Xerox/WH PDF – Part 15 ECF/CM workflow

I had noticed the lack of /ObjStm in the documents on ECF/CM and I have come to the tentative conclusion that the software actually decompresses these artifacts as they are PDF 1.5 onwards.

Additionally we learn that

PDF 1.5 introduced a (optional) new form of XREF; a cross-reference stream rather than a cross-reference table. A PDF 1.5 file may contain one or both for backward compatibility.

If that’s the case then I believe the work flow fully explains the artifacts found by Hermitian in document 35.1

A document was produced using a Fujitsu ScanSnap #S1500 and imported into a PDF using Adobe Paper Capture. [NBC: Note… See Vicklund comments below] As shown by Vicklund this creates the hidden black rectangles and the lines as well as the OCR. As I had pointed out, the document however contains compressed Object Streams, not found in the 35.1 document. But I looked at all the Florida files and none contained /ObjStm objects, even though they were created with Adobe Paper Capture. I will extend the analysis to the Mississippi files now that I have updated my scripts to allow for batch processing.

The resulting document was uploaded to ECF/CM where at least the following steps were applied to the PDF

  1. A label was added to every page indicating the document number and the page number
  2. The document was made compatible with PDF 1.3 by uncompressing XREF and ObjStm streams into their separate objects

Without the low level tools, it would not be easy or not possible at all to make these observations and thus come to these conclusions. Once again, the low level tools, used for this kind of forensic investigation show the power of looking at the ‘raw data’.

115 thoughts on “Xerox/WH PDF – Part 15 ECF/CM workflow

  1. A document was produced using a Fujitsu ScanSnap #S1500 and imported into a PDF using Adobe Paper Capture.

    Please note my long-standing objection to the claim that it was “imported into a PDF using Adobe Paper Capture.” It is my belief that a PDF was created initially by the Fujitsu scanner, the file opened in Adobe Acrobat, and then OCRed using Paper Capture. IMO, that workflow better explains the Creator/Producer tags.

    However, making it compatible with an earlier version (whether purposely or accidental) does make sense.

  2. Let me update the workflow.
    Depending on when the original ECF software was developed, it indeed makes sense to remove such features as they only serve to improve compression.

  3. Pardon me, as I may have missed this, but exactly what is Hermie’s point? The docket entries are informational for the parties, allowing them to save time an money in copying and mailing. Any official, certified document would be filed directly at the court and a scanned copy put on the docket for informational reasons.

  4. Northland10

    I could not agree more. This entire discussion of Henry Blake’s stupid affidavit is a complete waste of time. I know it is probably fun to pick apart Henry’s conclusions that make no sense but the time could be better spent writing one concise, documented, and well illustrated article on the LFBC released by the White House showing all the details of the process. It should contain a point by point debunking of Garrett Papit and Mara Zebest’s works that were prepared for the Cold Case Posse.

    I want to point people to NBC’s blog but right now they would have to wade through about two dozen articles to get to the important findings. The important articles are now buried by at least 15 articles about an affidavit filed by Henry Blake that anyone with half a brain could read and figure out was complete nonsense. I knew it as soon as I read it. Why would Scott Tepper and Sam Begley create a new version of anything to put in a filing in MS? The only purpose of the MDEC filing was to introduce the Hawaii verification letter when Orly Taitz stupidly left the door wide open by filing a crappy copy of the LFBC in one of her voluminous and rambling exhibits.

    If the purpose of these articles is to embarrass Henry Blake I have news. He did that to himself a long time ago. Let Blake and his partners in delusion like Doug Vogt and Paul Irey continue to make fools of themselves. No one except maybe Taitz is paying attention and I doubt that she is any more.

    I enjoy reading NBC’s work as well as the comments here by Kevin and others. But really folks, let’s keep or focus on the big picture. Why not obtain some green security paper, print a copy of the AP JPG on it and run it through the process that we know was used to create the WH LFBC. Then take the results and publish them in a YouTube video. It would be devastating.

    NBC please take this as constructive criticism. You have done some excellent work. So have folks like Kevin and others. Henry Blake doesn’t have the ear of an elected albeit crazy sheriff of a large county. Mike Zullo does.

