Awesome point about info vs. data:
friend sends you 2 emails, so you have 2 emails worth of data, but only 1 email worth of information. Info and data are not the same thing.
Run length encoding:
-using numbers instead of repetition of same characters to save space
-uses markers for specific characters
Lempel-Ziv compressor family:
--> uses markers for "patterns" of characters to encode with a smaller number of bits
Entropy coding:
-assigns codes to BLOCKS of data, works well for codes w/ variable lengths
-uses probability
Predictive compression is better for more "natural" images (i.e. photographs that have gradual shifts in color/value)
compression might not contain ANY of the original data
lossy preserves meaning, not exact data. how does a computer decide what is meaningful? it can't.
quantisation: division of a continuous signal into discrete steps
video compression:
each frame is compressed independently
vector graphics describe images as a set of geometric shapes, textures, and forms
This article opened my eyes to the various types of encoding currently in use. Seeing the difference visually between lossless and lossy encoding really helped me to grasp the concept much better than after just reading the wikipedia article.
Subscribe to:
Post Comments (Atom)
1 comment:
Rachel,
Reading your post made me think of something else...we transferred over to Google Apps this summer at work and they have a very similar tutorial to show why it will be important to use Google Docs. It used the same idea of a friend sending two emails, there being two messages with the same amount of information and it just continues and continues, but really there is only one piece of information. That is why we post the documents and share them now instead of sending emails. I never thought of it that way, until you made the comment about the two emails...
Rian
Post a Comment