Understanding the Digital Aspect of the Zodiac Letters – A Rebuttal to the Findings of Nanette Barto
Most of us researching the Zodiac case view the Zodiac letters mainly via two means. We can view the images in books, like Robert Graysmith’s Zodiac and Zodiac Unmasked. Many of us view them on our computers from sites like Tom Voigt’s ZodiacKiller.com. Not many of of us has had the opportunity to view the actual letters. So, in the transmission of these images from their original source to how they are viewed by our eyes, it is only natural that there are some issues that need to be considered.
When certain issues are not taken into serious consideration, misunderstanding and false conclusions are bound to be the result.
I have been, quite in vain, attempting to reason with Nanette Barto. I wanted her to take a closer look at her findings, taking into consideration the digital aspect of the images in question. These attempts have been a failure. She will just not accept the facts. While being able to communicate directly, instead of site to site, allowed for a better understanding between both parties, the end results were not very different.
I sent Ms. Barto a letter from the year 2000, hoping to prove once and for all that she is mistaken. I have done just that, but not to her. She refuses to acknowledge that there are simple explanations for the things she is seeing.
While she clings to her conspiracy theory, she does say that she has not accused anyone (at least in particular) of altering the letters. The implication, however, is there, in reports from the press and on Dennis Kaufman’s forum.
Her comparison of the two images, the 2000 letter, and the high resolution scan available at ZodiacKiller.com, led her to the following statements:
Ms. Barto writes: "This scan that you presented me with is an actual scan of the original document. I can tell this due to the buckling in the paper."
The older scan, of course is of the same image as the newer one, albeit different resolutions, file formats, color settings, and various other processes that make them appear different.
She continues: "What you also see is what we call pen pressure and how it varies in the stroke as it it executed. This causes points of the writing to appear lighter and others darker. When this happens the ink stroke is not continuous thus laying down a different amount throughout a single stroke."
Of course you can see the strokes on the new copy –it is a high quality image. The older copy is not.
Ms. Barto explains: "Line 7 word Shall, look at the back side of the first l and you see where there pen did not make a clean stroke. Look at every letter for lighter and darker spots, once you find them compare to (the newer scan)
Thick dark lines with no variance in stroke or pen pressure. That is because it was traced at a slow speed and the pen remained flat and in contact with the paper; this is unnatural for anyone to write this way and you proved it! Also, I made transparencies of these letters to overlay on the light table and in TV’s scans there is period dots at the end of every letter which a sign again of someone writing very slow with precision.
This explanation, in light of certain considerations that Ms. Barto chooses to ignore, is a complete misunderstanding of what is actually being seen. I tried very hard to explain, by getting down to the nitty gritty:
I wrote: I do not see how you can come to the conclusion that the 2000 letter, attributes below, depicts different (original) handwriting, when it is obviously a lower resolution copy. Both images depict the same letter, but by very different means. There has been no tracing. you don’t see the "dots" on the earlier image because it is a grayscale photocopy saved in a low quality file format. On the other hand, the newer image is a color scan, scanned digitally, and saved in a high quality file format.
Consider that the earlier image is:
591×836 px for a total of 494,076 pixels. Its a Compuserve GIF image of 149kb, bit depth = 8 and a 72×72 PPI
You are comparing the above letter, quite unfairly, to the high resolution scan that is now available:
1000×1417 px for a total of 1,417,000 pixels. It is a JPEG image of 611 kb, and a bit depth of 24, which, as most anyone knows, is a higher quality image. The PPI is 250×250
A JPEG image saves with little perceptible loss in image quality. A GIF image is severely limited by the 8 bit depth, which limits the colors that it is able to display. As I have pointed out in earlier communications, this is sometimes quite noticeable along angles and curves, where you seem to be finding what you believe to be "alterations."
Once again, the irregularities you are seeing are due to the limitations of a lower pixel resolution. By refusing to acknowledge that fact, you are making an unfair comparison, and depreciating your talent as a document examiner. The differences between tangible paper documents and computerized, digital images are immense, and those variables should be taken into serious consideration before presenting your findings. Nowhere in your findings have you taken into consideration the digital aspects of the letter images.
