll be doing with wax cylinders. ^M00:01:20 [ Pause ] ^M00:01:31 >> Eric Monroe: All right. Thanks [inaudible], thanks to everybody for coming and those who are online. What I wanted to talk to you about today was the work I've been doing with wax cylinders. Before coming to the Library wax cylinders, I said what is those? So what they are is turned out to be quite exciting. They're the first successfully commercialized recorded sound carriers. So about four inches long and like LPs, vinyls, they groove media. But unlike modern disc groove media these grooves are very wise and very shallow. This presents a challenge I'll talk about later. And also unlike the normal grooves the audio is on the bottom of the floor, of the groove. So you can see kind of a topographical map on the bottom right or actually this is an optical image from a cylinder. You can kind of see how the needle of the stylist would write along this groove, it then vibrates a plate and that produces the sound. So this was a massive change in humanity basically. Now we can record our sounds and go from there. So, one of the manufacturers was Thomas Edison's companies. It had a couple of different names over the years. And he had a group that would produce the material, they would sell the material to the company that would then cast it and put the audio on it. But over time they had a couple of different iterations. They started off in 1889 releasing this brown wax on far left. They were the first manufacturers to use a homogeneous composition. They kept that for a while. Had some difficulties in how they were doing the reproductions such that it was an instantaneous recording. So, for the most part the band was in the room with the cylinder when the cylinder was made. When we, toward the later portion of the time before they moved to the black cylinders in 1902 that they came up with a nifty way of using pantograph kind of situation where they could have a number of master cylinders and then record a handful of others, five or ten at a time. So needless to say they got, decided that that wasn't necessarily commercially feasible in the long term so they worked really hard. Came up with a way of making molded cylinders, that is that gold molded cylinder in 1902. From there they changed their formulation quite dramatically to make a four minute cylinder. So instead of having just two minutes of audio they can now put a whooping four. They did this just by making the grooves smaller and upping the density. And then they finally moved to using a celluloid based material in 1912. This talk is going to focus entirely on the really old stuff, the brown waxes. And so here is kind of images of how they were doing these recordings in these commercial cylinders. So you might the entire band, I believe that's the Navy band on the left and they are recording nine cylinders at a time. That's high throughput for them. So they would play the song, pull it off, put on new cylinders, fire up the band again. So they were making relatively few of these at a time. And particularly, you know, they were useful, interesting and the, were used a lot as entertainment. The very first applications were using these auto phonographs on the left where they would be touring casements with a phonograph player inside with one cylinder. It would sometimes be changed out once or twice a week. You go in, put your nickel, put the earbuds in and listen to the song. They also had of course home models with a reproducing horn. And then often times they would be used in kind of parlors. Now very interesting, we have a lot of cylinders that have music on them. That's not where this project came from. It was learned and utilized very early on that these wax cylinders could be used to make ethnographic field recordings. So the Library has approximately 10,000 of these things. And they start, they date back to as early as 1890 with the [inaudible] recordings. So we have all of these. They're down NAVCC, the National Audio Visual Conservancy Center. And we're slowly going through and looking to do digital transfers of these cylinders again. They've been transferred a while ago but now that technology has gotten better we can get much better audio of the material. Unfortunately sometimes when we open the box the cylinders have damage. Sometimes it's a crack. Sometimes it's a service [inaudible]. But the biggest concern was the cracking. When I mentioned before that those grooves are very shallow, they're only about 10 micron deep, that's the 10th of a human hair and about 125 micron wide. It's almost the reverse for vinyl. So if you try and play that with a stylist it's going to skip and often times damage the rest of your cylinder. We don't really like doing that. We are, of course do have a developing Irene which is a noncontact playback method. But for kind of high throughput and some of the best recordings come back, still back playing with the stylist. So that's where this entire project started was, the question is, why are they breaking? Why are they breaking? What can we do to stop it? And really is this a ticking time bomb? These cylinders were commercial or basically a commodity good. They weren't meant to last forever. They actually in the very early stages when that particularly the auto phonograph, it was kind of a selling point to the for bands and musicians that these things actually have a limited playback time. So you're guaranteed to be called back in to record with us again if you sign now. Long term employment. It's great. So when I was thinking about how can we, how should I develop a project to try and address this situation. Just coming in from industry, it was okay. I'd been doing reverse engineering of materials. I've got all the tools in lab. So I can go do [inaudible] analysis, identify what's there, make the materials and then test them again. Are they the same, are they different? Also thought of well let's go back into the records. What do we know? So