 I'm going to introduce to you our host, Dr. Susan Barger from the FAAIC. Go ahead. Hi, everyone. We're going to have a very interesting webinar today, and I'm going to get started. I'm not going to talk much. There is a handout that is in the lower right-hand corner of the screen. Be sure to download that. It has lots of information. It has all the websites that are going to be talked about, and other information. And it has also all of the previous Connecting to Collections webinars on digital production. So let's get started here. The very best way to keep informed about what's going on with Connecting to Collections care is to join the TTCC Announce List. It's just for announcements, only two or three messages a month at most. This is the web address that you need to go to to sign in. And if you were on the list there, and somehow you think you're not receiving messages, you may have been bumped off. So go here and sign up again or if you changed your email. So please join that. You can also like us on Facebook and follow us on Twitter. And if you want questions answered at any time, you can go to our discussion forum, and there are always people there watching and to make sure the questions get answered. And then you can always contact me. This is my email address. I'm happy to answer any questions or problems. And next week we're going to have a webinar on NAGPRA. And for those of you who have been waiting for the ivory webinar, it's been put off until early next year because of some things that couldn't be avoided, but we're going to have one on storage environments on December 5th. And that will be the last webinar for this year. And we're working on the 2018 schedule. So stay tuned for that. And without further ado, I'm going to hand this over to Brittany Stratton and Wendy Martin. And if you have questions, put them in the questions, comments box. I'll grab them and I'll make sure that they get answered at the end of the presentation. OK. Bye-bye. Go ahead. OK. Hi, everybody. This is Wendy Martin. I'll hear me OK. Yes. OK. Then I will go ahead and jump in. Well, hi. Thanks for joining us today for this Connecting to Collections Care presentation. As I said, my name is Wendy Martin, and I'm the assistant director of stewardship for the University of Texas Libraries. The stewardship department at the UT Libraries encompasses many functions, including traditional preservation activities, but also includes digitization and digital preservation, which we group together as digital stewardship. Joining me today is Brittany Stratton, our librarian for preservation and digital imaging. She supervises the day-to-day work of book and paper scanning in our digitization unit, and she has lots of hands-on experience with digitization projects. I also want to note that another colleague, Ana Lamphere, our head of digitization services, is listed here as a presenter and her contact information is listed in our handout. Unfortunately, she couldn't be here today, but I wanted to give credit where it's due. She and Brittany really developed this webinar, and I looked in to cover her role. So together, Brittany and I will describe the main elements of our digitization operation, including the setup of scanning hardware and processing software, quality standards and best practices as defined by the Federal Agency's Digital Guidelines Initiative, and the scanning process itself. We'll wrap up with an overview of how we steward our digitized assets for the long term, including some details about collecting and storing metadata. We'll contextualize our talk using real digitization projects from our lab, but we'll also offer alternative solutions that may be more applicable in smaller institutions. So before we dive in, I'll give some quick context about the University of Texas Libraries. The University of Texas at Austin is among the largest research universities in the country, and we have over 10 million items in our collections. We have materials representing the full spectrum of physical formats, including bound volumes, both general and special collections, large architectural drawings, maps, manuscripts, archival papers, a variety of audio-visual materials, and an ever-increasing stockpile of digitized and born digital assets stored on physical media. Our collection is housed in our main library and 10 branch libraries across the campus. Additionally, a significant amount of materials are stored in an off-site, high-density storage facility about eight miles north of our main campus. We have been digitizing materials for many, many years, but we continue to refine our workflows and processes all the time. Our digitization efforts help ensure greater ease of access to select physical items, especially for those items that are too fragile or rare for repeated handling or circulation. We typically treat the output of digitization projects as unique objects unto themselves, meaning whenever possible, we employ digital preservation principles to digitized assets. I'll go into much more detail about that later in the presentation. So our in-house digitization unit is equipped with two large-format scanners, useful for maps and architectural drawings, a K-Bus automated page turning scanner with a cradle that's used for bound items, a couple of standard flatbed scanners, including one that's outfitted to handle film and transparencies, and one multi-feed flatbed scanner. We also have capacity for AV digitization at a limited scale. Our lab employs up to six student technicians and two full-time digitization staff members, one of which is my co-presenter here, Brittany. For this presentation, we are discussing the digitization and long-term preservation activities specific to two digital projects, one recently completed and one ongoing in the digital stewardship unit of the UT libraries. The first project involves bound materials published by the university, and the second involves unbound scrapbooks from a dance collection housed at our Fine Arts Library. The nuts and bolts of how we prioritize digitization projects is a subject unto itself, one that is ever involving, in fact. But to summarize our strategy, we at the UT libraries tackle digital projects as a group effort, involving the subject expertise and big-picture thinking of our subject specialists, cooperation from our IT staff who help with behind-the-scenes system infrastructure, as well as the access portals that we use, the description and discoverability work of our content management team, and, of course, support from those of us in digital stewardship who not only manage the work of digitization, but also stewardship of the digital content for the long-term. In our digitization projects, we're lucky to have the support and buy-in of libraries' colleagues as we tackle what would otherwise be an insurmountable task, because the reality is we will never be able to, nor would we really want to, digitize everything. For these two projects in particular, it took all of us not only to identify the materials with research significance, but to coordinate the project all the way through to access. Next, I'll give a few more details about the physical materials from each collection and how we ultimately planned and implemented their digitization. So the first of our use cases involves our role in scanning what we refer to as the UT publications. These are publications produced by the Office of the Registrar here at the University between 1886, which was very early in the history of the University, and 1992. These items represent many decades of the scholarly work of the University in all subject areas. Before we embarked on this digital project, the publications were very difficult for researchers to identify and access, despite the wealth of information that they contained. In many cases, these are small pamphlet-sized works, and they are frequently bound together in a commercial binding. Unfortunately, there is little consistency in the binding approach. Sometimes they were bound by date, sometimes by subject. The catalog record was often not detailed enough to lead to discovery of individual titles within a bound unit. So two librarians, now retired, worked with the Office of the Registrar to gain access to individual copies of the pamphlets for scanning. The fact that we didn't have to scan bound volumes or try to disbind those volumes, many of which were fairly brittle, saved us a lot of time, effort, and potential damage. To date, we've scanned about 1,200 items in this collection, and the resulting images have been made available through our institutional repository, called Texas Scholar Works. Our second use case example includes materials scanned from the Shereer and the Shereer Bustamonte Dance Collection. We're still in the midst of this digital project, and we're currently scanning unbound scrapbooks from the collection, which include press clippings, photographs, small pamphlets, letters, awards, and advertisements. The scrapbooks themselves are part of the personal archives of UT Theater and Dance Professor Yaakov Shereer. These unbound scrapbooks, along with at least 20 linear feet of other paper-based materials, are being digitized to supplement the already digitized analog video component of the archive, which includes about 300 performances, rehearsals, and footage of actual UT Austin classes taught by Shereer and his colleagues. Both