Replies

Just found this at Atlas Shrugs:

http://atlasshrugs2000.typepad.com/atlas_shrugs/2008/07/omg——blogger.html#comments

Posted by: Underzog | Friday, July 04, 2008 at 05:27 PM

As a real forensic computer investigator (board certified, investigated thousands of cases, access to a full forensic computer lab, yadda yadda...) I decided to jump into the fray over the fake vs real discussion a week ago when a friend of mine challenged me to see what I could find (since according to him, the document was clearly a real one). He is what one would call a slightly rabid Obama supporter – he even has the tattoo to prove it.

First things first...I used the latest online KOS version (downloaded 6/29/2008) as it was supposedly the first released digital copy and the Decosta certificate from Israel Insider’s blog (also downloaded 6/29/2008) as an exemplar of a real certificate.

Of forensic interest on the KOS version is the EXIF metadata (encoded in the digital image data) showing that the KOS certificate was modified or created with “Adobe Photoshop CS3 Macintosh” on “2008:06:12 08:42:36”. Assuming the EXIF data was added after the document was scanned (assuming it is real) and the document modified only by placing a black-out section over the Certificate Number or if it was converted and saved for web use this would not be too unusual. The Daily KOS released the story on “2008:06:12 08:44:37” which would be 2 minutes after the graphic’s EXIF stamp was created.

Before I became a forensic geek I worked for [a herein nameless publicly traded company] that designed counterfeit detection hardware and software for the banking and retail industry. The company was very high profile and we received training from [a certain herein nameless department of the Federal government that knows a thing or two about counterfeiting] - but I do not claim to be an all around expert in Questioned Documents but after several years of working with them I do know what to look for to spot an obvious fake.

That being said...I too was able to see “something” that looked like a seal - however even using highly specialized software (specifically designed for law enforcement and forensic examiner use for cleaning up digital images and video) I was unable to recover anything more than what others have previously released using a basic find edges and modification of the contrast. Even when the background security paper pattern was removed there was not enough of a “seal” to view anything usable or verifiable which is a bit odd but still might be possible if the seal had been completely flattened out in the mail, etc. There is also a fold which runs vertically across the top of the certificate which is close to the same location on the Decosta certificate.

At this point I was beginning to believe the certificate was real until I resized and overlaid the Decosta certificate on top of the KOS version. All things being equal there was a 3.82% difference in the size of the KOS version vs the Decosta certificate - but again depending on the optical distortion from the scanner this too was explainable. But upon manually stretching them to match edge to edge I caught a glimpse of what I and apparently everyone else had simply not noticed. The security borders do not match. Literally. They are not even close to identical. For instance “Decosta” contains five 10 pixel wide “diamonds” per vertical row while the “KOS Obama” contains 2 to 2 1/4 36 pixel wide diamonds per vertical row. These differences can clearly be observed even with the naked eye although you may need to enlarge the graphics on your screen. Taking the measurements further - Decosta’s “Certification of Live Birth” heading is centered between security diamond pattern and is 762 pixels wide @ 0% angle while Obama’s “Certification of Live Birth” heading is not centered evenly between security diamond pattern and is 794 pixels wide @ 0% angle. Decosta’s “Any Alterations” footer is centered evenly between security diamond pattern and is 1244 pixels wide @ 0% angle while Obama’s “Any Alterations” footer is not centered evenly between security diamond pattern and is 1294 pixels wide @ 0% angle. I kept the comparative screen shots in case anyone wants them.

I am unable to explain the differences between the security diamond sizes and counts and the un-centered portions (meaning the diamond pattern ends on an odd pattern instead of even where it meets the edges of the header and footer boxes). Looking closer at the KOS certificate (magnified to 400%) clearly shows inconsistencies in the security border such as cut and paste marks and overlaying of the side borders where they meet the top and bottom. This effect is not observed in the Decosta certificate at any magnification. Another point of interest, removing the background security pattern did not remove the background area from underneath the security border on the KOS certificate. The color and hue values of the background pattern located and viewable through the security border are also not a match to the rest of the certificate background. I can not explain these discrepancies. I then noticed there were some indications that the background pattern had been duplicated and placed in various locations to clean up the document. Now at some point I just started to laugh and went out for a smoke and gave up looking for more.

I am convinced that the certificate is a fake (and not really a very good one) and I went into this with a completely open mind (something the Obamanationalists seem to have lost). I also have to say that everyone who has been looking into this federal crime (and it is a federal crime even if the certificates were never meant to be used for identification) have done a stupendous job and I wish they all worked for my lab. Talk about a winning team.

Posted by: techdude | Friday, July 04, 2008 at 09:19 PM