I started looking at some 20th century reports, various books, papers, things like that. And the stated composition was not the same in any single one of those records. Well okay, that was a kind of strike. And I'm thinking okay, maybe I really need to do this whole reverse engineering. This is going to be much more difficult. Then I got lucky. I got lucky because at Rutgers University is currently digitizing all of the Edison Company's papers. And as the material was initially developed by Thomas Edison's companies, that's in there. I say I got lucky because they had digitized to 1905. And in 1905 there was a lawsuit. The lawsuit was between, so not, sorry reverse. Not only is it a lawsuit but I also can then go through and grab lab notebooks, which as a scientist is really cool. So the lawsuit. So I could spend probably a good hour just walking you through the ins and outs of this and kind of the funness with that. Unfortunately there aren't many images that can go with it. So I'll try and keep it relatively brief. It's still kind of funny. So National Phonograph was Edison's company. As I was mentioned they had, they had released the brown wax in the finalized form in 1889. At the same time American Graphophone was producing their own cylinders. Large, their cylinders were paper core with a thin mineral wax coating. These were not very robust. They were prone to cracking. So they were really not having much of a market share. They were basically being taken to the cleaners. So a guy by the name of Thomas MacDonald got in touch with American Graphophone, signed a contract with them to produce a much better material. He bet the farm on a lead based composition and said this is the way we need to go. That composition had a lot of problems. They would mold the cylinder very similar to the Edison cylinders and it would look fine initially. But in a day or two it would haze over the surface. Haze over an [inaudible] surface when you're trying to play a stylist with very tiny little bumps at the bottom means it just doesn't work. So they were getting lots of complaints. Basically were going to have to do a recall. So MacDonald's bosses came to him and said okay you have three months to make a composition that is as good or better than Edison's. Get it in the factory and get it sold otherwise you're fired. And if you look at the details in your contract you owe us damages. Needless to say he got a little bit creative at that point to the tune of three different mechanisms that were effectively industrialized [inaudible]. Two industrialized [inaudible] and one things that happen in industry. So he got in touch with the procurement agent for National Phonograph and said I'll give you $500 if you tell me what's your ordering into the production facility. He then hired away one of the mold makers. So he was just dipping cylinders all day every day saying I will hire you. I'll pay you three times what you're worth. You won't even really have to work for us. All you have to do is go across the hallway to where they actually make the composition and bring me that recipe. The third thing is he got in touch with a soap maker out of Evansville, Indiana. Sent him a couple of cylinders and said please replicate this. The soap maker turns around, decides yeah I'll do this. What we'll do is I don't know how much this will cost us but just pay, I'll keep track of how long it takes us, what the costs were and just reimburse us for that. This is kind of an interesting challenge for us. It's something we don't get to mess around with a lot. A couple of weeks later that guy has reverse engineered the material, recreated Edison's composition and sent the cylinders and the formulas back to MacDonald. MacDonald then now has these three different recipes. All are basically the same. There is a slight difference with another one because Edison ended up making their composition two different ways. I'll talk about that in much more detail later. So MacDonald then turns around and patents it. He applied for the patent in 1896. It was awarded in 1898. It has only his name on it. He did not of course let the patent office know any of those three levels of trickery when he received the information. So now this is a lawsuit in 1905. That's nine years later. What happened was when, so the two companies ended up just manufacturing the same material over that time period. Edison knew that he didn't really, he couldn't do anything because he didn't patent the material. He was using it as a trade secret saying no one will understand what my secret sauce is. So there's the patent. And so the only way that National Phonograph could not lose this suit very badly and particularly since the lawsuit's basically saying you've done, you've been using this composition since I invented the composition so you've got now nine years of infringement. But what triggered this was Edison's move to the two minute black wax cylinders. The gold molded process. The two compositions are very, very, very similar. All they do is they changed a little bit of the ceresin wax to ceresin and carnauba wax. That is a derivative work. We would probably both invalidate Edison's patent and possibly lead to a large loss on Edison's part. So in order to get around that all they had, or they needed to show prior, they needed to show, no we actually developed this back in 1889. And we've been doing that ever since. So in the deposition is many, many, many, many pages of Jonas Aylsworth's testimony. And in this process he says I am the one that invented this. In 1888 we finalized the formulation in 1889. And these were all of the steps that I took to get to that point. And we were producing this form, see lab notebook experiment X at XY&Z places from these time periods. And then we moved to here and did this. We did this. We did this. We did this. And then we changed the formula because of this reason. And these are the two that are in the patent. So this got quite interesting. And