  5. I enjoy reading NBC’s work as well as the comments here by Kevin and others. But really folks, let’s keep or focus on the big picture. Why not obtain some green security paper, print a copy of the AP JPG on it and run it through the process that we know was used to create the WH LFBC. Then take the results and publish them in a YouTube video. It would be devastating.

    NBC please take this as constructive criticism. You have done some excellent work. So have folks like Kevin and others. Henry Blake doesn’t have the ear of an elected albeit crazy sheriff of a large county. Mike Zullo does.

    No offense taken. I just like to do research of anything minor or major🙂
    I am listening to your show right now and you raised the same issue.
    I will see what I can do here.

  6. The purpose of these postings are to explore the hypotheses and figure out a likely workflow based on the data, not necessarily based on our beliefs.

    I enjoy the challenge but I hear your points.

  7. NBC

    I understand perfectly. I am as guilty in getting caught up in debunking. I have spent more time than I should have at my blog responding to and picking apart Adrien Nash’s crazy theories about birth certificates and the Selective Service records. I noticed that my article referring to your work on the Zerox Work Centre has now scrolled off the front page at my blog. The Birther’s offer a cornucopia of things to debunk.

    One suggestion might be to write an article as a guest author on Obama Conspiracy Theories. Doc made the offer a while back for guest offers. I think an article summarizing the findings on the work flow would make a terrific one. I think Kevin Vicklund and others would be willing to review it and make suggestions as would I. Doc’s blog probably has the largest audience of any of out blogs.

  8. “NBC

    As I indicated previously, the trial scan made by NBC’s of a printout of the WH LFCOLB has several features not found in the WH LFCOLB PDF file.

    “See: A friendly reader has provided me with a scan on a Xerox WorkCentre 7535: WH LFBC Scanned Xerox 7535 WC

    “The document’s creator and producer are identical: Xerox WorkCentre 7535

    “The document contains 17 layers. The jpg again contains the comment string YCbCr”

    The most obvious is my observation that when opened in Adobe Illustrator CS6 or Adobe Ilustrator CC the composite image is rotated 90 degrees clockwise to landscape orientation. However, the Link panel indicates that the rotation angle of each of the 17 objects created by the Xerox Workcenter 7535 is zero. However the same PDF file opens in the correct letter position when opened in Adobe Acrobat XI Pro, Adobe Reader XI and PDFXchange Viewer Pro.

    Whereas the WH LFCOLB PDF image opens in the correct letter orientation in all of the above programs.

    A second observation is that the Xerox MRC algorithm failed to separate the text from the background as compared to the separation achieved in the WH LFCOLB PDF image. This strongly suggests that the WH LFCOLB was not created by this platform (at least not without human manipulation of the scanned image to increase the text separation by manually moving text from the background to the image layer.

    Strong evidence that this likely was done is the creation of a new, different type of image mask object in the Xerox scanned object list. I have decided to call these objects “skyline text image masks. These are associated with text that adheres to the background image layer. There are two different monochrome colors assigned to each of the four objects of this type that were created by the Xerox scan. One color is near-White and the other is dark Green mottled background pattern. These two colors are both assigned to the sky within these rectangular objects. The switching between the two colors is controlled by a clip group object. Each of the four objects is a rectangular “window” containing the “skyline” of background text in the bottom half of the window and sky above the profile of the text. The color of the sky is switched by the clip group for each of the four objects. When first opened in Adobe Illustrator CS6 the “sky color is near-white. When clip group object is switched off the sky changes color to the dark Green.

    Because of the complete absence of this new type of object in the original WH LFCOLB object list these objects would have to be manually removed from the background layer and placed in their appropriate location in the mostly text layer.

    An important initial conclusion is that these relocated objects may explain the source of the White Halos.

    The bottom line is that NBC has likely given up the very evidence needed to solve the riddle of the creation of the WH LFCOLB.

  9. Herm – “Whereas the WH LFCOLB PDF image opens in the correct letter orientation in all of the above programs.”

    You are ignoring the fact that the WH LFBC was saved as a pdf by Mac Preview. That corrects the orientation to portrait in Illustrator and shows each layer to have been rotated by 90 degrees.

