NOTICES:



A WARNING: Those site visitors of Aboriginal and Torres Straight Islander Culture should be aware that there are photographs and images of the deceased.


AFIO DISCLAIMER:
The author of this blog is a member of the Association of Former Intelligence Officers and as such the views and opinions expressed here are those of the author and do not represent the views and opinions of the Association of Former Intelligence Officers, its staff or Directors.

Learn more about the Association including membership requirements at www.afio.com.


The Somerton Man Case. The body of a man found on an Australian beach close to a major Atomic Testing ground, he was probably poisoned, a copy of the Rubaiyat of Omar Khayyam and an unbroken Code page found and associated to him. Set against a Cold War background in 1948, was this man a spy? We think so and this blog focuses on the evidence that was left behind and in some cases missed, the Code page, Dry Cleaning numbers, A Poem and a small, torn piece of paper bearing the words TAMAM SHUD.

Saturday, 7 March 2015

Somerton Man: The Letter Q Continued



This post is really a response to a comment and some valid questions from Nick Pelling, you'll see his comments later in the post below. Effectively if I read it correctly Nick sees the micro letters and numbers but needs to know just how I actually got to them So here I will do my best to provide first of all a quick overview and will follow up when back in a good coverage area to do a second more detailed job

Before I start let me explain that I am out bush at the moment with some very 'iffy' internet and even worse mobile phone coverage so apologies in advance.

Here is a short but not complete list of items that I made use of:

1. Code page from the University Tamam Shud wiki, same dimensions as Nick's copy and at 400 DPI. I should mention that earlier attempts at delving deeper into the code page were frustrated by there only being a 72 DPI and then 96 DPI version available. I then took 3 copies of the 400 DPI page, one at 10% size reduction, a 15% reduced version and finally a 20% reduced version. I did this to provide some kind of comparison as I understand that by using a slightly reduced size you effectively will boost the DPI proportionately.

2. I made use of a 20 megapixel Olympus camera

3. I used a 28 Watt X 200 mm diameter LED natural white downlight inverted as a back light

4. I used a 20 Watt CFL ES fitting UV globe fitted into a standard desktop light.

5. I used a couple of different tripods, one small and one large. Important to keep the camera as steady as possible.

6. Also made use of natural daylight for some images not included at this stage.

7. I also made use of the 12 megapixel camera in my mobile which showed up some surprisingly good images.

In all over 4 years I took more than 4500 pics of the code page and individual letters from many different angles and using many different techniques including oblique lighting, various angled shots, back lit images and combination back lit and UV images. Relatively recently I read of how a US Library team had recovered an altered word from the Declaration of Independence which was at first thought to be a smudge. This team used a 32 Megapixel camera and an array of LED lights in different colours in their efforts. Not being the US Library of Congress and short on budget I used what was closest to hand and available.

I am not surprised at Nick's comment about not getting the same result, I made many attempts to get the required detail to show up without success so don't feel too bad about it Nick :)

Very quickly, and I will follow up with more detail, my first step was to make slight adjustments to the brightness and contrast of the whole 400 DPI page and the other 3.

Then using various lighting techniques including the LED backlight combined with UV light I took a series of images of many of the letters on the page. For our purposes I will focus on the Q which was part of a group as can be seen in the previous post.

The resultant images showed clear signs of letters and numbers that were still partially obscured by the overwritten markings of the Police.

I decided to remove the lighter shades of grey that covered the darker markings. The aim was to get those darker markings to show up more sharply and by removing the lighter shades of grey (10th March. Note, this is actually a colour replacement so th grey was replaced by lighter colouring) I was able to get the appearance that we see in the image. Nick makes a comment of a process that produces an artificial element which I don't agree with or could well have misunderstood. Nothing artificial was created, some colouring was removed and replaced with a lighter shade which served to sharpen the images of letters and numbers already observed. In fact in the past, Prof Abbott made comments that the images I had shown were too blurry and smudged and needed to be sharper to show details. That's what I have been working on.