so, on the bottom right is just a, from the brief were saying that the letters patent hereupon is invalid and void because the same was obtained surreptitiously and unjustly for the invention which was in fact invented by the author or others who were using reasonable diligence. Basically we created it and you didn't create it. Your soap maker made it for you. So this now gives me a great wealth of knowledge. Not only do I have all of the legal or legal lead, basically binding recipes, I also have reference to all of the lab notebooks which I can go into and science geek out at. So what is this material? Don't get concerned by all of the chemical structures. It's relatively easy. Basically he used the process used to make soap with two minor changes. Instead of using a large fat which has three of these fatty acid chains, they used a purified steric acid. Back then it was a one to one blend of the steric and palmitic acid. You melt that down and you add in some base, sodium hydroxide or sodium carbonate. They used both at one point in time. That is the same process for making normal soap. The magic here was they didn't add so much base that it turned everything from the acid form to the sodium form. They also did this in the presence of aluminum. The aluminum was their, was their secret sauce. That's the, that aluminum stearate gave them the right physical properties for wear resistance and the ability to mold and basically make nice little chips. If you had too much sodium stearate it goes crystalline. If you don't have enough of the sodium stearate it is too soft. And so this was that thing that was in the middle to give them some leeway during production. They also then had to add a little of the ceresin wax to basically soften the material and to make it water resistant. So I want to go back in the lab notebooks. And so having these kind of lab notebooks gives you an eye into what they were actually thinking when they were developing this material. And I actually highlighted a number of things that weren't in the patent. A lot of the failure mechanisms that we see now with the cylinders nowadays, they knew all about back then because they were actively developing their materials to not have those problems. The various hazings, the surface defects, they were all noted. And well I've got his lab notebook on my iPad. I've read through most of them. Thankfully he has pretty reasonable handwriting. So this experiment 857 is his ah-ha moment. Up until this point he had been using, he had been going similar route as American Phonograph had using various mineral waxes in various combinations trying to make just the right properties. [Inaudible] say well what if I make some soap? So he did. And it, what this was was he took his steric acid and sodium hydroxide and did just the partial saponification. So he did, he just added, you know, about 50% of what he needed for the base. And when he cast that carb to groove he says it cut to almost perfection. He then continues, until basically a couple of days later when it became a little bit crystalline so it made a lot of pops and it was very sensitive to water. So he's like okay well but this was a changing moment. This was the basis for everything that happened over the next 20 years. So he realized at that point if he added another metal he could have, start to tease out various properties. So here 871 he stumbles upon aluminum acetate. So here he is adding the aluminum in in this form in the hot molten magma bath it's basically like napalm. It's about 400 Fahrenheit by the time you're done with it. So it's, you don't want to get it on you. But in the process this releases acidic acid so he doesn't have to worry about having something else in the soup other than his pure aluminum. It worked nicely. He's like well let's do some other things. So he spends the next couple of months looking at all sorts of other metals, lead, nickel, iron, you know, the list goes on and on. He had about another 30 or 40 different metals he had looked at. And decided no, aluminum was the best thing for us. He then went and had to optimize his method so he could then scale this up from doing it in a couple of pounds to doing it in a production scale in large metal kettles. So then they start making that material. That was in December of 1888. They then cast the number of cylinders, took them to their recording department. And the recording department was complaining because it kept wearing down their shaving knives. So it would come out the mold and then it be made smooth. And they couldn't keep a sharp edge because the material was just too hard, they couldn't cut it in a production scale. So, they went back, he's okay, I'll just add oleic acid, it'll soften, it will be great. It was great for a couple of months until summer happened in New Jersey. So June, 1889 now they're getting all sorts of complaints because well yeah it's softer, it takes a groove great, it sounds great. But then humidity comes along, a little bit of heat and now the surface is no longer shiny. Now it's just exuding what he later figured out was the oleic acid. So he knew he had to get rid of that but still put in something softer. So that's he stumbled upon the ceresin wax, it's just a hydrocarbon. He can throw that in. And it has the added benefit of not only making it softer but removing the oleic acid. From there he wanted to change the way they were adding in the aluminum in a large part because they were trying to up the process and if you didn't boil it long enough you had acidic acid there. If you had any acid leftover as soon as it came out of the mold and saw a little bit of moisture the entire cylinder would just haze over. Again, not what he wanted. So he'd also decided okay well let's make the aluminum [inaudible] offline. This will allow us to, we'll just make a bunch of the aluminum stearic because we added in relatively small amounts. So we can make it in big batches instead of in one