Jul 20 2008

The End Of The COLB Cult – Myth Busted Again

Update: Reader Ray from Australia noted this article wherein an official of the state of HI looked at the BHO COLB and said it looked to be real.  

When the birth certificate arrived from the Obama campaign it confirmed his name as the other documents already showed it. Still, we took an extra step: We e-mailed it to the Hawaii Department of Health, which maintains such records, to ask if it was real.

“It’s a valid Hawaii state birth certificate,” spokeswoman Janice Okubo told us.

As shown below, how could anyone miss the unique style of the border? There’s more – there may now exist another COLB with the same border:

We circled back to the Department of Health, had a newsroom colleague bring in her own Hawaii birth certificate to see if it looks the same (it’s identical). 

As I said below, just one more COLB formatted like Obama’s and Pam’s expert has egg on his face. Update: I have a question for Techdude and others who have multiple COLBs in their position: Did you ever get one that looks like the BHO COLB with its style of borders?  And if so, why didn’t you publish that information?  You may not, but these witnesses indicate there were two formats in circulation.  Kind of hard to believe you folks never saw one, but it is possible. – end update

Why are there two formats? Either one was a prototype discontinued or HI is just running through the last of their old stock. But the point still stands – to accuse someone of a crime you have to more the claim it is possible. If anyone has a COLB that has the same border pattern as below I would be interested in hearing about it [ajstrata@strata-sphere.com]

Finally, Hat Tip to Allahpundit at Hot Air for the link.  Hope your sin of support doesn’t cause too many headaches from the cultists.

Update: Well, if someone wants to know why this myth-mess is good for Obama, here’s an exampleend update

Well, it seems we finally have the final report, representing the final act of the COLB Cult today. Techdude has produced something that impresses the non-technical groupies of the movement. But does it hold up to the argument there is evidence, beyond a reasonable doubt (my standard for when I confront smears)? No, not really.

One thing to recall is the track record of failed claims from various cultists since this began. Each claim debunked, followed by wilder claims, which were debunked. Here is a recap of the trend:

(1) The image on Kos was a forgery because it was not a BC but a COLB (took people a while to realize HI had moved to issuing COLBs instead of BCs).
(2) The image was a fake because there was no stamped seal – later discovered existing in the image.
(3) The image was a fake because there was not signature seen on the back – later discovered in the image.
(4) The image was a fake because there was no date stamp for when it was issued – which was detected and evident to the casual viewer.
(5) The image was fake because it was photoshopped – later shown to be a cropping of the excess paper area and blacking out of the ID number
(6) The image was fake because it was different from the BHO version – later discovered to be the fact BHO saved the image to a lower resolution, smaller web friendly version.
(7) The image is a fake because it was made from a forgery started by Opendna – that lame claim lived for weeks, but in the end was proven impossible.
(8) The image was a fake because the borders did not match a 2002 COLB – one of my favorites because of this line from the aforementioned Techdude:

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.

Here’s the image of the two borders – how could anyone not see that? Subtle – eh? Techdude later admitted THIS claim was wrong – there had been updates between the 2002 and 2007 COLB formats. He also debunked the Opendna claims.

(8) The image was claimed to be a fake as more and more people compared the 2002 COLB to the newer 2007 COLB – all of which proved nothing more than we already knew, Hawaii was updating their vital records and COLBS in response to 9-11 and the Real ID Act of 2005, as can be seen in the variety of COLBs found for the period

(9) The image was a fake because the document format version control number never changed while the format did – which we now know is because the format control number covers the data fields and contents, not the look and feel.
(10) The image is a fake because a date field has a format anomaly in it – but then the exact same anomaly was found in other COLBS from the same time frame.
(11) Anti-aliasing haze around the text was a sign of forgery – when it is actually only a sign of text being printed.

Today Techdude has his report out, so let’s see if he has a smoking gun or wrecked professional credibility.

Sadly, Techdude doesn’t provide an analysis, he produces cherry picked data from one side only. A real analysis compares the evidence that supports the opposite conclusion. From the list above we see tons of reverse examples. Assumptions of what real image would contain, which later proved to exist. All those debunked claims are also evidence that the COLB image is real. It from the totality of the evidence that we decide which side is MORE COMPELLING. A real technical analysis would address all the evidence and discuss why some crazy marginal border measurements outweigh things like the impressed seal (hard to fake), a signature area, etc. All things COMMON to valid COLBs and images of them, but hard to fake out using graphics programs.

And much of what Techdude discovered PROVES points I made way back – which supports the claim the image of the BHO COLB is a scan of a legit COLB. For example, I predicted that as part of the upgrades to the COLB that came on line in 2006, there was a change in the paper which was probably a more secure paper-cloth hybrid. This new paper was thicker and made the seal impression harder to detect from a front scan. Techdude confirmed this again today:

The embossed seals and ink stamps in all of the pre-2006 images are clearly visible in the scans however none of the post-2006 seals or ink stamps are visible without extensive manipulation to the digital images. Even when scanning the physical post-2006 certificate in my possession using multiple resolutions and using multiple scanners I was also unable to produce an image which would allow the seal to show though the image. The ink stamps on the rear side were also not visible in the front side scans without digital modifications to the scanned images. My scans of the physical certificate also produced the same results using multiple resolutions and using multiple scanners.