the A, B, and paper materials were donated to the Fine Arts Library in 2015 by Dr. Shereer. And as with the UT publications, we are making additional output from this project, available through Texas Scholar Works, our institutional repository. Digitizing these collections will help make them available in the short term, but to ensure enduring access will need to be good stewards of the digital content for the long term as well. And planning for the digital preservation begins as soon as the materials are identified for scanning. Part of our strategy for digital preservation is to collect some descriptive information, such as title and owning location, and the intended use of the digital output directly from the requesting digital project manager, archivist, or subject specialist. They typically have greater subject knowledge of the materials than digitization staff do, and this saves research and data entry time on our end, and most importantly, ensures accuracy and consistency in the data. Before our scanning technicians handle the materials, we assess each item to make sure it can be safely scanned without causing damage. To determine this, a staff member from our preservation unit physically examines each item and takes condition notes. For example, he'll note how far a binding can be opened without causing damage, if the paper is brittle, if there are tears or loose pages, or if there are foldouts or other anomalies to watch out for in scanning. These notes, along with informational metadata provided by the sender, are entered into our workflow tracking system, which happens to be Microsoft SharePoint. Here on the screen, he can see an example of one of our SharePoint records for a Shureer scrapbook. So it isn't crucial to keep this assessment information in the same platform as our workflow tracking information, but for us, assessment really is a critical first step as we begin any digitization effort. So it makes sense to keep all this information together. Knowing the condition of an item helps inform how it should be handled and how it should be scanned. So from our perspective, condition assessment is very much part of the digitization process. So from here, I'm going to hand it over to Brittany, who's going to go into greater detail about standard specifications and the scanning process used for each of these use cases. Thanks, Wendy. And hello, everybody. So our goal in digital stewardship is to reach Fajji four-star level of standardized output. However, this goal is not always attainable or even desirable for everyone. I should remind you that these are guidelines only and the needs of the collection, resources available, or the goal of the digitization project will help determine which star rating is best or achievable for you. Institutions with smaller file storage budgets or digitization projects focusing only on access output may warrant lower resolution and bit depth to decrease file size. Fajji standards are specific to the material format. In this slide is an example of a Fajji standard for oversized materials. But standards for photographic materials, newspapers, bound materials, or rare manuscripts are slightly different. Our advice is to familiarize yourself with the Fajji standard for the materials you are working with before you begin scanning, and then use some of this information as a baseline to suit your needs. I understand Fajji and digitization terms can be a bit overwhelming. I'll also be going over a few basic concepts that should give you a basic understanding of some of the terms so that you'll be better prepared to select standards, equipment, and imaging software that will suit your needs. If you'd like more information about anything described here, the Fajji website has a searchable glossary which has been very helpful to us in trying to figure out what all the jargon means since many of the terms go by other terms. So first off, file formats is a standard way that information is encoded or organized for storage in a computer file and is indicated by the file extension, such as .tif or .jpeg. In our workflow, we typically create at least two encodings for each scan, an archival master file and some type of derivative or access file. The archival file format we select relies heavily on what Fajji and the Library of Congress determined to be sustainable digital formats. Access derivatives can be less stable, but are often easier to access. For master files, we choose a file format that is uncompressed, such as .tif. We want to make sure we get as much information as possible in our archival masters. And we prefer .tif because it is uncompressed, like I just said, but it's also viewable in many image viewers and processing software applications, whereas raw file formats, which is also uncompressed, often requires re-encoding or conversion to a different file format before it is viewable and editable. Not all digitization hardware workstations allow for .tif encoding at capture. DSLR cameras, for example, capture in proprietary file format, which will most likely be a raw file format. These files can be re-encoded to be a more useful .tif later on. For the derivative files, we create both JPEG and PDF files. JPEG is a relatively sustainable, compressed file format, meaning it yields much smaller file sizes than .tif, but they can still be of high quality. PDF is a highly compressed image file but allows you to combine multiple images into a single file, which has advantages when you're looking to provide access to file materials. Keep in mind that the different file formats organize data for specific uses, so it helps to research your access platform and the types of file formats that work best in that environment before you select a derivative file format. There's more information about file formats in our handout. Resolution is measured in PPI. PPI is pixels per inch. Often, DPI dots per inch is used interchangeably with PPI. The more pixels per inch, the more information you will get, and this is going to increase your file size. When you're looking for equipment for digitization, you want to make sure you look at the optical resolution in the technical specs of the scanner. This is the true resolution the device can achieve. Some devices will claim to be able to get very high resolution, but only with software-making adjustments. This is called interpolated resolution. You don't want the software to be making decisions and adding and subtracting pixels. You want to have the control over your images. Determine what resolution you want to scan at by looking at Fajji for their suggestions on what ideally you should scan at. Take into account how large your files will be, and then decide what resolution you need. Once you know this, only acquire equipment that can get you the resolution you desire without software adjustment. Okay, bit depth, which is also referred to as color depth. It's basically the amount of shades or values of a color that can be represented in a single pixel. The more bits or shades that are packed into a pixel, the more color range you will have. This will affect file size. This term can be confusing, since it's sometimes used to represent bits per pixel, and at other times, the total number of bits used multiplied by the number of total channels. So yeah, I know that's confusing. But for example, a typical color image using eight bits, a typical color image using eight bits per channel is often referred to as 24-bit RGB, or 8-bit RGB. This makes sense because 8 plus 8 plus 8 equals 24. When looking at a scanner's technical specs, make sure you understand how they are referring to bit depth capabilities. Color scanners and digital cameras typically produce 24-bit depth, or 36-bit depth. In 36-bit depth, there are 12 bits per color in each pixel. On the side, you can see examples of the different bit depth makes. Here, we have a typical 24-bit RGB image, and a lower quality 8-bit RGB image. So there's eight shades of each color in this one, while there's only two shades of color in this one. In the 24-bit depth image, we can have around 16 million possible color values. You can really see the difference in image quality that you're going to get there. The effect of bit depth is even more obvious in the gray-scale images. The image on the left just has one value of black and one value of white. This is called monochrome, or one bit. And the image on the right, an 8-bit gray, has eight values of both black and white, creating 256 possible gray values, which is allowing us to see much more of the detail in the photograph. So here in this monochrome background is just completely black, and her smock there is you're not getting any of the detail, because you're asking the computer, you're asking the scanner to just pick either black or white. But over here, we get way more detail. Color space is an in-depth and often very confusing topic. It determines the range of colors that can be processed and displayed for an image. Most of the time, devices are set by default to sRGB. This is the most compatible, but not necessarily recommended due to its relatively small color area. The image in the side illustrates the difference in sizes in the different color spaces that are typically referred to. As you see, sRGB, color space does not even include all the colors in a 2200 matte paper printer ink. Profoto RGB contains more color than are included in the visible color space. So Adobe RGB, while still lacking in range of colors, is a good middle of the road choice. So if possible, use Adobe RGB. However, if sRGB is the only option, this is better than not assigning a color space at all. If your scanner allows you to calibrate, you should do so and create a profile that is unique to your device. Every scanner, even if it is the same, make and model have deficiencies. Once you have selected the color space, you should also attach an ICC color profile. The ICC color profile basically allows the color values you create with your device and that are collected there to be truthfully rendered on other devices or programs. The main takeaway here is always, if possible, calibrate your scanner to set your scanner to a color space and to attach an ICC color profile. And this is usually just checking a box in the scanning software and I'll show you that a little bit later on. If you do these three things, you should have better results in the long run. So now I'm going to go into some of our work clothes for our test cases. So now that we have a basic understanding of what determines image quality from a technical standpoint, we can move on to describe some of our actual digitization workflows. For our UT Publications project, we used a Kira Toskabus Bow material scanner. This scanner is basically a book cradle with an overhead camera plus some specialized functions such as a page turning vacuum arm that automatically turns the page and clamps that help keep pages flat as the camera takes each shot. This scanner is ideal for bow materials with high page counts. In this image, you can see our Kira Toskabus machine and workstation, including two monitors, one to select and adjust software settings and the second used to quality assure the images as we scan. We can scan around 1,700 pages an hour when using this automatic page turner. Items that have more fragile bindings or pages can still work well on this machine. We can modify it by turning off all of the specialized functions, but it's still going to be in a nice supported cradle and we can still get really clear images of those pages. I understand that this Kira Toskabus machine may be impractical for smaller institutions due to its high price and specialized functionality, but the basic workflow is analogous to using a camera with copy stand or tripod. And we will go into details specific to true camera and copy stand workflows, but I'll describe our Kira Toskabus workflow because most of the settings and processes details apply to either hardware setup. So first, whenever you're going to scan with a camera, including our Kira Toskabus, we need to make sure we select the correct camera settings. Our Kira Toskabus uses a DSLR Canon 5D Mark II camera, which only allows for an out a proprietary file format output that is called a CR2. Because CR2s aren't in line with FODGY standards, we select the highest quality image possible. With this camera, it is called Raw Large. We recommend choosing the highest image quality, preferably Raw, for any camera that doesn't allow you to select 400 PPI TIF at the point of capture. This high quality ensures our ability to reformat the proprietary file as a 400 PPI TIF in post capture. We are also scanning in Adobe RGB, as you can see here, for our color space. And we do have a calibrated color profile. An ICC profile is attached. In this scanning software, it's called Picture Style. You're going to see that a lot with scanning software. They are going to change the names around on you a lot. And so it's just going to take time to get to know your scanner and scanner software. We also need to ensure proper focus before we get scanning. To do this, we zoom in to 100%, like we have here. And we scroll all over this page and make sure that the text is crisp. We recommend checking your focus with each new item that you're going to scan. Any time you need to pause a single scan for an extended period of time, like you go away to go to, you go get lunch. When you come back, go ahead and check your focus just to make sure nothing has changed. It's better to over check the focus than not check it enough and then have blurry scans later on. The use of an overhead camera, especially our Kavis machine, is considered a high production environment. So we scan the color checking target only once for each publication. Here we are using an X-Rite color checker classic. This once a batch approach is acceptable as long as settings are consistent through the entire batch during capture and image processing. For a Kavis machine and for most overhead camera setup, you should use your camera software to check the colors or to adjust the profile as needed. We recommend scanning cover to cover for bound materials. Our Kavis machine uses angled mirrors to allow us to capture two angled pages flatly and in order and the Kavis software also displays each page in a larger view in real time. As noted earlier, we rely on our initial condition assessment to decide whether or not to use the automated functions of the Kavis scanner. While the scanning technician checks each image to make sure that no pages are missed, we want to make sure that their hands aren't in the shot and that we really want to make sure that we capture the entire page. We don't want to cut off any content and so that's something we really look for when we are scanning. So this is an alternative process and it might be more suitable for smaller institutions and this is the basic equipment setup needed to digitize using a camera on some sort of a stand. In the resources we have provided some more in-depth instructions and guides from institutions who digitize with the camera. We don't use this method currently but we have helped make setups for people needing to have more transportable ways to digitize. The upside of having a camera is you can use it to digitize anything in your collection. So if you think about all the three dimensional objects you have, this could be a really good way to digitize them as well. So things that you want to look for, first off you're going to want to look for a good quality DSLR camera. DSLR stands for digital single lens reflex. A good quality DSLR camera will be key and to be honest it's going to be a big purchase. It can be a very big purchase but you're going to want to look at the sensor size of the camera, the larger the sensor size, the greater the resolution you're going to get and the less digital noise it will produce. So basically a better image. You will want to get a camera that you can access the manual controls. Again, you want to have control over the image that you're going to capture. And it's really great if it has the grid, the screen on the camera has a grid screen view and this is going to help you make sure that the item you're imaging is straight which is going to make a big difference in post-processing. A remote shutter switch release, it doesn't seem like that big of a thing but if you don't have to actually touch the camera when you're taking the picture it's going to cut down on any chance of movement that could create blurriness. For the tripod you might want to consider getting an extension arm. Copy stands can be used in place of a tripod but they are limited by the area that they come with and also the height. They can also be heavy so if you want a more portable and versatile option a tripod may be a better choice. The copy platform is what you will put your object on to be digitized. You're going to want something clean and plain, a neutral background that's going to allow your object to be the main focus of the image. Lighting is really important to getting an image that is true to the original object. Controlling the light will be necessary to create consistency from one day of scanning to another. Two lights like you see in the image here at 45 degree angles is a typical setup. So I know these are really basic instructions but you should really figure out what specifications you want to achieve and then go to your local photography store and become really good friends with them and they're going to have a lot more information on the specific lighting and other equipment that are really going to suit your needs. So getting a camera is sometimes scary and really overwhelming and it's going to require you to get comfortable but this is doable with research and time so I encourage you to give it a try if you can. So for our second case study, the scrapbooks, they were originally bound in three ring binders and we chose to disbind and remove the items from plastic sleeves and scan each page individually. Some of the scrapbook pages were laminated and there was no way to delaminate them without destroying the object. So we scanned these intact. We chose to scan both laminated and non-laminated paper materials on our Epson Expression 1100 XL, I'm sorry, 11000 XL flatbed scanner. A flatbed is a good option for objects that are glossy. They are also useful for scanning photographs, print material, and smaller bound pamphlets which were all in these scrapbooks. I should note that the photographs that we scanned were all in excellent condition. If a photograph is in very bad shape, it might not be a good idea to put the emulsion side down on the glass scanning bed. You might want to consider a camera stand set up in this case. Photographs