    BTW, it you open President Obama’s 2010 taxes (created on a .Xerox WorkCentre 7655),

    http://www.whitehouse.gov/sites/default/files/rss_viewer/POTUS_taxes.pdf

    In Illustrator, they load in landscape mode with the layers for page 1 showing no rotation

  10. Whereas the WH LFCOLB PDF image opens in the correct letter orientation in all of the above programs.

    Which of course happens after you save the document using preview…
    Duhh.. You really should pay more attention.

    A second observation is that the Xerox MRC algorithm failed to separate the text from the background as compared to the separation achieved in the WH LFCOLB PDF image. This strongly suggests that the WH LFCOLB was not created by this platform (at least not without human manipulation of the scanned image to increase the text separation by manually moving text from the background to the image layer.

    Hermitian makes poorly supported assertions and forgets that we are not scanning the original but rather an already highly compressed document so these discrepancies can be best explained by these simple facts.

    These two colors are both assigned to the sky within these rectangular objects. The switching between the two colors is controlled by a clip group object. Each of the four objects is a rectangular “window” containing the “skyline” of background text in the bottom half of the window and sky above the profile of the text.

    And again our poor Hermitian is constrained by his limited tool sets.

    If you look at the raw data, the objects are quite similar.

    The bottom line is that NBC has likely given up the very evidence needed to solve the riddle of the creation of the WH LFCOLB.

    Yes, it’s called a Xerox Workcentre.

    It’s quite funny that you resist the inevitable my friend.
    If you only had done the research by looking at the raw data and you would have realized that your objections do not hold.

    Let me also point out to you that the embedded jpeg string found in Xerox Workcentre documents such as Obama’s WH LFBC is so far only found in document created by a similar workcentre.
    None of the many other jpegs I have collected show any sign of this string

    YCbCr

    The smoking gun that totally destroys the myth.

  11. You are ignoring the fact that the WH LFBC was saved as a pdf by Mac Preview. That corrects the orientation to portrait in Illustrator and shows each layer to have been rotated by 90 degrees.

    Of course, but our poor friend has forgotten the details of the work flow.

    One’s findings are only as good as the validity of one’s assumptions and poor Hermitian, lacking perhaps access to a Mac platform, has been unable to pursue this avenue.

    Sigh… I keep telling him to familiarize himself with the PDF code rather than rely on these high level editors when it comes to researching the origins of a document

    I have shown now how such an approach has revealed the workflow related to at least two documents

    1. The Whitehouse PDF which Hermitian still believes to be forged in spite of the clear evidence.
    2. Document 35-1 which was once considered by Hermitian to be ‘forged/

    All that research using Adobe Illustrator, all for nothing as it caused the ‘researchers’ to miss essential clues.

    I hope the have learned a valuable lesson here.

    But somehow I doubt it.

  12. One suggestion might be to write an article as a guest author on Obama Conspiracy Theories. Doc made the offer a while back for guest offers. I think an article summarizing the findings on the work flow would make a terrific one. I think Kevin Vicklund and others would be willing to review it and make suggestions as would I. Doc’s blog probably has the largest audience of any of out blogs.

    Excellent point. I will see what I can do here. Makes a lot of sense to provide a high level overview of the debunking of the ‘forgery of the WH PDF’

    So far the few people that have visited my blog and commented have now gone silent. One offered to bring my findings to Zullo but why should I help Zullo do his job…
    I’d rather see him defend the ‘work’ by the Posse in court, but sadly there will be no such opportunity.

    Many tell tale signs were missed… And people’s comments that they should look into MRC were largely ignored.

  13. This strongly suggests that the WH LFCOLB was not created by this platform (at least not without human manipulation of the scanned image to increase the text separation by manually moving text from the background to the image layer.

    Not only is there no logical reason for anyone to do this but worse, the background is the image layer. Or does Hermitian mean moving the text from the graphics layer to the text layer?

    And still leaving quite a few pieces behind? It just does not make sense and Hermitian will never explain why any forger would do this, while we have credible evidence that it was all software that caused these multiple layers that according to ‘researchers’ could never exist…

    Much to their dismay, they were shown to be wrong… Very wrong…

    Too bad that they appear too much guided more by their conclusions rather than by the data.