Barto responded by informing me that she can’t explain what I see.
She writes: It is a b/w scan u sent me. It is not due to low pixel resolution. The FBI can see it I’m surprised u can’t.
Does she not realize that there are even more color limitations within a black and white image? I don’t think so. it seems that she is implying that because the image is not in color, color limitations are irrelevant –an elementary mistake. So, you see how difficult this has been. The futility of using logic in trying to reason has become all to real.
Even though I tried to explain the differences in pixel resolution between images, I didn’t even approach the subject of display resolution. This adds even more variables to the equation, or perhaps inequation would be the proper term.
If the FBI truly accepts these findings, that does not bode well for the forensic capabilities of the agency. I’ll accept that they may be looking at these images. They might even humor Ms. Barto a bit. However, I strongly believe the FBI has come too far in their forensic capability to accept these findings to any serious degree.
Anyone interested in a quick review of digital imaging, is encouraged to follow the links within this post. A better understanding of the nature of these images can be quite enlightening, and perhaps clear up some of the confusion and myths surrounding this issue.
In addition to the above, here are some excerpt of prior emails sent by me, explaining resolution and what she is actually seeing:
I attempted to recreate the so-called alterations by normal, innocent means. I succeeded in this task. This was done by merely viewing the image at different zoom levels. While the photo copy effect I tried to explain earlier has been employed, it was actually a mere change in resolution that resulted in the change seen in the "a." The changes in the "a" were a completely benign irregularity that was the result of very simple and innocent action.
Try this: Look at the image you believe to be unaltered. View it at different resolutions. Notice that the more it is enlarged, the smoother it appears. Notice that the "a" in watch loses what you refer to as a hook as it is enlarged. While you may not be able to recreate what you perceive as an alteration exactly, we don’t know exactly how the original images were processed or enhanced. However, note that you can come close to "altering" the letter, as you refer to it, by merely changing the resolution at which it is viewed. A high res copy of a low res image will not improve it. A low res copy of a high res image will degrade it. There is nothing to indicate that these letters have been intentionally altered –they have only been saved with a different number of total pixels per inch, along with other enhancing effects that may make notable, albeit, innocent changes. A lower resolution save of a hi res image may create anomalies at odd places in the image. This is especially true in curved areas. If the pixels aren’t small enough, a curve can display the right angles at the pixel area, creating what looks like notches or "hooks" or other faults where the higher res image displayed curves.
Despite the letters being cleaned up, the individual characters have not been changed in any discernible means other than pixel resolution. As you should be aware, there is a huge difference between paper documents and digital images. Dramatic changes in digital images can appear quite innocently after digital effects are employed. Similarly, subtle changes can also appear. If you’ve ever selected pixels "by color" in an image, you will know that sometimes a small number of pixels can be selected in addition to those targeted. This could also explain some of the differences you are perceiving. I process all of the images on my website, including Zodiac letters, and always have. It is from this experience which I speak. It is my opinion that the changes you are seeing are due to resolution and "clean-up" processing. I do not believe there was nefarious intent involved. I have a copy of Graysmith’s book, Zodiac, from the 1980s, as well as the new copy published when "Zodiac" the movie was released in 2007. The photos of the letters are the same, specifically, the blotchy scans that were cleaned up when Zodiac Unmasked was released. I believe if there was an attempt at cover-up Graysmith would have updated the scans in both books. This he did not do. The cleaned-up version in Unmasked was presented as an improvement to the earlier "blotchy" scans in Zodiac, which still exist even in the newer editions.
As far as handwriting being a science, I can agree to some extent. However, the subjectiveness and difference among the handwriting experts in this case has been displayed to a degree that leaves me doubting. The Maupin allegations effect on the findings of experts proved the subjectiveness very well. Also, you have a lot invested in the results, which is surely bound to affect your perception in some way or another.
I keep asking myself why I am even bothering to send these (the older letters) to you. My only answer is that I hope it will serve as more proof that there is no conspiracy involved in the appearance of these letters and that the "alterations" are not intentional, not by devious design, but a result of differences among various resolutions and processed images.
Related posts:
