Technical Aspects of Digitization and Project Management

• All items were digitized using 400 dpi resolution and a 24 bit color scale. All master copies were saved as a .tif file.

The resolution standard are common in large digitization centers. However we chose to use
color bit scale for all scanned items, weather black and white or in color. This was decided
off the suggestion of the PCL's digitization center's director - Uri Kolodney. Since learning more about digitization I have been playing around with old family photos. In my own observation, using a bitonal grayscale for black and white photos seems to maximize the quality of the finished project. If I were to do this again, I believe I would choose a grayscale for the black and white photos in my collection (*I would play around with the objects and our equipment before deciding of course).

• All scanners and monitors were callibrated before use.

I still have a lot to learn about callibrating equipment properly in order to obtain the best quality scan. However, I have definitely learned the importance of its function. This really hit home when I work on quality control of the items, which I will discuss further in the next bullet point. Some of my items were oversized and needed to be scanned by the digibook in the PCL digitization center. After looking at the finished product that this type of machine produced, it was painfully clear how much attention you should pay to the level of quality scanner you purchase or use for your project.

• An inventory of all items were documented in an excel spreadsheet that included the: file name, condition notes (such as tearing, scratches, folds, stains, etc.), material type (i.e. photograph of artist), treatment, receive date, and return date. This type of inventory is necessary in digitization centers to not only keep track of the items in the collection but to hold the employees responsible for handling the materials with care.
  

I was surprised by how long the documentation portion took, it was not something I would have accounted for in terms of managing a digitization project before.

• All scanned items and documentation were saved on UT server space set aside for this specific project. There was an overall folder for the finished project entries and subsets that kept individual students' work separate for easy access and organization.
  

Having never worked with this type of storage method, it took a while for me to get used to the computational path to retrieve and upload files. It was a good experience learning to keep the folders straight and how to route newly scanned items into their appropriate folders.

• All digitized items were reviewed and edited in Photoshop for quality control and then saved as a .jpeg file for publication on the web.
  

Quality control was much more time consuming and draining than I had anticipated. The documents provided by the digitization center were extremly helpful in knowing what to look for and how to edit the digitial object for the best results. Before working on these items I did not notice the quality of digital objects, especially in terms of their color. I really noticed the difference in quality callibration while handling this portion of the project as well. The items in my collection were not all scanned at one time. The first set of objects that I scanned required quite a few adjustments in terms of the color levels. When I came to the next set scanned on a different machine (callibrated by a professional), the levels needed little adjustment and took less time. At this point I also had to "bandaid" dust bunnies and other obstructions on the scanner (often invisible to the naked eye, but affect the overall look of the finished product) which demonstrated how important it is to keep your equipment squeaky clean. I decided to add an extra step by creating another folder and saving the QC'd documents as a .tif to start out with. I did this because I am new to Quality Control and thought if there was something I wanted to correct with further precision, it would be best to work off a .tif rather than the final .jpeg and save me time overall. I used Photoshop's batch processing

• All students in the class had to come up with a naming convention for the digital records so that consistency was maintained.
  

The naming convention was much more problematic than I anticipated. Choosing a name that described the collection and objects accurately but concisely was difficult. Here is what we came up with:
blam_exhib_0003_1969_m_001.tif
The first four letters indicate the BLAnton Museum. The next section identifies the collection as an exhibit. Each exhibit was assigned a number, that we came up with in class, and this is indicated in the next section. The year of the exhibit was including to also give the items more identifying information just by glancing at the file name. The next letter indicates the type of material. For this, we conformed to the Harry Ransom Humanities Research Center's convention for file-naming construction: a = Art: sketches, paintings, engravings or woodblock/metalblock prints, posters, etc. m = manuscripts (correspondence, letters -- handwritten or typed, pamphlets, brochures, telegrams, drafts, legal agreements, etc.); and p = any photographic material (slides, negatives, photographs, etc.). The number of the individual objects were indicated using a three digit standard and then of course the file type finished off the file name.

Early on we had discussed a letter extension that indicated if you were scanning the front and back of your item. We decided against that and just increased the total number by one if you had to scan the back of an item, or an item had more parts to it (such as a pamphlet or a document with several pages stapled together).

We really should have decided on this earlier in the process, and stuck by our decision so as to save time overall. It was a good learning experience nonetheless.

• All students created a MODS record of their collection.
  

This aspect of the project required a steep learning curve if you are unfamiliar with metadata schemes and how xml works. Both of which I know little. However, this was a very good experience. I managed to populate most of the fields in the MODS scheme. The ones I left out included: table of contents (because I was only describing the item as a whole and not its individual components, classification (due to the fact that this item has never been cataloged), related item, part and extension.