ideally need to be scanned at a higher resolution of 600 ppi. This scanner is capable of achieving the 600 ppi 24-bit depth color and TIFFs, and we are at least able to attach a color space to the images. The Epson Scanning software makes it easy to switch from the 600 ppi for the photographs to 400 ppi required for the print material. For most scanning devices, including our Epson flatbed scanner, the device scanner must be powered on to adjust software settings and configuration. This has gotten me so many times, I think, something is broken and I just need to turn the machine on to access the software. And most scanners do come with their own scanning software package, offering basic color profiles and image tools. Although you may choose to purchase more robust scanning software such as Silverfast. In the example images here, we are scanning a printed item. Therefore, we scan at 24-bit color at 400 ppi resolution, and also we're going to do a TIFF in accordance with the FODGY standards. Selecting the source and targets over here establishes our ICC profile in the color tab of the scanning software. In some scanning software packages such as in our example, you may encounter ICM, which stands for Image Color Management. The standard file extension for ICC profiles on Windows is ICM, which is a function of Microsoft proprietary software. The file format is the same one ending in ICC, and they are completely interchangeable. You should not have any difficulty using either type of profile in any ICC or application. We also need to attach a color profile and have it have selected this Epson standard, which is specific to the scanner. This will ensure that the scanner software is set to capture in a color profile that matches the scanner's hardware capabilities. However, this color profile will not necessarily transfer well to other devices or editing software. So we're going to select a target color profile of Adobe RGB. This means most displays and post-processing software packages should be able to interpret the output faithfully. In general, in our scanning unit, we prefer not to introduce any pre-processing to our archival masters, but we have chosen to do an unsharp mask with a level of medium in our example here. After a few test scans, when we looked at the images, we ultimately decided that this level of unsharp mask, which is a sharpening, gave us the crisp output that most accurately interpreted the quality of our original items. Okay, a word about targets. There are many different kinds of targets for different purposes. Some targets are for calibrating equipment, and some are to be scanned with the object. They are not necessarily the same thing. Object scan targets are often called a color bar. In keeping with Fawjee standards, we include a color bar within each unprocessed archival master image. These targets measure the digitizing environment of the actual material and not the equipment. This target is important because it allows us to be able to communicate the physical size of the object when you're actually looking at the image. You can't always tell how big an actual image is, and so having that little ruler there is really helpful. We typically use the Kodak Q13 for our flatbed scanning. We position the scan target near the upper left-hand corner of the scan bed, right here. Then we position the object a little bit lower, here, with some space in between. We do this because we also leave a bit of room between the edge of the scan bed and the object itself, as you can see right here. We do this because with this particular scanner, we've noticed a slight curve in the images if the object is too close to the edge. This may not be an issue for other scanners of this same make and model, but it's something we've noticed with ours. Once you have your object and scan target placed on the scan bed, you should preview scan. Previewing before scanning allows you to draw a marquee around the area you want to scan instead of scanning the entire scan bed. This cuts down on scan time and ultimately file size. Getting the object as straight as possible and in the correct orientation will cut down on post-capture processing. You can also make sure you like how the image looks. You can check the color checker, you can check the focus, and then you can scan. If all of your items are the same size, you can pre-scan just the first item and position all subsequent items in the same position on the scanning bed. This will cut down on your overall scanning time. Up until now, we have discussed strategies for establishing a capture environment, resulting in the highest quality output master files. However, uncompressed tests that we've been scanning in may not be good for general use because they are quite large and may include target information and uncropped border areas that distract from the usefulness of the content areas. Your digitization workflow will most likely need to include some level of post-processing. Post-processing tasks may include image manipulation such as cropping out that target, attaching an ICC profile if your scanning device didn't allow for it at the time of scanning, and in rare cases where the original materials were badly yellowed or discolored, you may need to perform some level of color or contrast balancing, all of which will make display of the scanned images easier and more consistent. An ideal capture where the object is straight and your color and light are correct will make post-processing more efficient. Be aware, though, that you should generally not execute any type of post-processing on the master file image itself. Rather, you should create a copy or a derivative file before executing post-processing tasks and leave the master file untouched. This strategy will ensure that the archival master can be preserved so you can go back to the original scan in the future. Most post-processing tasks require the use of specialized software. We use Photoshop for image processing, resizing, and some types of compression, and we use Adobe Acrobat Pro to run OCR and create PDFs. Lightroom is another popular image processing software application. However, Photoshop, Lightroom, and Adobe Acrobat Pro are proprietary and expensive. For those with limited budgets or who prefer open source software, we suggest XNVIL, which is free to download and includes instruction resources. It's capable of executing most image manipulation, post-processing, and resizing tasks in a user-friendly interface. And for more information about XNVIL, you can see our handout for that. So back to our case studies. For our UT publications, and for all of our post-processing and derivative requirements, all that is figured out beforehand with the client in the planning stages of the project. For our UT publications project, one of the end goals was to allow access to these publications through our online digital repository, Texas ScholarWorks. As a small, easily downloadable PDF file, this involves several post-processing steps. If you remember from the capture process, our DSLR Canon camera settings saved the images as proprietary CR2 files. CR2 is not FODGIC compliant as an Archival Master file, so before we begin any image processing, we will re-encode or convert the raw CR2 files to an uncompressed TIFF 400 PPI file and save those. We either use camera raw software or Photoshop to re-encode the CR2s as TIFFs. Once we have our Archival Master TIFFs, we create our publication master images by cropping each image to the page edge and saving as compressed JPEG files. We compress our JPEGs to a quality of 9. As a final step, we combine the individual JPEG files for each page of the publication into a single PDF file using Adobe Acrobat Pro. We also run OCR or Object Character Recognition to make the publication searchable and easier to navigate. This PDF file becomes the access-derivative file which is uploaded into Texas ScholarWorks for access and use. The post-processing workflow for our Shereeer scrapbooks is very similar to the UT publications, except that during capture, the flatbed scanner creates uncompressed TIFF files. The publication master files for this collection are created in Photoshop by rotating and cropping as necessary and again saving as JPEG. We then combine all the JPEGs into a single PDF and run OCR on the text portions. The PDF is then delivered to the subject specialist who uploads the file into Texas ScholarWorks and adds descriptive metadata for the entire PDF. So far I have described the workflows and the standards we use for creating digital assets. Next, Wendy will discuss our next steps in our digital preservation workflow. Okay, back in my hands. Can everybody hear me? Okay, so I'm going to go off script for a moment and say that I can see there's lots and lots of questions being asked. So I'm going to zip through the second part about digital preservation with the hopes that we can address all the questions. At the end, I've tried to answer one or two that I could do quickly but kind of got lost and didn't want to take up time doing that while we need to talk. But any questions that we don't get to, we'll make sure that we get answers to them to you guys somehow. So with that, I will jump into the digital preservation part of this. Okay, so it's clear that we take a great deal of care in creating our digital assets and we want to take equally great care in storing