Compression algorithms are software programs that take advantage of redundancy in an image or data to reduce the size of the file and make it easier to store and transmit. There are many different such algorithms because different types of files are best compressed in different ways. Symbolic data (such as text, numbers or an executable program) compresses differently than image data, and has a greater need to be decompressed with no data loss. Color images compress differently from black and white images. Images compress differently from audio. Images and audio are far more tolerant to “lossy” compression techniques since they only need to accommodate the threshold of the human eye and ear.
Every single “anomaly” identified by the fleet of amateur Birther “image analysts” in an attempt to discredit the Obama long form is directly attributable to the compression algorithms applied to the PDF between scanning and publishing on-line. None of them are signs of forgery or fraud, and all of them are so obviously generated by a non-human process that it is sometimes difficult to credit even amateur sleuths with simply not noticing them.
A PDF (as opposed to a JPEG or most other pure “raster” images) is a complex file format, capable of containing raster, vector and symbolic data all within the same file. As such, it provided the software designers who invented it with both a compression challenge and a software design opportunity. The challenge was that a single compression algorithm would not work well for every PDF. The opportunity was not simply to allow different algorithms for different PDFs, but to also create a process that actually deconstructed even a “flat” PDF into different components that could use different compression algorithms at the same time.
The “layers” that amateur Birther “image analysts” have found in the long form PDF are artifacts created by this compression process. They are not the results of a human “forger” assembling a digital document which was then printed on paper, they are the result of a paper document being disassembled by a computer algorithm for easier storage and transmission on-line.
That this is what happened here can be clearly seen by looking at the details of the “layers” discovered in Adobe Illustrator. The color components of the image appear in one “layer” and the other large “layer” is completely black and white showing that the algorithm stripped the black and white components out of the total image for a different compression process.
Look at the “color” layer and notice the following details.
First… it contains almost every color component of the PDF. The other “large” layer is black and white… possessing not even shades of gray. (There are only five other tiny objects that were also stored in color… primarily the date and certification stamps… features that also were not identified by the compression process as perfectly black.)
And that’s the key for how features were separated. The very few “black” details that remain in the color layers are not actually black. Blowing up the detail to get a look at the individual pixels shows that they are composed of pixels of many different colors… with few of them being simply black. Examine for example the single digit of the Certificate number that was not stripped out of this layer. This is the last “1” in the number.
Whereas all the other digits are composed completely of pixels that are a single color black, this other digit (from the same number) is composed of many different shades of gray and green. It takes a lot more information to store that color image of the number “1” than it does to store the monochrome number “1s” that appears just four and five digits earlier in the same number. It was compressed in color, while the rest of the number was stripped out into a different layer that was compressed entirely (and far more efficiently) in black and white.
The second large layer is perfectly monochrome, consisting of pixels that are only white or only black. There are no shades of gray or hints of the green security image. Such an image is very easy to compress to a tiny fraction of its original size thus making it an obvious set of features for a computer program to strip out and treat as a different object with a different compression algorithm.
One of the ways we know these two “layers” were created by the compression process from an original single layer is that there is absolutely no overlap between the layers. Every single pixel in the black and white “layer” falls onto an empty (white) pixel on the color layer. In the color layer these features appear as white “shadows” where the black features were stripped out. If these were genuinely different layers created by a human forger, it would be expected that at least some of the black pixels should overlap green pixels from the security paper layer below. There are more than 2 million pixels in this PDF. Not a single one of them has two pixels overlapping.
Another way we can tell the compression process was performed by a computer and not a human forger is that the layers themselves do not even make sense from the perspective of human forgery. The final digit in the Certificate Number is the most obvious example of where objects are split between two different layers in a way that doesn’t make obvious sense. A more subtle but even more telling example is in the signatures. The entire signature of the Local Registrar appears on the color “layer” except for a single cursive letter “i” that appears on the black “layer.” A human forger would be expected to create (or extract) a signature as a single unit and them layer it onto the forged image. But who would ever forge a signature in multiple parts, placing every letter except one as part of one large color object, and then put a single letter of that signature in a completely different large black and white object?
Of course, they wouldn’t.
But a mindless computer algorithm instructed to strip out everything that looked perfectly black would show many such weird choices, unable to intelligently understand that the letter had anything to do with a larger thing called a “signature.” It stood alone. It was black. It became part of the black and white layer while the rest remained in color.
The bottom line is that the “layers” found by amateur Birther “image analysts” are completely unlike what would be expected from an actual forgery. They are simply the ordinary results of a scanned document being optimized as a digital file. Even very conservative sources such as WND, National Review Online and Fox News have reached that conclusion.
But here is where it gets fun… Miss Tickley’s “discovery” of identical pixel patterns across the birth certificate. How in God’s name did that happen?
Identical Pixel Patterns:
Among the reasons a black and white image can compress so much more efficiently than a color image is not merely because it only has to account for two colors; black and white. The vastly simplified image also provides opportunities to search for repetition and redundancy. If parts of the image are identical, the compression algorithm can store those multiple parts a single time rather than three or four or a dozen different times. An identical letter (for example) that is repeated 40 times can be stored in 1/40th of the file space as 40 identical letters stored 40 times.
So the black and white compression algorithm searches for objects or patterns on the image those are close enough to identical that they can then be tagged and stored that way.
This is called a “lossy” compression algorithm, because it actually does “lose” information in the effort to store it most efficiently. It counts on the fact that human eye would never have been able to tell those objects apart anyway, so the data lost is not meaningful. In the compression process it actually compares the objects (two check boxes for example, or two letters “T”) and determines that if they are close enough, they will be stored as two identical objects a single time.
As in the separation between “layers” already discussed above, the choices made by the computer are mindless and therefore often don’t make sense from a human perspective. Look for example at the check boxes identified by Miss Tickley.
Two of those check boxes are pixel for pixel identical. This could certainly be a result either of a human being cutting and pasting the same check box multiple times or a compression algorithm deciding the objects were close enough to make them identical for storage. But what then about that third check box? The third box is not identical, which doesn’t make sense if this was human forger creating the document from scratch. A human would most likely simply reuse the same image over and over… certainly use the same one to place three such boxes so close together on the form.
A mindless computer algorithm doesn’t even know what a “check box” is, and certainly doesn’t care about how close they might be together on a “form” that it also doesn’t understand. It did not recognize that third check box as being close enough to store them as identical. Certainly, to your eye and mine they look the same. But the computers do not do what we want them to do, they do what we tell them to do. And the algorithm we wrote told the computer that this box was different enough to be stored separately.
Again… the bottom line is not just that the (same) explanation of compression algorithms accounts for both the layers and the identical “objects” and letters found on the form, but that human forgery does not. No human forger would ever create the suite of characteristics seen in this PDF and “discovered” by our fleet of amateur sleuths. On the contrary, they are further evidence that the long form birth certificate released by president Obama is absolutely authentic.