  14. “gorefan

    July 19, 2013 04:42

    Herm – “Whereas the WH LFCOLB PDF image opens in the correct letter orientation in all of the above programs.”

    You are ignoring the fact that the WH LFBC was saved as a pdf by Mac Preview. That corrects the orientation to portrait in Illustrator and shows each layer to have been rotated by 90 degrees.

    BTW, it you open President Obama’s 2010 taxes (created on a .Xerox WorkCentre 7655),

    http://www.whitehouse.gov/sites/default/files/rss_viewer/POTUS_taxes.pdf

    In Illustrator, they load in landscape mode with the layers for page 1 showing no rotation

    Well Gore! I know that you and Al invented the internet but when it comes to everything Obama I’m a strong believer in “TRUST BUT VERIFY”. I might have given your claim some weight but because NBC has already blamed it on the secretary placing the printout of the WH LFCOLB upside down on the glass then I remain a skeptic.

    You see, by now, it would have been sooo… easy to demonstrate your Preview fix then we’ll just have to continue to doubt.

    [NBC: And thus a real doubter would have done the experiments, just like we have and found that there was no reason for doubt]

    And then there is the problem that Preview would make other changes to the image.

    [NBC: Yes, changes you would not observe at the Illustrator level. But tell tale signs. I may have to introduce you to a 101 of PDF analysis… Note for instance that the image objects from the Xerox scanner are called XIObject and those from Preview Im0 etc… You could look at the same document and not realize that they are very different. With different compression, different object numbers, different embedded stream objects. All these details would remain totally lost to you.]

    So where’s the proof?

    [NBC: Well, you claim that it makes a difference, would it not be great if you could do the simple experiment yourself? You do realize that we already went down this path in our analyses?]

    Two wrongs don’t make it right. Now, thanks to you, I know that there is not just one but rather two images that exhibit this same problem. And you and NBC haven’t tested your fixes on either one.

    And we know for a fact that the WH LFCOLB opens in letter orientation in all the popular PDF software tools.

    [NBC: Yes because it was touched by Preview. That’s not too hard]

    Of course, to demonstrate either fix, one has to produce graphic images. That’s where NBC’s method fails because his freeware parser doesn’t render content streams to screen or print images.

    [NBC: Again, you are wrong. The NBC method shows how the document continues to remain rotated internally but this is corrected in a manner most viewers understand but Illustrator somehow messes up. Luckily after saving it through preview, the document looks just like the WH PDF]

    And now that Vicklund is off-line, NBC would have to use one of those dreaded Adobe programs to produce the required images.

    [NBC: Not really… But I am looking forward to you running the WH Tax document through preview and do this simple experiment yourself. That would be such a magnificent scoop if you could show that we are wrong… Say what Hermitian… Why did you not do the experiment? And I do not dread Adobe programs, they serve their purpose, just not when it comes to forensics. I understand the limitations of these tools.]

    And that would never do because it doesn’t fit his story line.

    [NBC: Show this to all of us my friend because so far you have done no experiments to back up your claims here. This is finally your chance to do something useful and present the experiments you claim to know the answer of already… Such poor research approach where you let your beliefs guide you and not real data.]

  15. NBC

    Let me also point out to you that the embedded jpeg string found in Xerox Workcentre documents such as Obama’s WH LFBC is so far only found in document created by a similar workcentre.
    None of the many other jpegs I have collected show any sign of this string

    YCbCr

    The most important two words in your comment are “so far”. You are talking to a very experienced Engineer on my end NBC. I know very well how far “so far” goes. Especially when it’s stated as a fact by a very sloppy novice.

    [NBC: Any scientist would understand and acknowledge the tentative nature of their findings. They would also never make claims that software cannot do something…]

  16. “NBC

    “This strongly suggests that the WH LFCOLB was not created by this platform (at least not without human manipulation of the scanned image to increase the text separation by manually moving text from the background to the image layer.

    Not only is there no logical reason for anyone to do this but worse, the background is the image layer. Or does Hermitian mean moving the text from the graphics layer to the text layer?