pot. Set it aside and just kind of sprinkle it in as seasoning whenever you need to. He then cut costs by changing from sodium hydroxide to sodium carbonate. That's also a whole lot safer because to add the sodium carbonate as a powder instead of liquid sodium hydroxide into 400 Fahrenheit boiling pot of effectively oil. This was a method that they then ran with for a number of years. Jonas Alysworth actually left this project. And I think at that point in time went on to go work on I think it was the lightbulb production line. Until they started to have a lot of more problems about six years later, so 1895. It turns out this is also the same time that MacDonald was snooping around. So I said that there was two different copies that he had. One was the second, this other method where when Alysworth came back and was identifying well what the problem was, he looked in the vats. And when they were making the stearic acid they're supposed to put them into pans and cover them. Well production had gotten enough and they had been doing this for long enough. They said awe we'll just cut out the whole cover it problem. So they were making the aluminum stearic putting it in pans and then leaving it in the room for a couple of months until they needed it. So it was just dirt and grime and dust because it was an industrial situation. So he went back and okay, we'll just use this all in one pot and we'll just throw in a chunk of aluminum. It turns out it actually works and it was quite simple to do. It's all one pot, throw everything in, boil it until it stops boiling, cast it and we're good to go. So these were the two methods that he had used. And needless to say I've got a lab, I'm going to make these. I made not only those last two that they did the long scale production but each of the ones that I had just talked about was something that they had produced commercially outside of that very first one at some point in time. When I mentioned that the material's quite easy to make. The ones that are on the bottom for reasons I'll explain here shortly were all made by a high school graduate student that I had as a summer intern last summer. So we made these materials but we actually learned something in the making of the materials. One of which was that my materials were too pure. In my first iteration of trying to make these materials I was using stearic acid that I purchased from a chemical company. It's 100% pure almost. In going through that process I always saw crystallization. It wasn't until I was flipping back through the notebook and he noted, our stearic acid, we had these specifications. It can be 50% palmitic acid, 50% stearic acid and no more than 2% oleic acid. Oh, yeah that'd do it. Just having that slight different length was needed to make that amorphic material. When I started blending the two together it went from something that had always formed crystals on the surface to something which is perfectly amorphous and by hand is identical to the materials I was seeing from way back. Also learned that the color of the cylinders is a, it's a function of the time, temperature and metals that the material sees. So these things were made in a cast iron kettle, kept warm for a while. They then were poured out into pans to solidify. They were then taken across the hallway when needed, the cylinder molding room, they were melted down again and then they were cast into cylinders. So by the time, the amount of time I just kept this thing boiling can dramatically affect the color of my material. This was somewhat novel in looking at those people who have been looking at cylinders and mostly home hobbyists that are interested in these materials. Some of them had been thinking that because there's such a variation in color of the cylinders that it was actually that they were always tweaking the recipe. Well the data or I can replicate that change here as well as looking back in all the records to know it was a very, very stable composition that really only had two compositions they were ever really making. So I also mentioned metals. So when we started looking at the composition of cylinders from our collection we saw that more than just sodium and aluminum as the metals which we would expect to see. So here using x-ray fluorescents, basically it's a little x-ray gun, put it up next to the thing, pull the trigger, the x-rays excite the metal atoms that are in the material which then emit another photon with certain energy that we can measure. And that energy is, corresponds to the elements. So we can identify what they are. In this case we can't do it quantitatively but we can see what elements are actually there. So we don't have much of a response for sodium and aluminum here, that's because it's a limitation of the method and how I was running it. But I do see a lot of other things. I see a fair amount of Sulphur, that was a contaminant in the ceresin. Calcium shows up from time to time, nickel was an interesting one as was the iron. So those came from the kettles in which they were making the material. More often than not they were producing in cast iron kettles so just a contaminant from that. The nickel was from a very brief time period. I think it was, it was less than a year where production was moved to the Silver Lake facility because they were renovating another manufacturing plant and they had nickel kettles. So I actually had one that likely came from there. We also had a couple that hit positive for lead. Those were likely from the American Graphophone or some knockoff there who was using their composition. So we have all these metals and going back to the lab notebook. Oh, oh, what's going on? Do these metals have an effect on the properties and material? I can make the really pure stuff but does that really simulate what I have? Does it matter? Well it turns out that the having iron at levels which we detected in the material did change the color. It actually made it a whole lot more similar