• noise: Adding noise to text changes the letters in a manner which accentuates or adds a feature to their original shape. Occasionally noise obscures or removes a feature of a character, but not very often.

• blur: Blurring a character does one of two things. If a character is sufficiently narrow and straight, it makes that character narrower and straighter. Otherwise it makes the character rounder and shorter.

• enhance: Enhancing text made up of random characters attempts to turn that text into something vaguely recognizable and somewhat pronouceable. This creates the illusion of semantic content in something that, in fact, has none.

Blur and noise are done using per-character mappings. Enhance is done using bastardized Perl regular expressions which attempt to create letter groupings commonly seen in the Romantic languages.

To use this applet, enter text in the big box area and choose a text filter from the menu of choices. Then enter the strength of the filtering (a number between 0 and 1, inclusive) and press "Apply Filter" to see the effect of applying the filter to the text.

Try applying many iterations of the "Enhance" filter to text that has been obscured by noise or blur. A value of .5 is a good setting for "Enhance".

NOTE:
There may be a fairly long delay when applying the "Enhance" filter for the first time. This is caused by Java loading several classes over the network.