This destroys many forgery claims by Polarik and others who compared the 2002 COLB and BHO 2007 COLBs based on these features. And in fact, the similarity in the results by Techdude on taking modern COLBs and seeing the same telltales in his scans as shown in the Kos image means he proved the Kos image is legit. When you use real world COLBs and perform the same scanning process using a range of variables and you create the same result this is confirmation.

It should have been noted as such in any professional ‘report’. Also, if you want review of your work you need to provide the data. Figure 3 is a low quality image of the ‘smoking gun’. When I tried to down load it and blow it up it was impossible to see. Sorry, but peer reviewing requires the actual data. I will forego any detailed comments until Techdude and Pam decide to provide all the data.

But let’s step back again to the blatantly obvious fact the BHO COLB has unique security borders – not even close to the others. Techdude goes into minute measurements to prove the obvious – the borders are not the same. Why? We know they are not. Not the same color, not the same pattern, not the same dimensions. What Techdude has not proven is that the 2007 BHO COLB is the one and only HI COLB with that pattern! Techdude has two other COLBS (from 2006 and 2008) with telltales IDENTICAL to the BHO COLB. These include version numbers, data fields and contents, border lettering, seal imprints etc. Commonalities that cannot be ignored.

Finally Techdude repeats his horrible mistake from day one – he goes back to comparing apples to oranges. Remember, Techdude has some original COLBs, which he or someone else scanned. They were not scanned by the same person who scanned the BHO COLB. They were not scanned on the same machines using the same settings. Why is this important?

Because the tools Techdude wanted to use to do image analysis are used to detect image manipulation WITHIN an image. These tools and methods look at inconsistencies with in an image to show where the image was manipulated. an image should have consistent sharpness, depth, resolution, etc if it has not been tampered with. Areas where there is known tampering would show up easily (for example where the ID number was blocked out). A real expert in this field did the analysis on the BHO image and concluded they were reasonably untouched (minus the known manipulations).

What did Techdud do? To my shock he took three different images and tried to compare across them!

The 2007, 2008, and the two KOS images were then analyzed by creating a heat map showing where each pixel changes as jpeg quality decreases from 100 to 0. A change was considered relevant once the sum of the changes to the red, green, and blue values exceeded 10%. The heat map created from the 2007 and 2008 images showed the fonts, seal image, and security border are all identical consistent values. To eliminate any subjective presumptions and to increase the number of comparative tests the same analysis was then conducted on the 2006 and prior certificate images which all found the fonts, seal images, and security borders to also be saved with identical consistent values. The same analysis on the KOS images showed the security border having a substantially different RGB quality value than the fonts and the seal image.

It is a rookie mistake with these kinds of tools, to try and make assumptions across uniquely created images of different formatted documents. There is no conclusion that can be drawn when doing this except the documents are different and the process of image capture was not the same. It doesn’t prove anything beyond that (and Techdude knows this). Here are the images:

Original here.  Clearly there is are differences, but I want to note the blacking out of date fields to show what a ‘forgery’ we show. You can see in all three where the modifications were done – they are impossible to miss. They have sharper edges, they have different colors.  While all three show different RGB results, all of them are the same thing – efforts to black out personal data. You need to look at changes in the context of the image itself, not across images.  For example, the fact COLB 2 has bright borders on the black outs and COLB 1 doesn’t is not a sign one of them is forged. 

The BHO COLB is unique to the set of COLBs so far seen. But we can end this mess in a heartbeat. If anyone has a HI COLB that is of the format and structure of the BHO COLB please share it with us. If you got a COLB around Jun 2007 my guess is you might have the variant we saw with the OBH COLB, which seemed to be around for a while as an experiment or prototype version. Just a scan of of the border is needed – nothing more.

Techdude has given his groupies more false positives to run around and play their games with.  He also showed, unintentionally, why the BHO COLB image could be real.  He confirmed the paper upgrade and showed empirically that all the new COLBs, post 2006, shared the same traits regarding the impression of the seal.

Now what are the odds someone who screwed up the borders so badly also caught the paper quality detail? Basically they are – zero!  No one who did enough investigating to make sure the seal imprint was light due to new, thicker paper would miss the borders.  What are the odds a person who screwed up the borders so badly was able to know to put the date field anomaly in?  Zero as well. Thanks Techdude, but as usual you proved my point.  When you claim such an obvious and glaring disconnect is the work of a forger who was able to get the other mountain of hidden details right – you have debunked yourself. 

Myth busted – again.

81 responses so far

81 Responses to “The End Of The COLB Cult – Myth Busted Again”