them for the long term. We are very fortunate to have a digital archivist in our staff who has helped us refine our strategy for digital preservation over the last few years. Our strategy is of course unique to our own circumstances but in talking through it, I think we'll hit on some key points that are useful in any digitization setup. So in thinking about digital preservation, we want to focus on what makes the products of our digitization efforts usable into the far future. This guides the high level planning around which pieces of a project's output are critical to the understanding or making use of the whole and what sorts of accompanying information need to be preserved along with the images in order for people in the future to be able to use them. Essentially, we want to be asking the question, if someone wants to use this data and these images in the future and none of us are here any longer to explain what the project was or describe the content, what do we need to do or provide now in order for them to make sense of it on their own later? With this question in mind, I'll spend the next several slides talking you through our strategy for digital preservation which includes the five points listed on this slide. I'll discuss the reasons we do what we do and describe the tools and methods we use to accomplish this work. So at this point, I'll just note that currently we have a piecemeal approach to our digital preservation work. There are digital preservation systems or software packages and some are even open source that you can use that will manage and automate all of these functions in a systematic way. To date, our processes and procedures simply haven't evolved that way, so I'm going to take you through our steps step by step, but just know that that's an option that's out there. Okay, so Brittany has already described how in creating our digital assets, we choose well documented common file formats that meet national and international standards. We try to use open formats rather than proprietary ones when we can, and when that's not practical, we'll make sure what we are using is ubiquitous and well documented, PDF for example. This is a critical step in thinking about the long term longevity of a file. As a next step, we also make sure that we gather technical metadata about the file formats used in a given project. We package this information together with the digital assets for long term storage. So what exactly is technical metadata and why do we want it? Well, technical metadata describes the technical characteristics of a file. In most cases, this information is either in a header in the file or embedded within the file itself. We choose to extract that information and store it separately. Why do we do this? Well, down the road, a future user may encounter the image file and not be able to open it. Having the technical information about the file pulled out separately in a human readable format may allow them to figure out a way to access the data in the file or to find legacy or contemporary tools that will allow them to open it and use it. How do we go about getting this technical metadata? Well, collecting it is relatively straightforward as long as you have the right tools at hand. We use the free and open source file information tool set, or FITS, developed by Harvard. FITS is actually a set of tools bundled together that do a lot of things. They identify, validate, and extract technical metadata for a wide range of file formats. The slide here shows the various tools at use in FITS. And here's an example of the type of information output by FITS. Even though we aren't looking at the actual image, we can tell all kinds of things about it from this metadata. For example, we can see that this file format is uncompressed. We can see the width and the height of the image, the color space, the ICC profile version used, as well as the bit depth. All of the scanning specifications that Brittany just described are captured here. There's even information about the type of scanner and the software used to create the image. Because FITS outputs is XML, like you see here, the information is both machine and human readable. So this is just scratching the surface. There's definitely much more to say about FITS and other metadata extraction tools and how they could be used to support digital preservation. So I invite you to take a look at the links in our resource list if you're interested in learning more. Okay, as you've learned through this talk, we produce a variety of output and information as materials move through our workflow. We have the image files, master copies, as well as access derivatives. The technical metadata I just described, as well as the descriptive information that was gathered at the outset of the scanning project. I'll spend the next few slides describing how we package that information for delivery and for long-term storage. Our packaging strategy is informed by the Open Archival Information System Reference Model, or OAIS, shown here in the diagram on the slide. OAIS is a high-level model that describes the components and processes necessary for a digital archive. It outlines requirements and responsibilities of the archive, models its functions, and defines an information model. It's a conceptual framework and not a technical or methodological prescription. From the diagram here, you can see that the model involves information producers submitting data into the preservation system or repository, where the data is then managed and preserved, and then can be disseminated to information consumers on the other end of the system as needed. Lots of activities happen within the system, including planning, administration, ingest, and the provision of access. As the digital stewardship unit, we are doing the management of the work that takes place in the gray box here. So this is information coming into the system, information inside the system that is being managed and preserved and taken care of, and information coming out of the system to the consumer. So again, we could talk about OAIS for hours, but nobody wants to do that. The basic takeaway for this webinar is that OAIS specifies three different types of information packages illustrated here in this diagram by the... Let me get my arrow going here. By this arrow, this turquoise box here, and this arrow here. And they are the submission information package, the archival information package, and the dissemination information package. So when I say package, I simply mean how we organize and arrange the digital assets and the metadata into directories and folders. In our environment, the submission information package, also called a SIP, which I'll say from now on, is what we think of our basic record-keeping package. We establish this package at the outset and include any metadata sent by subject specialists or archivists about the content. And by the end of the digitization process, we add reference thumbnail images of the content, technical metadata, and more administrative metadata. The package ends up including everything you need to understand the digitized images except the full-size images themselves, and more on that in a moment. We keep these SIP folders locally on readily accessible server space so we can get at them any time we need to. The archival information package is the object of long-term preservation. This is the most robust of the three packages in that it contains all of the data and metadata needed to make sense and reuse of the content in the future, including the archival and publication master tips. We wrap the files in this package up in a digital bag, which I'll describe more fully in a moment, and the process of bagging generates additional files and metadata that are included in the archival information package. In our digital preservation system, these packages are written to LTO tape for long-term storage. I'll just note that information on LTO tape is not readily accessible the way information on active server space is, which is why I previously mentioned that we do keep our SIP packages on active server space, whereas our archival packages go onto tape. Finally, the dissemination information package is the package of information that's delivered to an end user of some kind. In our examples, the dissemination information package is the group of access files delivered to our institutional repository staff for ingest, in this case PDF files of the things we've scanned. Okay. In thinking about these packages, it's important to be consistent in how we structure and organize the data. Right from the beginning of any project, we have a system for organizing project folders and nesting data files. Each piece of media, such as a single scrapbook or an individual pamphlet in our two use cases, is assigned a SIP number as soon as the items are sent to our unit. Our naming convention for the SIP number is the year followed by consecutive accession type number. So in the example you see here, this was the 350th SIP number created in 2017. The SIP number is then used as a directory name into which all files are placed throughout the various stages of processing. The top level directory will contain subfolders for content files, derivative files, the technical metadata for each class of files, as well as any descriptive or administrative information created throughout the project. In