Maricopa County sheriff Joe Arpaio held a press conference this evening presenting his conclusions that Barack Obama's birth certificate and selective-service registration are frauds. Normally this isn't something I'd give the time of day, but it involves things I like to play with (scanners and image-editing software), so it seemed like a pleasant Thursday evening diversion.
Among the evidence released at the press conference are five videos – which can be seen at the end of this article – to demonstrate why the Obama long-form birth certificate is suspected to be a computer-generated forgery.
The videos consist of step-by-step computer demonstrations using a control document. They display the testing used by the investigators to examine various claims made by supporters of the April 27 document.
The investigators contend the videos illustrate their conclusion that the features and anomalies observed on the Obama long-form birth certificate were inconsistent with features produced when a paper document is scanned, even if the scan is enhanced by Optical Character Recognition, OCR, and optimized.
Arpaio tweeted his press release. The press release has links to the videos. The one I'm interested in is this, which deals with the "fact" that when you scan an image to PDF and open it in Illustrator, you only get one layer.
This has been debunked over and over again, but DIY is always fun. Fortunately, I have to scan old documents to PDF all the time. So I had plenty laying around. Like this one, of an article I posted recently. I scanned it on a Xerox WorkCentre and opened it directly into Illustrator. Here's what it looks like with all the groups visible:

And with all but one hidden:

Poof. Random parts of her dress disappear.
If you actually managed to sit through the video, you may have noted how they briefly skimmed over the fact that one layer has "most of the text." This video shows what happens when you don't skim over it. And it's a slightly better example since it's an old-school form, plus the guy has a pleasant German accent:
Yes, it's silliness. But I'm always interested in the structure of silliness and how it spreads, even to big-city sheriffs.
The most common rebuttal to the claims of fraud comes from an Adobe Certified Expert, Jean-Claude Tremblay, who states:
——————-
(Jean-Claude Tremblay) said the layers cited by doubters are evidence of the use of common, off-the-shelf scanning software — not evidence of a forgery. “I have seen a lot of illustrator documents that come from photos and contain those kind of clippings—and it looks exactly like this,” he said.
Tremblay explained that the scanner optical character recognition (OCR) software attempts to translate characters or words in a photograph into text. He said the layers cited by the doubters shows that software at work – and nothing more.
——————-
The obvious question to ask is this: If the layers are the result of OCR at work, then where is the text? OCR output results in editable text. It also results in a font list. Neither of those things is present in the pdf.
Now explain the white glow around the writing and why the background stays straight when the writing & lines are curved on the left.