    Yup! I said it and I meant it. The proof is in the PDF that you provided here:

    “WH LFBC Scanned Xerox 7535 WC”

    Your friend’s scan is actually proof that the Xerox Workcenter could never produce the dergee of separation of text from background that the forger achieved by manipulation of the WH LFCOLB.

    [NBC: The scan proves nothing of the kind. It proves at best that a scan of a printout of a compressed document does not recover the same features as found in the original scan of the original object. Big difference.]

    “And still leaving quite a few pieces behind? It just does not make sense and Hermitian will never explain why any forger would do this, while we have credible evidence that it was all software that caused these multiple layers that according to ‘researchers’ could never exist…”

    The pieces left behind would be near-White background where the image box was selected and relocated to the text layer along with its “skyline text image masks”. These near-White text background blocks are scattered all over the background.

    [NBC: Skyline text image masks.. My goodness, can you speak in proper English to communicate your findings. What are the objects in question you are talking about.]

    Much to their dismay, they were shown to be wrong… Very wrong…

    I must have missed that part.

    [NBC: Obviously. But that is not my problem now is it?]

    Too bad that they appear too much guided more by their conclusions rather than by the data.”

    But it’s your data NBC and your conclusions! You don’t remember posting the 17 images which revealed that the Xerox Workcenter creates image mask objects that don’t exist within the WH LFCOLB PDF image? I immediately recognized that these objects are something entirely new. But you missed it completely, because you have no skill level using the high-level tools that break down the image into its component parts.

    [NBC: Of course there will be additional objects, the scan is of much lower quality… My goodness sakes. Is that your ‘argument’ now?]

    And then your defense is that these objects look the same at your microscopic code-line view.
    [NBC: I do not rely on high level tools manipulating the objects. I rely on the actual raw evidence]

    So here’s the bottom line. You fail to even come close to overcoming one of the main findings of the CCP’s experts. No software has “as yet” demonstrated the degree of separation of text from the background as exhibited in the WH LFCOLB PDF. And neither has your Xerox Workcenter overcome this hurdle.

    [NBC: I have shown how simple software can do what these ‘experts’ said it would/could never do: create more than the three layers found typically in MCR. As to the level of separation, any expert would realize that the experiment was limited as it relied on a printout of a highly compressed document. Funny how you cannot see the forest through the trees…]

    And it can never do so without the assistance of a human to manipulate it’s image.

    [NBC: That is an unsupported statement, just like the claim of forgery… Hilarious]

  17. “NBC

    “Hermitian”

    “July 19, 2013 12:20

    “Let me also point out to you that the embedded jpeg string found in Xerox Workcentre documents such as Obama’s WH LFBC is so far only found in document created by a similar workcentre.”

    None of the many other jpegs I have collected show any sign of this string

    YCbCr

    More smoke than gun here…

    So if you are correct that this JPEG tag is unique to the Xerox Workcenter 7535 then you have managed to provide solid proof that this scanner could never have created the WH LFCOLB PDF image.

    [NBC: Huh… I have shown that Xerox Work Centers software embeds the colorspace into the JPEG and that the same colorspace comment is found in the WH PDF and this somehow proves the opposite? You are not making much sense here.]

  18. “NBC

    And before you make some other wild claim about your Xerox forger, you should know that it took me about 5 seconds to discover these new objects with my code-level tools. A verified tool using verified query profiles.

    The tool identifies only these four new objects as “Black bitmap images set to knockout”

    The other 12 non-background images are normal “image mask” objects.

  19. “NBC

    The four unique objects discovered by me within the Xerox PDF scan is the smoking gun that disproves all of NBC’s conjecture about the Xerox forger.

  20. “NBC

    The four new unique objects that were created entirely by the Xerox Workcenter scan of a printout of the WH LFCOLB exist only because of the poor separation of text from the background image demonstrated by the Xerox MRC algorithm.

    We know that as a fact because we have proved that the new objects are associated only with the text that remains on the background image.

    The remedy for this necessarily requires the intervention of a human to manually move text box images from the background image to the mostly text image layer.

  21. Hermie – “Especially when it’s stated as a fact by a very sloppy novice.”

    Don’t get so down on yourself.

    Hermie – “And we know for a fact that the WH LFCOLB opens in letter orientation in all the popular PDF software tools.”