to the cylinder material that we see in the collection. So it does have an effect on color. Next question would be well how does this relate to overall properties? So really it's now time to get down to the nitty gritty and the science of what do, what did I make? Are my materials similar enough to the old broken box of sad things? Thankfully most of those are blanks. Unfortunately not all of them. And in our collection again sometimes we see cracks. But the question again is well what changed? Is the things that I've just made are they the same or as what I have in the box whereas when the box broken because things are degrading? In which case it's all hands digitalis time to just digitize everything used and the best techniques we have in order to not lose any more information. So one way to do this is to look at the exact or the quantify the level of metals in the cylinders. So I can do this very sensitive or very sensitively by taking a small scrapings from the inside of the bore of a cylinder or of a fragment. In this case this was a blank with a big hole in the back you can't see. And this amount of material is enough to do about ten analyses. So I can use a very small amount of this material. We then digest it down and shoot it into this fancy piece of equipment, inductively coupled plasma with optimal [inaudible]. Basically it throws a liquid into a plasma similar to what you see if you put a fork into a microwave except this is in an argon atmosphere instead of air. So it's an argon plasma. It has a little coil just like your microwave does. It's very, very hot. It excites the metal elements similar to the XRF that I was talking about before. It emits another photon. We measure those. And now we can identify what elements are there and at what quantity. When we do that we get a whole lot of numbers. Don't really need to go too far into this one. The most important thing is to look in this top right. This is the only cylinder that I have tested in this way that has a known provenance. This was a cylinder from 1890. That was when they were making composition 1058. My aluminum and sodium levels are very, very similar. We were making what we thought we were making or I was making what I thought I was making. They were making what they thought they were making. Using all of these numbers is where the project for my summer student came. And so he doped in all of these materials or all of these other metals into the compositions in order to make those, that second set that was much more brown. So our metals are there but none them were radioactive metals. So they're not decaying, they're not going anywhere. They are what they are. So then it's what about the other portions of the cylinders? What about the organics? So for that we can take a little glass rod. Touch it to the interior of the bore, of the cylinder and put that in front of our mass spectrometer. So in this case I've got yet another plasma, air plasma, not actually helium, sorry, coming through the blue piece of instrumentation on the right. That then bounces ions off of our glass surface that transfers the charge. And desorbs our analyte which then gets sucked into the vacuum of our mass spectrometer which tells us the molecular weight of the things we just threw into it. Once we do that we get some other data that looks like this. What it is is molecular weight on the bottom and relative intensity on the y axis. I can identify almost all of those peaks as salt adducts or fatty acids themselves through this process that sodium and aluminum stearate component. The metal gets knocked off so we just see the acid forms. And these three higher were here are complexes of stearic and palmitic acid. So it's two palmitic, two stearic or one of each stuck together. The really important thing here is I've looked at this in great detail. And in many different ones of these I've taken from different cylinders and all of mine. And I really only see two differences. In the period brown wax on the top I see the presence of oleic acid. I don't have that in mine because I didn't add it. I'm using pure materials. Later on I did add in oleic acid to get that kind of signal, it's right in that one or two percent which was the specification they were using in the way back. Kind of cool. I also know where my suppliers are getting their stearic acid supplies from. It's coming from cows because I see margaric acid, margarine. So but I see no evidence of any degradation products. I would expect to see either something down here or something in through here that's different between the two if I was seeing some degradation. Either a reaction that things come together, gets a slightly larger mass or things fall apart, they get smaller. So the stearic acid and palmitic acid aren't degrading, okay. What about the ceresin wax? Didn't see that in using the previous method, the [inaudible], so I dissolved the stuff up in some chloroform and shoot it across a different instrument, a gas chromatography, mass spectrometer. Basically it separates the molecules out by roughly molecular weight or hydrophobicity and then puts them into another mass spectrometer. Breaks them apart so I can see what they are. This is just a trace of the separation. And you can see that the two period fragments are, well, pretty much indistinguishable from what I had made in lab. The bottom is the raw ceresin wax that I was using. I've used a couple of different ceresins, the slightly different molecular weight. But even the top period looks more like the ceresin than the period, to period comparison but it's still the same as my lab prep. So now I've looked at everything and I see no differences outside of some of those additional metals. Time to look at some physical measurements to see if I can explain that. But it doesn't appear that we're actually chemically degrading. This is great. It means we likely have some time. But it means at this point I didn't know what the cause was. So let's