As you can see I included parts of adjacent letters in the image, this was done to show that they too have underlying markings that have signs of letters and numbers and, in fact, that is correct because when they were similarly examined, letters and numbers show up quite well. In fact, I would say each letter on the code page contains micro written letters and/or numbers. I should explain that during the 4 years, I examined 100 or more samples of handwritten letters, I will discuss them in a later post.

I will have to sign off now, the connection is going from bad to worse. I hope this helps Nick and I will post the rest in a day or so.

Nick's Comment:
Gordon: working from the most reliable (1802 x 1440) images of the scan I know of,

If that's either the one I forwarded to you or the one from the Wiki and it's 400 DPI then we have the same base document

I can't get anything like the images you are showing here, no matter how I sharpen or contrast enhance the 'Q', particularly around the area you're focused upon.

Don't be embarrassed about it, it took me many attempts to find a way to properly reveal the darker markings beneath the Police write over marks. There are other things that you should do as per the earlier comments in this post above.

It seems as though you originally contrast-enhanced the image, visually annotated it (with boxes), printed it out onto paper, examined it under a magnifying glass, and them photographed the Q at a slight angle.

Not quite, a few extra steps involved as per the comments in the above post. Interesting that you mention the 'photograph at a slight angle' I took the picture in the earlier post from the same angle that I believe the micro writing was written in to the larger letters. Look carefully at it and you'll see they are left to right and in caps. Using caps was standard procedure for writing code certainly in WW2 and if you look at any of the Russian coded messages or the code names of those mentioned in Venona, they too are all in caps. the image appears the same whether you include the red box annotation or not, that was done to simply highlight each letter.  Be interested to hear your feedback on the slight angle issue.

 Is that an accurate description of what you did? Can you say (even roughly) what processing steps you took to get to that final image? It looks to me as though you have selectively removed bands of grey-scale levels to produce a set of artificially quantized edges, but it's hard to be completely sure.

The removal of lighter coloured shades was to bring what was already visible into greater relief. I wouldn't say it was to produce artificial edges, it was done to remove any blurring as per the Prof Abbott discussions. 



14 comments:

  1. Gordon: having spent many years doing digital image processing, I'm 95% sure that what you've done is artificially removed a set of grey levels from an image to try to yield a sharper-looking edge. But that's not actually digital sharpening, it's a quantizing digital manipulation that also has sharpening as a side-effect. I suspect that what your process is bringing to the fore is other stuff in the image (such as JPEG artefacts), rather than microwriting.

    I'll try to reproduce what you've done on the Q directly (i.e. without all the intermediate steps), hopefully it will become a little clearer then.

    ReplyDelete
    Replies
    1. Thanks Nik, Interesting, what you'll find is that no layers or levels of grey were removed, the process is known as colour changing so dark grey is turned to a lighter grey. Most importantly though is the fact that that part of the process took place only after the letters and numbers had been identified. By skipping the 'intermediate steps' as you put it you are missing out some very important parts of the process. It's also important to note that if you are testing the way that I have done this you should follow every step in line with normal academic practice. I fear that all you will be doing is to show what your process does and not disprove what mine does. A pointless exercise. I also feel that you need to revisit the issue of JPEG digital artifacts, this was dealt with some time ago and discredited as a possible cause, they simply don't behave in the way that you have described.

      If you are going to pursue your method I would respectfully suggest that you do it with another letter from the code otherwise you could be accused of creating an artificial result due to the use of my base image of the Q.

      Nonetheless, this is a free world and you are absolutely entitled to do whatever you wish and whilst I look forward to seeing your results it does seem to me, from what you have written, that you are predicting the result without first doing the work or taking the already defined steps, not sure that's altogether good practice.

      Delete
  2. Lossy compression - is that what we are talking about here?