    As does the President’s 2010 taxes. Only in Illustrator does it open in landscape mode. Which is just like the pdf sent to NBC that was copied on a Xerox 7535.

    And guess what, Hermie?

    When opened in Mac Preview, it opens in portrait mode.

    And when you “save as PDF”, the saved pdf opens in portrait mode in Illustrator.

    other changes:

    The original 2010 tax return pdf was created with pdf version 1.5, but after saving in Preview the new pdf was created with pdf version 1.3.

    The 8-bit objects (there are 59 pages in the tax return pdf) use a “Filter[/FlateDecode/DCTDecode]” for the original pdf but this is changed to “Filter /DCTDecode ” for the pdf created by Preview.

    The 1-bit layers use a “Filter/JBIG2Decode” for the original pdf and “Filter /FlateDecode” for the pdf created by Preview.

    BTW, if you open page 1 of the original tax return in Illustrator, the 8-bit background layer contains the first “0” in “1040” and the “2” and last “0” in “2010”. And all the layers have 0 degrees rotation. But for the Preview created pdf, all the layers have been rotated 90 degrees.

  22. “Gorefan

    “BTW, if you open page 1 of the original tax return in Illustrator, the 8-bit background layer contains the first “0″ in “1040″ and the “2″ and last “0″ in “2010″. And all the layers have 0 degrees rotation. But for the Preview created pdf, all the layers have been rotated 90 degrees”

    But you haven’t demonstrated your claim by making images of the Xerox scanned PDF run through Preview and then opened within each of the software tools that I have identified.

    [NBC: We have shown how such a document, when run through preview once again opens as expected. You too can do the experiment as you have the WH Tax Form PDF and you can run it through preview, just like I and others have done.]

    And don’t use the excuse that you don’t have these tools because you won’t get any sympathy from me.

    What does your 90 degree rotation in Illustrator (after Preview) do to the image orientation in all the other tools?

    Words are cheap but facts are as hard as diamonds.

    Like your claim about Preview converting JBIG2 to DCTDecode — so are you claiming that Preview just changes the label of the filter from JBIG2 to DCTDecode or to the contrary Preview re-compresses the already compressed image. If the latter then Preview should warn the user that the PDF was already compressed.

    [NBC: Anyone familiar with PDFs would know that a conversion from JBIG2 to DCTDecode would not make sense. So one would have checked the original comment which in fact mentions FlateDecode… Big difference.]

    If the images are doubly compressed then you or NBC should check the image before and after the second compression with DCTDecode.

    I fundamentally avoid compressing a PDF more than once.

    [NBC: Who cares what you do… Software is not constrained by your work flow practices. Nor do you even know what the software does, deep down. ]

  23. “Gorefan

    “And guess what, Hermie?

    “When opened in Mac Preview, it opens in portrait mode.”

    I could care less how it opens in Preview. Apple has been mangling PDF files ever since they developed Postscript. I don’t trust any PDF file created on a MAC.

  24. Just an afterthought…

    I wonder what the double compression does to NBC’s “smoking gun” deeply buried JPEG label?

    [NBC: PDF 1.3 supports DCTDecode and thus nothing is double compressed. Yes, Xerox uses a flatedecode to compress a JPG which is a bit of an overkill but anyone knows that FlateDecode is lossless compression. Preview uncompresses the JPEG and then does not recompress it. The content remains the same. But these are good questions an interested researcher could have resolved himself, just like I did.]

  25. Like your claim about Preview converting JBIG2 to DCTDecode — so are you claiming that Preview just changes the label of the filter from JBIG2 to DCTDecode or to the contrary Preview re-compresses the already compressed image.

    No one has made that claim – it is JBIG2 to Flate, and it doesn’t affect the background image.

    I wonder what the double compression does to NBC’s “smoking gun” deeply buried JPEG label?

    Nothing, because only the background image is a JPEG, and thus wasn’t modified by Preview. No double compression required.

  26. he tool identifies only these four new objects as “Black bitmap images set to knockout”

    What you are seeing is the interpretation by the high level tool.