look at hardness. Hardness was a big thing that they kept measuring and harping on in the notebooks. So I have all these previous cylinder fragments to the left that were melted down into little pucks as well as two of my compositions, that 1058 and the 38 were the commercial forms. And I'm right at the right hardness as the period, okay, that's good. That 892 and 957, those were the materials prior to adding of the ceresin wax where they said it was too hard. Okay, it actually was too hard. So then I decided okay well one thing is when I cast these things in these little cups, little melt aluminum cups they shrink and I can easily pull them out. It makes analysis really easy and you can see how that would be in a little mold. But how much is it actually shrinking? So I did a series of experiments where I took these cylinder pucks and measured the diameter in multiple places across it and multiple temperatures and humidity's. Did the hardness of the same thing with temperature and humidity. These things really do not care about the humidity, the differences. But I do see a sizeable expansion when I change the temperature. I can quantify that by using the coefficient thermal expansion. Basically it's a measurement of if you change the temperature of an object x degrees it will expand or contract by this fraction. And what I found was these materials, so the period cylinders as well as the 1058 and 38 compositions are, have a coefficient of thermal expansion of about 200 parts per million per degree Celsius temperature change. What does that mean? Well it means that these things are changing quite a bit. A cylinder will expand or contract about 10,000ths of an inch at the 20 degree Fahrenheit change. To put this in perspective modern plastics their coefficient expansion is about 150. These materials do expand and contract if you're making a mold to produce a glass or something. You need to account for that shrinkage otherwise you're not going to get the right measurements. But those polymers that make up the plastics have molecular weights in the millions. They're incredibly large. Stearic acid and palmitic acid have molecular weights between 200 and 300. There's just not a lot of interaction there that can hold the materials together. Seeing this data in the first iteration, so this is about 20 different samples and there's 55 across all of these. Most are in the 1058, 38 compositions. When we first saw that information we immediately called down to NAVCC. How are we dealing with moving the cylinders from the collection spaces into the recording studio? Because I think we might need to talk about this. So in doing so, it's like okay, let's make cylinders to make sure before I undergo any big process change that the theory at, this is just a thermal effect could be plausible. So I don't have any molds from the late 1800s sitting around. So I went down to Home Depot, picked up some iron pipes with a little bit of silicon in the middle to hold it together so it doesn't leak all over my bench. And cast some pseudocylinders. I'm calling them pseudocylinders because the diameter's not quite right, it's a little too small. But the wall thickness is spot on for period, for historical cylinders. Could then take that and be harsh to it, really mean and cycle it from zero degrees Fahrenheit to 100 degrees Fahrenheit in the time it took to walk across the room. I'd done some measurements on how rapidly the waxes will equilibrate to a temperature change. And the answer is normally about five minutes. So we're doing this much harder, much faster. But in just cycling 8 to 12 times we were, we're causing cracks in our cylinders that look just like what we're seeing in the period. So it's not that the materials are degrading but it's been thermal shock over the years. Now we're taking it really quickly but changing even 20 or 30 degrees or being in an attic in the summer or a garage in the winter and the summer. And having that thermal cycling in the course of 120 years it makes plausible sense. Our cylinders are living in a very acclimatized space. They're back in the vault at NAVCC. They're actually back behind this facility in that mountain in Culpepper, Virginia. And so the vault conditions are very stable at 50 degrees Fahrenheit, 50% relative humidity. The cylinders had been going through an acclimatization room which is at 65 Fahrenheit and 30% relative humidity so slow the transition of the humidity, to slow that transition. Because all of the other recording materials down there they are more sensitive to the effect of humidity changes on them rather than the temperature. Here we're the opposite because then they were going to the recording studio, you know, at 72 and 30. We did a very, okay not really. So sometimes the best solution is the easiest one. Now instead of our cylinders going from the vault to that acclimatization space on a cart in the little boxes they now get to take that trip in a fishing grade Igloo cooler which can slow that temperature shift from a couple of minutes to now 24 to 48 hours. So we're trying to make that transition as slow as possible and limit the number of cycles that they see that anyway. The same time it makes it very important that when people are handling the cylinders they do what they've been doing since they were taught how to handle cylinders which is to keep them moving. Because just the touch of your hand can change the local temperature. The four minute black [inaudible] are really notorious for that because they're much more brittle. It turns out they have the same coefficient thermal expansion as these other ones. So not only were we able to say that yeah our collection is in relatively good hands, it's now in its happiest place they could possibly be as the temperature and humidity are now nice and constant. And we take steps to slow any sort of thermal transition that we may have. With that we think we can reduce the risk and has come together what I think