*Breaking* Techdude Finds Original Text on Alleged Obama Birth Certificate -(prior owner "Female")

texasdarlin.wordpress.com ^ | August 3, 2008 | Techdude

[supercat]

They are not “non-matching fonts” One is reproduced at one magnification, the other at about 3 times that.

I don't know why the term "non-matching fonts" is used. The issue in that picture is the text near the blue arrows. The 'magnified' version does not look like a magnified version of the smaller. It could just be a scanning defect, since the quality of the smaller scan is so crummy, but it does seem somewhat suspicious.

[MindBender26]

You are right, it is scanning artifacts, just as exists on the other listings.

[supercat]

Techdude results cannot be reproduced according to this:

Techdude does not state his case very well. What he is looking for are not letters per se, but rather anomalies whose edges line up with various possible texts. Not a bad approach, but one which must be approached cautiously. A good way to approach such things would be to have one person clip out a dozen or so areas of what should be identical background, and give those clips, along with a randomly-ordered list containing some texts are supposedly found in those places and a few texts that aren't, to someone else to examine. If the person examining the clipped-out regions can determine which texts are associated with which reason, that would suggest that there may be 'something there'.

A key aspect of the test is that if it is done correctly, no amount of wishful thinking on the part of the person examining the clippings would be able to determine which ones contained which text unless there were indeed something in the clippings to support such identification. Implementing the test protocol is not quite so easy as it might seem, though, since there are a number of ways the results could be biased. Better than any other method of validation I can think of, though.