    So far you have done nothing more than detect a ‘new object’ but refuse to address the simple facts that

    Both the WH Doc and the Xerox Work Center doc contain
    1. Jpeg objects with YCbCr as an embedded comment field
    2. Multiple bitmap layers which ‘experts’ had claimed could never exist

    All you claim is some bitmap image set to knockout.

  27. I could care less how it opens in Preview. Apple has been mangling PDF files ever since they developed Postscript. I don’t trust any PDF file created on a MAC.

    But you should care because, as you may remember, the WH document was finally touched by Preview, something you awkwardly seems to have forgotten.

  28. I wonder what the double compression does to NBC’s “smoking gun” deeply buried JPEG label?

    Double compression?…

    The comment label is present in both the WH PDF and the Xerox PDF.

    Your point?… Do you need to do a real experiment? Have you even be able to see the comment ?

    Sigh…

  29. “NBC

    So you already forgot that your trial Xerox scan was never touched by Preview?

    Or maybe it was but the results didn’t fit your story line so you deep sixed them along with Vicklund’s Acrobat printouts of his hidden Strikeouts, Black Rectangles and Line elements.

    I wonder which typed text was struck by Vicklund’s fat Green rectangle?

    And you Obots are always crying about Zullo not showing all of his evidence.

    Talk about the “pot calling the kettle Black” !!!

  30. Like your claim about Preview converting JBIG2 to DCTDecode — so are you claiming that Preview just changes the label of the filter from JBIG2 to DCTDecode or to the contrary Preview re-compresses the already compressed image. If the latter then Preview should warn the user that the PDF was already compressed.

    Let’s try to educate our friend about PDF and how preview works.

    Remember that PDF version 1.3, which is what is produced by Preview, does not support JBIG2 encoding format.
    Preview however can read PDF’s from any version but when it creates a PDF document, it takes objects that cannot be represented as JBIG2 and recompresses them as FlateDecode not DCTDecode which would not make sense really.

    You too can do the simple experiment by opening the document created on the Xerox Work Centre and then do a print to PDF using preview.

    Preview has no reason to ‘warn the user’ that the document is recompressed, it is merely doing what the user asked. Take the PDF and print it as a PDF version 1.3 which, as you should have known, does not support JBIG2 format.

    What Hermitian fails to understand is that these compressions are lossless. JBIG2 is uncompressed and the objects are now represented as FlatDecode

    I fundamentally avoid compressing a PDF more than once.

    Which is why you have missed these tell tale signs, which can not be easily detected in a high level tool which would display the PDFs as being identical…

    And finally

    But you haven’t demonstrated your claim by making images of the Xerox scanned PDF run through Preview and then opened within each of the software tools that I have identified.

    We have shown that documents which are created on Xerox WorkCenter, once touched by preview, open correctly even though internally the images remain rotated.

    You can observe this by looking at the low level code which still performs the rotation using the coordinate transformation matrix (cm)

    So we have now again established that

    1. Our friend is unfamiliar with Mac Preview
    2. Our friend does not understand that it supports PDF 1.3 when creating PDFs
    3. Our friend does not understand that decompressing JBIG2 and recompressing it as FlateDecode produce the same results, but slightly larger file size
    4. Preview does not ‘warn’ the user when it is told to print a document to PDF
    5. Without these low level tools, Hermitian remains unable to verify and test his flawed hypotheses.

    Not bad for a few hours of effort.

  31. So if you are correct that this JPEG tag is unique to the Xerox Workcenter 7535 then you have managed to provide solid proof that this scanner could never have created the WH LFCOLB PDF image.

    Because…

    You make a pretty silly statement. I find that Xerox Workcentres produce JPEGS with an embedded comment identifying the colorspace and I find the same one in Obama’s PDF and you claim that this is solid proof that the Xerox Workcenter could not have created the LFCOLB WH PDF…

    You have just abandoned reason and logic my friend.

    Let me help you out:

    1. Xerox Work Centers embed a finger print in the JPEG
    2. So far these finger prints are only found in such Work Center created PDFs
    3. This finger print is found in the WH PDF.

    You do the logic… If you need some help finishing the logic, let me know. It’s reasonably straight forward and well known to many forensic researcher.

  32. The four unique objects discovered by me within the Xerox PDF scan is the smoking gun that disproves all of NBC’s conjecture about the Xerox forger.