the slide you can see the organizational structure for one Shariah scrapbook. The archival master files are in this folder. The technical metadata for those files is here. The publication masters and the access images or access PDFs are here, and the technical metadata for those is here. We've got an additional text file that scraped our information from a SharePoint and a descriptive metadata spreadsheet that was provided by the subject specialist. So that's all packaged together in this consistent way. Before putting our archival package away in long-term storage, we need to take one final step to wrap up our data in a way that ensures that it remains intact during transfer. This is the bagging step that I mentioned a few slides back. Bag It is a hierarchical file packaging format for digital content developed by the Library of Congress, which wraps content data and associated metadata in a digital container, the bag, so that it can be safely transferred to and from a storage platform or across networks. It arranges files in a standard directory structure and generates a check sum, which is a unique digital signature for each file, as well as creating a manifest of the contents of the package. Essentially, a bag is just like a computer folder or directory, but with specific formatting and contents. The manifest that's produced lists each file in the bag along with its check sum. When you transfer bags from place to place, you then regenerate the check sums for the files. If the newly run check sums match what is listed in the original manifest, you know your data transferred without any loss. If they're different, you know something went wrong. The bag info file contains a small amount of administrative metadata, such as contact information for the collection owner and a brief description of the collection. You can kind of think of this like the luggage tag on the bag. So here's a look at our luggage tag. This is an example of our bag info fields and their input guidelines. It's designed to provide enough information for someone to understand what's in the bag, who put it there, how it can be identified and perhaps found elsewhere, and a right statement addressing how it can be used. And again, this is really, we designed this with the future user in mind. We're imagining a scenario when somebody is looking at this and none of us who created this, we've all been long since retired and are on a beach somewhere and they've got to make sense of this on their own. So that's the idea behind the information that's provided here. So you can run bag it from a UNIX command line if you have the relevant technical know-how, but you can also use a more user-friendly graphical user interface utility called bagger to create these bags. Developed and maintained in part by the Library of Congress. It's an open source, well documented, and is available for free from GitHub. Bagger allows you to input the metadata information using these profile forms as shown here that are included in the software. They automatically populate the bag info text file with all the proper formatting just by filling in this form. The form keeps data collection consistent and it's customizable to various institutional contexts and practices. Again, our handout points to more information on bag it and bagger if you're interested in learning more. So once our archival packages are bagged, they are ready for transfer and ingest into long-term storage. I've already noted that we store our archival packages on LTO tape. Our archival master files tend to be very large and we produce terabytes of data each year. Because of that, keeping the master files on active server space or even in cloud storage isn't practical or financially feasible. Instead, we rely on LTO tape as our primary preservation storage method. We often refer to this as vaulting to tape. LTO stands for linear tape open. It's a magnetic tape data storage technology pictured here in the slide. The LTO consortium directs development of the medium and manages licensing and certification of the manufacturers that produce the tape, so we're confident that LTO is a stable archival medium. However, LTO has its drawbacks. It's a tape cartridge that depends on specific equipment for recording and retrieval of data. And it requires specialized expertise. In our situation, we have system administrators from the library's IT unit running our tape system. Also, you can't readily view files and folders and their content on tape the way you can view them on a server or a hard drive. The data has to be retrieved or restored before it can be viewed. Retrieval of data from tape can be slow, especially when it falls outside of our unit. Still, because this is for long-term archival storage and frequent retrieval is not the intention, the constant infrastructure savings associated with archival tape meets our needs. LTO tape is the best choice for us, but it may not be a good fit in your own institution. Depending on the scope of your digitization program, cloud storage or the use of local servers or external hard drives might meet your needs. Whatever your storage solution is for your archival packages, you'll need to keep multiple copies. Those copies should be stored in geographically separate locations in case some kind of disaster happens at one location. Ideally, you would have not only two locations, but also two different platforms, maybe one copy in cloud storage and one on an external hard drive. And unfortunately, you can't stop there. Data degrades over time, so digital preservation cannot be an out-of-sight, out-of-mind endeavor. You'll need a plan to check your data periodically and perhaps migrate it to new physical carriers as needed. That's where the check sums and bags will serve you well. You can check the integrity of the data and migrate as needed over time. So finally, a few words about record keeping around all of these efforts. To this point, we have already delivered access derivatives to satisfy the needs of the immediate digitization project, and we've packaged and stored archival copies of the output for the long term. Now we want to ensure that those stored files are findable into the future. We do this by keeping careful records of all the materials that enter our digitization workflow and their ultimate disposition on LTO tape. I just talked about tape, and I think it's easy to understand that once data is stored there, it could be, if you don't take the right steps, easy to lose track of what's there. As I already noted, it's not easy to pull up your tape and look at what's on it. So this is the reason that we keep a thumbnail reference image in the submission information package, and we keep that package on Active Server Space where we can pull it up and look at it. That submission information package, the SIP, is a very robust record of the digitization work that has been written to tape in the archival package. We can look through the SIP directories and find out everything we want to know about the vaulted images. We then just need to retrieve those images from tape if we actually want access to the archival master file. We also keep a comprehensive SIP submission information package spreadsheet that serves as an inventory of everything we digitize. The fields in the spreadsheet are called out in yellow here. We enter the information as soon as the item enters the digitization workflow and add data along the way, like the tape number, as it moves through the process. This detailed record keeping gives us a way to search for and locate information about our SOAR digital assets without having to read through tape manifests. So that wraps up all of our prepared remarks about this. I know we covered a lot, and some parts are perhaps more relevant to your immediate needs than others. But even if our situation is different than yours, I hope we've provided some useful information on scanning equipment, standards, and procedures, as well as some insight on digital preservation. And so we are happy to take questions if there are any, is what I wrote here. And of course, there's lots of questions. So we will begin working through those. Do you want me to read those? Yeah, can you do that, Susan? Yeah, I will. Heather Potter asked in the very beginning, how are you scanning the unbound scrapbooks by the page or individual items on the page? So we scan them unbound. We have needed to do bound scrapbooks in the past. And for those, we've had to use a larger bound materials cradle. And that might not be available for smaller institutions. But if you had a camera and tripod setup, you could image each page individually with one side propped up to where it wouldn't have to be open all the way. And it would take a little manipulating of the actual objects back and forth, but it would definitely be doable. And I know that people do do bound materials that can't lie flat that way. Okay. Deborah C. Annithica said, did you have a workflow in place or did you have to create one? And then Heather Potter asked along the same lines, can you please provide us with copies of your workflow process? So our basic workflow processes are based on if it's a bound material, if it's a large format material, or if it's a smaller item. And it's really based on the same scanning applies regardless of what the item is. But our access derivatives are going to be different, whatever the