    ReplyDelete
  3. Pete, JPEG images can suffer from loss. The JPEG artefact is observed in the pixelated image, sometimes a 'zigzag' effect but never has there been a case where this results in clear sets of letters and numbers. Nick has found a site somewhere that also discusses 'Quantization':
    Digital image quantization is the process of determining which parts of an image can be discarded or consolidated with minimal subjective loss. Image quantization is inherently lossy, however (i.e. the image quality is reduced due to the loss of some information).

    Colour quantization is a method of reducing the number of colours required to represent an image. For example, converting a photograph to GIF format requires the number of colours to be reduced to 256.

    In images scanners, quantization is the process of forcing similar grey levels to the same level as a result of limited tonal resolution. Most often apparent in the shadow areas of scanned images.

    None of which is applicable to the process I used, it being 95% use of light and oblique angles, very much a physical effort as opposed to a digital one. Nick has decided that he didn't need the 'interim' steps which, of course, includes the 95% of the effort and it is within those steps that the letters and numbers are seen. Perhaps Nick has tried it, saw the letters and numbers and hurried back to his leader for advice? I have little faith in Nick's claimed ability in this field.

    ReplyDelete
  4. GC, harking back to the rubaiyat - you say it had a dustcover and the code was written on that and impressed through to the back cover.
    I always thought that the rubaiyat that was handed in was a cheap soft cover, they are very small as you know (I have a Lothian edition 1944) Did they have dustcovers? It seems unlikely.

    ReplyDelete
    Replies
    1. If the book, as is generally believed, was a pocket edition from the 'Courage & Friendship' series published By W&T then the dust cover would have been in use. It's not what I have said, I have restated the views of others and in this rare case I tend to agree with DA., it had a cover. Unless you have other evidence of course?

      Delete
    2. Hi Gordon, What is the number of the verse that is commonly believed linked to the 'code' (moving finger/half line one) in the edition thought to be the one the SM had? Does it remain 51, like in other editions? thanks ~j

      Delete
    3. Hi Jenny, Here's a link to a copy of what is thought to be a book from the same series as that found and associated with the Somerton Man:
      https://drive.google.com/file/d/0B4FgVwct8qNUXzlIMVJYUlNkZmM/view?usp=sharing

      I think that it might be worthwhile to review the Marshall case and the page to which the ROK was opened when found. Then I would review the sequence of verses in that edition which, by the way, was a false imprint/unauthorised version.

      Delete
    4. You might be tired of hearing it.....but Thanks again! lol.... I appreciate all the help you offer, Gordon....

      Delete
    5. Hi Jenny, Thanks for your comment, I appreciate it. There will be some interesting new posts on this and other aspects in the next few weeks and they will be factual ones. The focus is on using known pieces of evidence and examining them in minute detail. What comes to mind is a phrase from 'Tuesday's With Maurie' where the then young student was asked repeatedly 'What do you see?' Keep looking hard, rest a while and then look again. I think you have a good approach. Thanks again for your kind comments.

      Delete
  5. I would rather see the Professor's evidence, Gordon.

    ReplyDelete
    Replies
    1. You could always ask him. I think that GF is of the same view. Then again no harm in questioning the authors.

      Delete
  6. Gordon, it would appear that nobody can be certain of the actual surface that was photographed to produce the published image of the code, some say it is a photograph of an overmarked photograph, others say it is a photograph of the rubaiyat's back dustcover after the code was overwritten. The rubaiyat's 'glossy' back cover.

    ReplyDelete
    Replies
    1. Hi Pete, I think you'll find that the original photograph was taken on glass plate, Police or other agency camera and equipment. In other words a high, evidence, quality image. It had been marked over and the photograph we all see was taken of that image by a freelance photographer who was working for the Advertiser at the time. Think in terms of an original photograph that was extremely detailed and then a really good newspaper camera used to photograph that. They were obviously not digital cameras and the images would not have been distorted in any way. What the camera saw is what you got in its finest detail.

      Delete