The truth about the "fake" birth certificate

Many of you probably know that I spent too much of last night (while simultaneously trying to make a morning deadline on a for-pay project) arguing with the acolytes of one Techdude over on No Quarter.

The Dude is, we are told, a courtroom-quality expert in computer imaging. Although I don't know the man's name, I'll take his claimed resume at face value -- for now. (His followers are very quick to wave that resume in your face if you dare to question the Dude.) The Dude still insists that Barack Obama's Certificate of Live Birth -- the dreaded COLB -- is a fake.

His latest argument, here, comes down to this: He says that a forger created the Obama COLB by changing an original which belonged to someone else.

That's possible. In a sense, that was the tactic I used when I created the Wailing Wall "PS" note seen here. I enlarged a copy of the original note, used cloned background material to erase Obama's pious words, then overlaid my own (bad) attempt to mimic Obama's handwriting onto the "blank" surface. The results are sloppy, but so what? The intent was humor, not deception.

Techdude believes that something similar occurred in the case of the COLB. The Dude argues, based on the evidence of another Hawaiian COLB of known provenance, that the original text was once in a slightly different position.

Moreover, he thinks that the original text can be recaptured, much as one can use special techniques to read a palimpsest. He believes that he has done this very trick. He avers that he has found the word "FEMALE" where the sex should be listed. (In other words, the forger worked from a female's COLB.)

He also says that he knows the original name on the COLB -- and that this name is both familiar and surprising. He decided to hold it back for a bit in order to give the guilty parties time to confess.

Which means that, during that time, others should be able to replicate his work. That's what I tried to do. Unfortunately, Techdude's "simple" instructions are anything but.

He begins his piece by implying that his techniques require no special expertise. Even a child can do it:

That being said, it never ceases to amaze me how some people will always just refuse to see the facts when they are placed in front of them. Perhaps they have just been blinded by their own creative interpretations of what they want to see but I will simply put my money on them just being really stupid

Since a lot of people want to try some of these techniques at home all they would need is a copy of the full uncropped KOS image and a copy of Michelle’s 2008 COLB. Both of which are freely available on this site or on sites linked to from this site.

Well. Any man who talks in that fashion should be able to present a replicable procedure. He's not saying: "Trust me, I'm an expert." He's saying: "Hey, kids, go ahead and try this trick at home. If you can't do it, you're an idiot."

What, precisely, is he doing?

I've read his text closely, and I'm still not sure. I ask you: Do these sound like the kind of instructions a tech guy would give you?

First, change the overall hue value to red.

A-HEM. Which program are we using? The Dude, it seems, is too cool to divulge such information. I'll use Photoshop, the industry standard. (CS2. I never got used to CS3.) Which tool are we using? Which way are we pushing the slider bars? What percentage?

Most of you will consider this point inconsequential, but "hue" and "value" are two very different properties of color. The two words should not be pushed together in this fashion.

(All color has three properties: Hue, value and chroma. Next time an English major in a museum starts spouting pretentious nonsense about art, ask him to name the three properties of color -- and then ask him if someone should talk about literature if he doesn't know the difference between a noun and a verb. This is one of my favorite tricks.)

Okay, back to our story.

Next, adjust the contrast a bit until you can see all of the light colored pixels appear from the majority of the “white space”. They should appear as slightly off white with a tinge of red or yellow.

"Adjust the contrast a bit"? Not a very precise instruction. I used the contrast slider in Photoshop, and could get no results like the ones that the Dude displays. (You have to adjust brightness as well, methinks.)

At first, I didn't know what to make of the phrase "all of the light colored pixels appear from the majority of the “white space.”" He probably means that one should adjust brightness and contrast until one has squeezed out all possible latent information in those pixels which may appear white to the naked eye.

Okay, fine. That makes sense to me. But I'd still prefer some exact numbers.

After adjusting the so-called "hue values" and fiddling with the contrast (and brightness), the Dude gives us a document like unto this (click on it for enlargement):

Uhh...Dude? The text is red. In the original, it's black. And the background pattern is amber, not red. (It was green in the original.) No matter how you adjust the hue slider in Photoshop, you won't transform black into red.

In my book, the easiest way to transform black into red would be to overlay a layer of red and then set that layer to "lighten." But the Dude doesn't say we should do that. Nor does he explain why he did it. This scientist must believe in ESP, because he expects us to read his mind.

Are you wondering about all those boxes and numbers? The boxes and grey lettering indicate the areas where the text exists on an overlay of the reference COLB -- the one with the known provenance. The Dude uses this as a guide to tell him where the original text was.

(Which brings up an interesting question. Why would a forger not place his new text in the exact same locations as the old text? Doing so would be easy. You set a copy of the original as a top layer, set that layer to "multiply," then put the opacity at 10% or so. Instant guidelines.)

So: How, exactly, does the Dude think that we can extract the ghostly images of the original letters? Again, he is maddeningly unclear.

