Back to Home

Documentation on Website's Development

Christy Richards
INLS 210-046
April 1st, 2003
Dr. Tibbo

Introduction

This digital project was created in the fullfillment of the requirements for INLS 210-046, Digital Preservation and Access. The following paper which chronicles the development of this website, is meant as an accompaniment to the exhibit.

Selection Decisions

The selection process for this project was relatively simple. I first thought up a few criteria that I would like my selection to meet. First of all, I knew that the digitized items should not be subject to copyright law. I wanted to create a digital exhibit that would become a permanent part of my portfolio, without worries of rights violations. This limited my selections substantially, so I decided to choose something from my own family. However, I also wanted something visually interesting and engaging--something that would make someone want to re-visit the site. I thought of the mountains of artwork that my brother had produced over the years and my decision was complete.

Or was it?

I still had to choose what artwork to digitize. His artistic output could be categorized into several broad categories: drawings, photographs, wood sculptures, ceramics, digital art, and miscellaneous three-dimensional art. I immediately eliminated the categories of wood sculpture, ceramics, and other three-dimensional art. For me to digitize these images, I would have had to take digital photographs and then put them up on the web. Because I don't have long-term access to a digital camera, the point was moot. This left me with three possibilities: drawings, photography, and digital art. Although my brother's real strength lies within digital art, I eventually decided against including anything on my website because I didn't know any information about the images to produce successful metadata. Also, I got the distinct impression that the purpose of the project was to digitize, not to work with born-digital formats. Thus, my decision was made: I would select drawings and photographs.

From that point, selection was easy. I knew that I needed some sort of text, so I went through some of the stories that Matthew had illustrated for me as a child. I found a good selection from "Life on Planet X", so I chose that. I then went through a binder of his childhood drawings and pulled out those that best showcased his imagination. At last, I went through his photographic portfolio, pulling out selected images that I thought would be most visually appealing to viewers.

On an interesting note, I did actually come upon a copyright debate in my selection process. Originally, I had selected one of Matthew's childhood drawing that depicted a group of Civil War soldiers fighting during a battle. My brother had been a big fan of the Civil War, so I thought adding this image to the project would show his diversity. I digitized it, appended metadata and put it up on the web. However, 2 days before this project was due, my brother informed me that he had traced that image out of a Civil War book. It was not his original artwork at all. Because of this, I had to remove the image and put another in its place. I came to realize that even something as straightforward as children's drawings could have copyright issues.

Digitization Decisions

Because both the drawings and photographs were two-dimensional art, it was self-evident that all my digitization would be on a flat-bed scanner. Although scanning was simple for 1 page items, I found it difficult to scan pages from the copybook that contained "Life on Planet X". The images were nearly impossible to scan straight, and attempts to straighten them in Photoshop after the fact proved fruitless. What I learned is that it takes greater patience when scanning books as opposed to flat pieces of paper.

The original purpose of choosing the text images was to experiment with OCR software. However, my handwriting showing up as garbled language when I ran it through OCR, so I was forced to transcribe the text. I transcribed directly into my website authoring window, because if I had tried writing it in a Word Processor, the automatic spellcheck would have corrected many of the typos and grammatical errors that riddle the document.

Digitization Process

My digitization process was pretty straightforward--I got into a rhythm and began using the same process to digitize all the images. First, I would lay down the image onto the flatbed scanner and preview the scan, I would then select the portion of the image that I wanted to scan. (Because the scanner glass is larger than 8.5" x 11" which was the size of most of my images, I only selected the image itself.) Then I would open up my metadata file that I had saved in Microsoft Word. I entered a lot of the metadata right there at the creation of the digital object. I would record the size of the scanned image; change the dpi; alter the highlights, shadows and midtones of the picture, if necessary; and change the sharpening, if necessary. Also if the picture had no color in it, I would choose to scan in Grayscale. However, if the image had any color, I would choose True Color.

For all the images, I scanned at 600dpi. I had talked with several friends who scanned images on a frequent basis, and they said that 600 was pretty standard, and anything less than 300 was generally frowned upon. Because I wanted to preserve the clarity of these images, I took their advice and scanned at a high resolution. However, at the time I received their advice, I had already scanned in the text images for "Life on Planet X" at the standard 200 dpi. Although I considered rescanning these at a higher resolution, I decided against it, because the images look fine as they are. Moreover, they are just a sample from the story. If I ever chose to scan the entire story, I would do so using a higher resolution.

Once I had finished making these adjustments to the image, and once I had recorded the metadata for that image, I would save the image as a .TIF file. Eventually all of these .TIF files would be burned on 2 CD-R's for my own personal files.

Once all the images were scanned, I went back and opened the digital master .TIF in the program Kodak Imaging. I saved the image as a .JPEG and changed the resolution to 200 dpi. This automatically compressed the file by 10%. Then, I went into Adobe Photoshop and decreased the size of the image. The size of the item, as scanned, was too large for a web browser window. Once I found a decent size, I saved the image under a different name. I then returned to my Metadata File and entered the new information about this image: the pixel size of the image; the new dpi; the bit-depth; and the file-extension were all added at this point.

This stage of the image is what appears on my website as a close-up. Using this version of the image, I then generated thumbnails to go on the site. Once all of this was completed, I uploaded these images to the web.
 

Challenges of Metadata Creation

There is no doubt that creating metadata slowed down this procedure a great deal. I had to deal with an entire page of metadata for every image. The tricky part about this, was that I had to generate metadata at creation. This limited my ability to use previously-digitized images, or digital-born objects, since I knew nothing about how they were created. Moreover, I found the metadata schema somewhat difficult to understand. At first, I believed that the metadata was supposed to reflect the .TIF file, so I filled in all the fields for the .TIF, but then I realized that the metadata was supposed to reflect the actual image that was appearing on the website, so I had to change my metadata to reflect this. Also, some of the fields were a little uncertain. For example, the field that is listed "scanned height" and "scanned width". These dimensions in my metadata reflect the original size of the image when it was first scanned into the computer; it does not reflect the current size of the image. I am uncertain if this is correct.

I suppose my confusion comes from the fact that I don't know how much of the metadata is supposed to reflect the digital master .TIF and how much metadata is supposed to refer to the .JPEG version of the image.

Conclusion

I think that, over all, this has been a good experience. Indeed, this project is one of the reasons that I originally took the Digital Preservation and Access Class. I must say that the process was far more complex and involved than I ever could have anticipated. Even after hours of fine-tuning, I feel my exhibit is far from perfect. Indeed, I had originally anticipated scanning an additional five images from slides that my brother had produced. Unfortunately, time constraints hampered this goal. I still regret the inability to add these slides, because they exhibit another dimension of my brother's art and serve as an interesting contrast to his earlier work. Hopefully, I'll be able to add these images in the future. After all, this is a project that I want to continue developing, expanding, and improving as time progresses.