Week 5 Reading Notes

1.) Data Compression Wikipedia

Data compression is encoding information in order to use fewer bits. Both the sender and receiver of the information must be able to understand the encoding scheme. Compressed data is easier to transport, but most be decompressed before used. Lossless compression schemes are reversible without information being lost. Lossy compression schemes do lose information, but have a higher compression. Lossy compression is used frequently with photographs and videos because they can be compressed and much of the information lost is not noticeable to most people. Run-length encoding is used when a string of identical data values are replaced with one of the data values and the number of times it is repeated. Lempel-Ziv method is very common in lossless compression.

2.) Data Compression Basics

Data compression allows you to store the same information in less space, and also makes it easier to send information. Run-length encoding isn't beneficiary for compression of all text because it could make it larger in some cases. (eg: Mary had a little lamb.) Run-length encoding (RLE) really only works when there is a repetition fo 3 or more characters, which is rare in short texts. There are ways to compress only parts of the file that are designated by a marker (in this case a *). Can make"Item X...................................$100.00" into "Item X *[60].*$100.00. You want to choose a marker that is rarely used as to not confuse the computer. RLE is also useful for encoding images that have large areas of the same color.

Lempel-Ziv (LZ) uses chunks of information that are repeated to compress. It codes how far back to look and how many characters to copy. Endtropy coding uses a number or numbers to replace frequently used sequences. Prediction coding is used frequently for images. It is encoding the difference between the predicted value and actual value of a pixel. It is better for natural images such as photographs, and not as successful for graphic images like logos.

Movies can be compressed by individually compressing each still image that makes up the movie. You have to be sure that the receiver will be able to decode the images quickly enough for the video. There is a difference between a video file and video stream. A video file has a definite beginning and end, while a video stream is a constant flow of video (eg: a live broadcast). Vector animation can be used on certain types of video files to take up less space. Temporal delta encoding stores only the difference between the frames. I understood the use of macroblocks, but am confused about how the blocks move from frame 1 to 2. Something that was interesting was how some encoders can tell if there was a scene change and that the difference is too great to warrant a residual image. Instead it will create a keyframe. Part 1 was pretty straightforward and helpful. I couldn't access Part 2. The link didn't work, and it didn't work when I manually typed in the address. I'm not sure if anyone else had this problem or not. Part 3 was a bit more confusing to me.

3.) Imaging Pittsburgh

This project, a partnership of Archives Service Center of the University of Pittsburgh, Carnegie Museum of Art, and the Historical Society of Western Pennsylvania, focuses on making historic visual images available online. The author hopes that these photographs will help show the city's history and changes throughout time. Users can search keywords, or browse a particular collection. One interesting aspect is that users can order prints of the photographs. The program has come across some difficulties in communication between the participating organizations and people. They created a listserv, but it is not used often unless someone reports that they have finished a task. The author isn't sure if this is failed communication, or that people just understand what they need to do and don't need to communicate issues. There have also been issues about processes and language used to describe the collections because of the different focuses of the institutions. The project participants also had to coordinate how they would arrange their metadata and use a common vocabulary. Additionally the project group set a minimum standard for the master images to ensure consistent quality. They launched in February of 2004, but are taking until the Fall 2004 to review the website, respond to problems.

4.)YouTube and Libraries

YouTube is a fast, and inexpensive (free) way to reach a large audience. First you need to set up a user account and channel profile. Users can then select a video file of 10 minutes or less and upload it. Librarians will need to make sure they address copyright issues of any material they plan to upload. Librarians could post informational videos that could help patrons access online information when the library is closed. They could also create a YouTube video to introduce patrons to the library's services. Using screen capture software a Librarian could create a tutorial for students.