client desires or the subject specialist desires. So while yes, we do already have basic workflows in place, we do have to tweak them for each project. Okay. And can you... Your workflow or... It sounds like that's kind of complicated. Very. Are you there? Can people hear me? I can hear you Susan, but only very faintly. I'll try to speak up. Well, that's better for me. Okay. So can you provide copies of your workflow or it sounds like that's fairly complicated? It is. We do have some documentation, but it's very specific to our projects. We could take a look at what we have that could be generalized that we could share out later. Would that work Susan? Yeah, that'd be fine and I'll post it with the recording. Okay. Yeah. William Van Doyle says, could you describe how you're using SharePoint? Sure. So SharePoint, we use it not only to log information about the actual physical object condition and assessment and things like that, but we also have a way to mark its progress throughout its sort of digitization life cycle. We can mark if it is currently being scanned. So it works for us as a tracking method. Some of the subject specialists do have access to SharePoint, so they can see where we are in the process, so they don't need to contact us and ask us where is my stuff. They can just go ahead and look on SharePoint. It's just a way for us to sort of easily disseminate information instead of having to answer constant emails. And I would just add there that we don't necessarily recommend the SharePoint as the way to do this. It's what our organization had available to us. We've been using it for over 10 years now, and we've made it work for us, but I wouldn't want people here to construe that use of it as a suggestion that it's the way to manage digital projects. It is one way to keep track of what you're sending through your workflow. Okay. Mara Warnock asked if you could provide the name of the and type of your book scanner, and you did include that, but do you just want to give that information again? Sure. It's a Cabus II made by the Kirtos Company, and I noted in the questions and comments that we've actually had it for seven years now. So I assume they're still manufacturing that model, but I'm not sure of that. Brittany's nodding yes, so we've had it for a long time, and it's been a good work for us. Okay. Then, I've lost my place. Oh, please answer the evaluation before everybody leaves. I'll ask you to do that. Jessica Okolinsky says, can you see a preview of what the image will look like at different settings, f-stop, ISO, et cetera, or do you just have to take a picture and adjust until you get the results you want? So our camera software does have an ability to do a live view, and we can adjust things in that, and that makes it a lot quicker and more efficient to be able to really get the image quality that we want. And for the most part, our camera doesn't move. It is constantly attached to the machine, so the amount of adjustment we have to do on a day-to-day basis is pretty limited. But, yeah, having a camera that you can access a live view is really going to help to make those sort of changes without having to constantly take a shot and evaluate it. Okay. And it says, it looks like your scrapbooks are more contemporary than our 20th-century scrapbooks. Would you recommend a flatbed for older black-page scrapbooks? I'll take this one. I think it really depends on the condition of the materials in the scrapbook. I think there are certainly examples where using a flatbed scanner would be fine, but you just want to make sure that your photos are well-adhered to the pages. If there's any concern about what damage might happen in the act of flipping a page over to put it face down on the flatbed, you might consider a camera setup instead. Okay. Samantha Thompson asked, what if a page is unavoidably curved? How do you make sure it's all in focus with your camera? So, even when we need to scan pages, not on our KVIS machine that has clamps that hold the page down, we often use, sometimes it's a two-person job. You have to have one person using maybe a microspatch law, or occasionally we use a gloved hand to make sure that the page is flat. And this is acceptable on the Fodgy standards as long as you're not creating too much of a shadow on the object itself when you're taking the image. I would say though that the majority of cameras do come with some sort of depth of field, and so unless it is a very large curve, you're not necessarily going to have a problem with that. You can look at your camera specs and find out what depth of field it has. I know on some of our cameras we have up to an inch, so that's a pretty big difference, and we'll still get a clean shot overall. Travis Souther asks, is there a difference between monitor RGB and Adobe RGB? To be perfectly honest, I am not sure about the answer to that question. We can ask one of our tech guys, see if he has an answer, and we can get back to you. I would assume that there isn't a huge difference, but I'm really not sure. I would have to ask somebody. Kathy Gow asks, doesn't using Unsharp mask lose you some original detail, and images will be sharper than the original? Yeah, so you have to be careful when you're applying any sort of Unsharp mask. For some things it's definitely not advisable. Our images for the scrapbooks, we wanted them to be very sharp and crisp because of the variation between one type of print material to the other type of print material, and when you looked at them in person, they looked very, very sharp, but without Unsharp mask, they looked a little bit fuzzy around the character edges. For this specific project, it was okay. Other things, it might not be advisable. Asita Albert asked, if you're planning something that is a combination of text and images, do you recommend 600 PPI? So if the image is a real photograph, it's just not a print off from a regular computer printer. If it is a legitimate photograph, then yes, you would want to go with the 600 PPI for the entire item. If, like our scrapbooks, you have, oh, here's a photograph, a true photograph, and then the next thing is a print material that isn't a true photograph. You can go ahead and switch back and forth. Okay. I'm going to combine several questions here. There was sort of a discussion on the different software. So Kevin copy asks, XNView is Windows on the line. Are there other open source programs? Heather Potter said, what about EarthMView instead of XNView? And Rodney Carter said, I use GIMP as an open source Photoshop alternative. And so what are other things besides Windows-based programs? And what do you suggest for Mac users? Yeah, so it's going to say I'll answer that and then tackle Colleen's question too, which unfortunately isn't going to help her very much. We are largely a Windows shop, so I don't have any specific recommendations for Mac users here. So I'm sorry about that. We have used EarthMView in the past. I noticed in the comments somebody said that their IT folks have told them not to use it because it creates problems with Active Directory. And I don't know if that's the case. It's been several years since we used it, but we did use to use it with great effectiveness. So there are other utilities out there that you can use. If you just Google open source image processing, I think you'll find other suggestions. I'd be interested if other people who are listening have suggestions for what they might be using, especially if there's Mac users out there that have recommendations. Yeah, and if we get those, I'll post them with the recording. Jessica Opolinsky said, are there any certifications or training courses for learning digitization methods? And I suggested she check with Lyrasys. But are there any others that you can think of? Lyrasys was on my list, too. AnyDCC, I think, has some information on their website. And they offer, I think they're the ones that do digital directions, which is a week-long course that covers all kinds of things. But yeah, any of these regional preservation centers will offer some kind of course on digitization. They're not necessarily certification level, but there's lots of information out there. The Society for American Archivists, SAA, also offers, they do offer a certification course in dealing with digital preservation or born digital archives, where much of these same things are covered, only slightly different context. And we didn't really put anything about courses in our handout, but we could pull together some resources around that. Okay, thanks. Sure. And I'll make sure they get posted. Samantha Thompson, the SIP number is for one digital object or for a bunch of them? For instance, a collection. I missed this. So that's actually a great question that I just sort of glided over. But I'm glad you picked up on it. It depends. So ideally, we'd like the SIP to be describing the most granular thing possible. And that was our initial approach. But we learned pretty quickly that that's not scalable when you take in a collection of say, you know, 400, three and a half inch floppy disks that you're going to deal with. That ends up being a lot of sort of just data entry work in terms of creating the SIP. So we figured out pretty quickly that we needed to step back from that ideal and sort of figure out how we might group things. So Brittany can correct me if I'm wrong, but I think in