at least is a really kind of interesting story that all started off with the really kind of worrisome. Are we going to lose this entire collection? And so we have to just take everybody in do this mass digitization. What it really likes like now is the cylinders look like they're going to be quite stable. They're much more stable than any of the current discs that we have with CDs, DVDs, Blue Rays, things like that. So now we can decide what do we need, what kind of effort to we need to do the preservation by saying okay what is at most risk? In this case the cylinders are much better off than we thought they were to begin with. And so with that I need to acknowledge a number of people. Fenella and the folks in the PRTD have been great help. So was Andrew Almeida, he was my summer intern student. He's now a freshman at the University of Maryland. Peter Alyea for a lot of helpful communications as well as NAVCC and the Folk Life Center for really making this project possible. With that I'll gladly take any questions. ^M00:43:05 [ Audience Applause ] ^M00:43:11 >> Fenella France: I just want to say I think the only time in my life when someone on my staff has walked into my office and says I've just broken something. I went yes [inaudible]. So we'll open this up to questions and Eric will actually repeat the question for our online folks as well. ^M00:43:32 [ Pause ] ^M00:43:37 >> Eric Monroe: No questions, oh. >> Hi there. I think I saw a [inaudible] image, a 3D of that. I'm wondering if you're considering using visualization of the cylinder before, before subjecting it to imaging or [inaudible]? >> Eric Monroe: Sure so the question was whether or not, since I had the 3D image of the grooves if we're using that prior to doing a transfer, something like that. That is the Irene process in snapshot. So it was something that the library's been working on and developing with folks at, I really lost it, at Berkeley National Lab, sorry, which uses an optical probe to measure the depth and image the depths of those grooves which then we can play back with an algorithm. So it's been quite useful. ^M00:44:36 [ Pause ] ^M00:44:40 >> You talk about the [inaudible] still cracking. Do you ever see any flaking [inaudible] cracking? >> Eric Monroe: So the question was if we ever see any flaking kind of cracking rather than a straight crack? I haven't seen any on any of the fragments that I have. It doesn't mean it doesn't exist. I'll have to reach out to the folks at NAVCC to see if that's one of their failure mechanisms that they do see. >> You looked at the [inaudible]? >> Eric Monroe: We do on occasion. Mostly those are, the only time I've ever seen that is when there's a much larger crack and the pieces come off. I guess by, you know, abrasion in and out of a box you might see something like that. But as far as a failure mode itself I haven't necessarily seen it. I wouldn't expect it as the material itself should be amorphous so it shouldn't have any interior stress points that would just flake off a little particle unless that stress point kind of loops part way through and then comes back up. In which case it's possible, I just haven't seen it. >> I was wondering if you're accelerating and heat testing [inaudible] for how many of those thermal shocks it may take to cause some of this damage? >> Eric Monroe: The answer is some and probably quite bit. We needless to say haven't, I didn't inflict the port intern to doing 100 Fahrenheit to, you know, what we see over a normal cycle of going form say 100 down to 60 or 80 to 60 if it was in the sun and in the shade in a room. The fact that it does happen is promising. I mean, I'm not saying that this is the definitive reason why everything is happening. It's very plausible in this point in time. We do, I did have, do have some plans of kind of reducing that to see if I still get it. That can give us a better idea. At this point for preservation it's try and minimize it as much as you can just like anything else is try and minimize the risk. >> Very interesting presentation. I'm wondering, I'm sure you haven't done a massive survey, all media throughout the world. But from what you've heard of, from what you're familiar with would you be able to perhaps make a guess as to which other types of media might have similar issues? >> Eric Monroe: So the question is if I can think of other media which have a similar issue to the cylinders? I mean, there's kind of two failure modes that we see a lot. And those are more with the modern materials where there's delimitation of various layers. This gets around that by just being one material in a whole. That's some of the problems with the other cylinders. I'm trying to think. >> Fenella France: Unless you took the [inaudible] already looking at all the different types of audio visual materials that are different in the way they're played and the background and so we're looking at those [inaudible]. This is very specific. >> Eric Monroe: Yeah there was outside of the cylinders once they ended up with the wax cylinders then they went to plastics from there effectively. So everything went to completely different style of polymer. So I'm trying, yeah I don't really think of anything even used these materials once they move to the four minute waxes. ^M00:48:17 [ Pause ] ^M00:48:21 >> So if you haven't seen any evidence of chemical degradation, it all seems to be thermally induced. Do you think there's maybe a case to be made for not storing them at 50 and storing them or 50 Fahrenheit and storing them at 70 [inaudible]? >> Eric Monroe: Sure. So the question was since we're seeing that thermal is the likely problem, does it makes sense to not store them at 50? Part of the reason to store them cold if to prevent any other kind of degradation that might happen. There's some evidence that there can be some transfer from encasement or the boxes or something like that onto the surface. So trying to, you know, even if the cylinder's not