My guess is that his guess is that the presumed forger took a rectangular chunk of the background pattern, laid it over the text he wanted to cover up, then set that layer to "lighten." The black letters would disappear. (That's one way to do the job, although it's not the way I would have chosen.) But anti-aliasing means that small greyish pixels might surround the hole left by the vanished letters.

(Anti-aliasing is how computers smooth over the "jaggies" in text. See here.)

As I see it, we would get a workable number of those anti-aliasing pixels only if the background pattern were of a value closer to medium dark. (This is the proper use of the term "value.") The green background pattern is quite light.

Nevertheless, our fine Dude insists that -- after punching up the latent information in the white areas -- he is able to make those vanished letters re-appear. That is how he was able to determine that the original sex (in field 33, above) was FEMALE. Like so:

You may have noticed that we have a problem here. Techdude gives us a picture with nice, thick black letters: FEMALE. Obviously, this is not what cropped up when he played with the contrast slider.

That black FEMALE is not a latent image; it's an overlay. Those are the letters he thinks should be there.

Why doesn't he show what's really there? Why don't we see the actual results of his image manipulation? Techdude never gives us that image. Frankly, I find his reticence to show his results troubling -- so troubling, perhaps, as to call into question his integrity.

Now, I have tried to replicate his manipulations. I've done everything I can think of to squeeze information out of the same section of the COLB. Yes, I used the largest version of the COLB available. In the example below, I've tried my best to adjust color, brightness and contrast to match what the Dude has given unto us.

I tried to replicate what Techdude did in order see what can be seen sans overlays.

Here's my method: I used Photoshop. I reddened the black letters as described above. I then placed Techdude's "FEMALE" image (above) over my work as a guide. (Fortunately, he gave it at 100%) Then I adjusted the hue and brightness to match his results.

After much experimentation, I got my version of the COLB as close as possible to his by sliding the master hue control over to the -40 mark, then adjusting contrast +36 and brightness -45.

(See, Techdude? That's how you explain your work. Nice and clear. In this case, a child really could do it -- because I've explained how to do it.)

And here are the results. (For a larger, clearer picture, click on the image.)

Do you see the word FEMALE anywhere in there?

I don't.

Neither No Quarter nor Texas Darlin's site have dared to publish the actual results. I wonder why?

Sure, you can point to various pixel anomalies, but they exist throughout the image. The certainly do not add up to legible lettering.

You can also see pictures in clouds, in a textured ceiling or in a stucco wall. To quote the divine Groucho: "Oh, how you can get stucco."

I think that the people who are blindly applauding Techdude's work without replicating it are very much in danger of getting stucco.

Now, if anyone out there can come up with another technique which will bring out legible letters, then I will publish your work. That's a promise. You can't ask for fairer than that, can you?

Let's turn to the area where the name supposedly appears on the original birth certificate. This would be the signified by boxes 22 and 23 in Techdude's example. Today, we have the announcement that the name is that of Barack Obama's sister, Maya Kassandra Soetoro.

Look at the page giving this announcement. Do you see an explanation as to how the name was derived? No. Do you see an image? No.

We may get an image tomorrow -- but I'll betcha dollars to donuts that Techdude will give us an image with an overlay, telling us what we are supposed to see. The power of suggestion will do the rest of the work, for those committed to a predetermined outcome.

At any rate, here is what I was able to come up with.

Oh Maaaaaayaaaaa....! Where are you, Maya? (Again, click on the image for better quality.)

I suppose if you have a bad case of the "wanna believe its," you can scry her name in that urine-colored morass. And if you listen carefully to white noise, you may be able to hear Renata Tebaldi sing Carmen.

Some have claimed to have had success by overlaying 9 pt letters and fitting them over various parts of the image. It is said that "pixel anomalies" line up with the letters. This sort of procedure is unscientific and invites suggestibility.

Hamlet. Do you see yonder cloud that's almost in shape of a camel?

Polonius. By th' mass, and 'tis like a camel indeed.

Hamlet. Methinks it is like a weasel.

Polonius. It is back'd like a weasel.

Hamlet. Or like a whale.

Polonius. Very like a whale.

Again, I am perfectly willing to post your results if you can somehow squeeze legible letters out of all those ochre and greyish squares. My only request is that you accomplish this goal without overlays. If you use overlays, you will never persuade a skeptic.

In this case, I'll be overjoyed to eat crow if I have to. But I don't think I'll have to.

And what about Techdude's much ballyhooed "expert" status?

Hugh Trevor-Roper was one of the world's foremost authorities on Adolf Hitler, yet he screwed up royally when he vouched for the authenticity of the Hitler diaries. Dr. Bruce Maccabee had impeccable credentials as a photo-analyst for the Navy, yet he "verified" the Gulf Breeze UFO photographs, now known to be fakes.

I respect expertise as much as anyone else does. But science is a matter of replicable experiment. Reality is not determined by resumes.