the case of like the Shereer scrapbooks, A scrapbook, one scrapbook is one SIP. OK. Where does the accession number fit into the digital preservation file? So that number is what we use to name all of the files essentially. I think that answers the question, that it becomes the file name of all of the image files and associated metadata within the packages that we put away for digital preservation. OK. Laura Watson says, we use Pass Perfect. Should we create a separate master files for each item with the accession number? So I will say I'm familiar with Pass Perfect, but have not used it. So I'm not quite sure how it sets things up for you. I think our main point here is that you can do whatever you need to as long as you're consistent. So if it's naming files for you and you want to just use the name it assigns, do that. Just make sure you apply it throughout all of your preservation packages and metadata in a consistent manner. Does that answer the question? I think so. And Laura, let us know if it doesn't. Kate Swisher says, this flatbed scanning method also applied to standard copiers with scanning capabilities, standard office copiers. And then she says, we have a Toshiba copier. And would you have any concerns about using this to scan archival images? So I'll start my answer, but I'll let Brittany jump in if she has additional ideas. I think my concern there would be that you probably don't have ability to control your settings with that kind of scanner. Maybe I'm wrong. I'm not familiar with that model specifically. So that would be my concern, that you're sort of stuck with what you get, and it's probably not going to give you a TIFF. It's going to give you a JPEG, which is compressed and lossless. So that isn't an ideal archival master file. However, you work with what you have, and if that's what's available and you need to provide digitized images for access, that's certainly one way to do it. Yeah, I would agree with Wendy. You just need to look at what the capabilities of the scanner are and determine if those are, like I talked about, optical resolution and just really look at what your technical specs are before determining if they're acceptable for archival images. Okay. I also would ask, curious about migration checks and how practically that gets used. Suggested resources? So, I've mentioned that we obviously have a large team that works through all of this here at the UT libraries. In our case, as we migrate from, what's the word I'm looking for, a version of LTO tape to the next version of LTO tape, our systems administrators are checking all of that. So, everything we have is on LTO4. LTO is backwards compatible by two generations. The world is now on six. It's time for us to be moving, or the new readers won't be able to read our tapes. So they conduct that migration, and they do use those checksums and make sure that everything has transferred properly. That's not very practical advice for, you know, a smaller scale shop. I would recommend that if you are storing things, say on external hard drives or even in cloud storage, that you just set up a sort of a calendar for checking your material that, you know, once a year, this is what you're going to do. You're going to check to make sure that everything's okay. And these steps are really, you know, much better done when using automated tools to help you do it. So, looking into using Baggar and, I mean, Bagit and Baggar and how you run those tools automatically will help you develop workflows for this. Okay. Kathy Gao asks, what are your recommendations for DPI for scanning 35-millimeter colored slides? Okay, that is a great question because I have just spent weeks scanning 35-millimeter colored slides. And what we scan at is at a resolution of 4,000 PPI. So, yeah, it's a lot. Sarah says, do you employ a systematic approach for rechecking on digital files for deterioration over the long term? We are employing check sums, but I'm not sure what the next steps are. So, I think we kind of answered that with the previous question, right? We do have a systematic approach here, but I recognize that it probably doesn't apply to the situation most people are in. But again, I would just recommend setting up a routine schedule on which you check on your digital assets again. So, if you've generated check sums at the outset, then you want to run check sums again and make sure that they match. And there are tools that do this in an automated way for you so that you're not actually looking, you know, down the spreadsheet of these long strings of numbers. And that's what a check sum is. It's just a long string of an alphanumeric string of many characters. And there's all kinds. You know, we could do a whole webinar on check sums. There's different kinds, and they have, you know, different people like different ones for different reasons and so on. But yes, if you just set a periodic frequency that you do that, then you can keep tabs on how your data is doing. Now, what you do when you discover that the check sums don't match and there is a problem is where we hope that you've kept two copies of your data. So, you know that one is compromised in some way. You can go to the other one and hopefully it's still good. And you can recopy your archival images from there. Okay. Samantha Thompson says, sorry for the basic photography question, but what type of lens is best for taking camera pictures of manuscripts? Would that be a special lens for close-up photography? So, I'm going to readily admit that neither Brittany or I are photography experts, either. The camera setups we've worked with, the camera is part of the scanner. And so, we don't have a lot of decision-making in deciding what lens to use. It's what came with it. Brittany, do you have anything to add there? I do. So, again, when you're picking a camera for manuscripts, you want to make sure that you really are able to get a lot of detail, right? You're looking for all the pin strokes and everything like that. So, definitely getting a lens and a camera that can achieve that 400 ppi and maybe even more if you really want to be able to zoom in and get really close-up information from your image. I would suggest going to a photography store and asking a photography specialist. And I want to say to people the recording will be available in the next few days on the slides. And once you no longer see the advertisement for this webinar on our Connecting to Collections Care home page, you can look in the archive and it will be there and you can download everything. So, that just keeps an eye out for that. Excuse me. Camille Diab says, what are the differences when dealing with an image collection with a lot of scientific data attached to it? Do you have any recommendations as far as organizing this information in an accessible way? So, that's a hard question to just answer sort of off the cuff, but I think my recommendation would be to just, before you do anything with it, carefully think through what you want your model to be. I wish our digital archivist was here to answer that question because she probably has some more sort of tangible advice. So, you're dealing with different kinds of files, and I just think you want to consider, well, for example, like we have our image files and then we have technical metadata associated with them. Well, we've decided to keep them into separate directories. So, you want to think through that sort of thing when you're dealing with multiple sets of files. Do you want to group all the files about one thing together or do you want to separate them out into directories? And there's not a right or wrong answer for this. It just depends on what kind of data you're dealing with and how you think people will need to access it and use it now and how it might make sense to people in the future. Okay, we have two more quick questions. One, if you're scanning 35-millimeter slides at 4,000 PPI, are you keeping the target size the same? Yeah, so when you're scanning really anything, you want to make sure that you're scanning to the size of the actual physical object. So while 4,000 PPI seems like a lot of pixels, you're packing them all into whatever the size of your slide is for 35-millimeter. So that's, like, what, one and a half-ish by one? And so, yeah, you're going to get an image when you look at the metadata for that image. You should be seeing a size of one and a half by one or whatever the size of your, whatever the size of the actual 35-millimeter slide is. Okay. If Swisher says one last question, is there a difference between digitization and digitalization? My answer to that is I don't think so. I think those are two terms that describe the same thing, maybe different uses in different countries or translations. Although I welcome any thoughts that are different than that. Sometimes we get that question of what's the difference between digitization and digital preservation, and they're, of course, related, but one is more about the capture of the images and the other is more about the long-term stewardship of those images. But I think these two terms here really mean the same thing. Okay. I think we have to go. So thank you very much. I hope we'll see you guys next week for our webinar on NAG prep. Please fill out the evaluation, and that's it. Okay. Bye-bye. Thank you both. Thank you. You're welcome. Thanks to everybody for attending. Thank you.