degrading the box it came in might. And so I've got a summer student that starts next week who's going to be looking at what kind, what are the other failure modes that are on the surface of the cylinders as well as coming up with better ways of doing the cleaning. Now if you don't have access to a cold storage room keep them where they are. The cylinders will spend most of their life at room temperature rather than cold storage. But down there I think that's about all we've got because it's just underneath, it's in the mountain so it's like an almost cellar. >> That was my question. Yeah and I think that it comes with so many materials it's the fluctuation, you know, [inaudible] that keeps heating up those potential failure modes [inaudible] materials. >> Just a [inaudible]. It would be interesting though to say for an institution that was maybe able to house the material, the, put them in a temporary [inaudible] and stable and you can't [inaudible] climate control the areas that could be a new, new room for best practices as far as [inaudible] preservation for housing. And you keep recommending that you keep them at a higher temperature to avoid the [inaudible]. >> Eric Monroe: Right. So the statement was along the lines of if they can't [inaudible] place with these cylinders can't afford it, doesn't have the access to colder storage then, you know, rehouse them and then keep to into another space. And in that case it's yeah, you want to be as climate controlled as possible. Whatever it is just keep it as stable as possible. And thankfully most of the storage spaces are now designed to be rock solid for that, for special collections and things like that. >> Fenella France: And I think what's really interesting is that to be able to say you don't need to worry about humidity for these materials [inaudible] is huge. The other thing would [inaudible] look at is the cause quite often a small organization will transfer them to another location to be transferred. One of the implications into the transportation of trying to control and not we, you know, actually doing more damage. Though I think that's another question I'd like to talk with [inaudible] colleagues about and, you know, probably think about that and keep [inaudible] recommendations for [inaudible]. Do I see a hand in the? >> I was just curious. I understand it's about storage that you've been doing. What is the process for putting the broken pieces back together? >> Eric Monroe: Right now there's not, so the question was what process do we have to put the broken pieces back together. And right now unfortunately there's not a lot of options if it's really, really broken to a bunch of pieces. With Irene we've been able to stitch together, kind of if it's broken off a U chunk or a big piece we can piece that back together again. Image that and then fix everything basically using Photoshop for cylinders, Photoshop for audio. And so if there's large pieces we can do that. And I think the plan is with the [inaudible] iterations of the software is to be able to basically image individual pieces and then stitch them all back together in the computer. So as long as we had all of the pieces we could put them back together again. If you have a little bit of a loss it's much more difficult. So these things are spinning on the mandrel just short of 200 RPM. So even a small loss or small long scratch or crack can lead to quite a bit of data loss because it's now every half second you're seeing a click. >> Fascinating your presentation [inaudible]. >> Eric Monroe: Thanks. >> I was interested in the long [inaudible]. What are your thoughts on the [inaudible]. There could be two or three or perhaps four different kinds of [inaudible]. I wonder whether there was a kind of rudimentary [inaudible]. >> Eric Monroe: Okay, so the question was in the lab notebooks the apparent notching of the pages. They're not notches at all. Those are little weights that were used at Rutgers to hold it, the paper, the pages flat for imaging. And so they're not really notches, not labeling or anything like that. It was just to hold it flat. It's the reason why it looked like he wrote around it is because they put the weight so it didn't cover up anything for us to see. >> The entry [inaudible] notches though it looked like birds. >> Eric Monroe: I think it's felt. >> Thank you. >> I just love that we're all [inaudible] more curious about the notebooks too. And I'm wondering if reading those notebooks has changed the way you keep your notes? And the work you've been doing, do you have good detail? >> Eric Monroe: I do keep pretty reasonable with detail. My handwriting is completely awful. So mine's all on my or most of mine's on my computer. But yeah, that's one of the things that, you know, as scientists were taught to do is lab notebooks are important. And it turns out that lab notebooks are really important even for mundane things that, you know, you might not think that someone's going to come looking at your lab notebook 130 some odd years after you did the experiment but it does happen. >> Fenella France: And basically we've got a big approach it would be trying to extract the metadata from the instruments to sort of try and make it easier. Because the last thing you want to [inaudible] scientist with doing really cool stuff like this is make sure you've written all of it down. But we all know that we [inaudible] little things and so if there's a way to really capture that. And it's something that we're actually working through from some of the digital perspective to try and capture [inaudible]. Well if there's no more questions I want to thank you all for coming. Thank you to the online [inaudible]. So the talks will be [inaudible] probably sometime in June. So watch the online and watch the space for that. And we look forward to seeing you there. Thank you again to Eric. >> Eric Monroe: Thanks. >> This has been a presentation of the Library of Congress. Visit us at loc.gov.