    On the contrary. Let’s assume for the moment that you are correct, these objects do not undermine the findings so far, they just add another minor mystery that you seem to be unable to resolve.

    You have shown that a Scan of a print out of a highly compressed document does not look the same in all aspects and ignore how they are the same.

    Figures…

  33. We know that as a fact because we have proved that the new objects are associated only with the text that remains on the background image.

    You are still not making sense. You are saying that these objects had to be manually moved?… Fascinating how you mind does not allow for anything else than a forger even though I continue to show you to be wrong.

    It would help if you could identify these objects more accurately so that we can help you understand their origins.

    All the homework we have to do for you🙂 I guess that’s what happens when one lacks the proper tools for analysis.

  34. The 1-bit layers use a “Filter/JBIG2Decode” for the original pdf and “Filter /FlateDecode” for the pdf created by Preview.

    As expected Gorefan correctly identified the step but Hermitian failed to understand that JBIG2 would never be compressed using DCTDecode which is a lossy JPEG format.

    Such follies… These are basic mistakes my friend but I guess they can be excused by the fact that you may not be too familiar with the low level encoding of the PDF documents and instead rely on how readers treat the document.

    As such you have missed and continue to miss the tell tale signs found in the raw data.

  35. So you already forgot that your trial Xerox scan was never touched by Preview?

    On the contrary. The Xerox scan is the fundamental document and anyone with the proper tools can do the relevant experiments, as I have done and none of them show any changes in the findings.

    Your accusations are foolish and unsupported and the mere fact that you have to accuse me, once again, of lying, continues to affect only your own credibility as I and others have looked in depth at all the data.

    Somehow our researcher friend cannot even do the proper experiments… And yet he is trying to explain the artifacts found in a document created in preview but lacks the tools?

    Fascinating.

    If you are so certain that I am not showing ‘all the results’ because they do not fit my story line, then do the research.

    I have shown how people using limited high level tools have not been able to observe the full results.

    Poor Hermitian, now that his claims have been fully destroyed, he goes after the person.

    So predictable…

  36. The most important two words in your comment are “so far”. You are talking to a very experienced Engineer on my end NBC. I know very well how far “so far” goes. Especially when it’s stated as a fact by a very sloppy novice.

    Of course, as a scientist I do understand the limitations of my findings. But I have shown how all the data so far support my findings. The mere fact that these documents share an embedded comment in a JPG is highly relevant but of course in the end it was merely helpful in forming the hypothesis about the Xerox Work Center.

    Remarkably, the Work Center did something all these ‘experts’ said it could never do: Separate the document into more than three layers. Oops….

    Yes, I am a novice but not sloppy as I do not rely on high level tools that hide the underlying raw data but actually do all the experiments and report on them carefully.

    Sloppy are those who did not even look at the raw PDF data…
    That’s far worse than being a ‘novice’ my friend…

  37. ROTFL. Poor Hermitian, he loves to dig his own holes…

    Confirmation bias at its best, or just laziness…

  38. So you already forgot that he ran the trial scan through Preview?

    I find it fascinating how unable our friend is to do proper research into this.
    Lack of proper tools seem to prevent him to do the most relevant step: Creating the PDF using preview…

    He would have known through visual inspection of the actual raw data, that many objects are changed.

    Time to do what Hermitian failed to do: Present a more detailed analysis of the before and after, just for his benefits as he probably does not realize how much information can be gained from a PDF.

    Needless to say, one can create the same high level picture through an infinite amount of different raw PDF data, and you would not even know the difference at the high level of Adobe Illustrator.

    That’s why looking at the raw data is so important and a failure to do so causes one to miss an incredible amount of tell tale signs.

    And that, unfortunately, may lead to flawed conclusions…

  39. “W. Kevin Vicklund”

    July 18, 2013 18:51

    “A document was produced using a Fujitsu ScanSnap #S1500 and imported into a PDF using Adobe Paper Capture.

    “Please note my long-standing objection to the claim that it was “imported into a PDF using Adobe Paper Capture.” It is my belief that a PDF was created initially by the Fujitsu scanner, the file opened in Adobe Acrobat, and then OCRed using Paper Capture. IMO, that workflow better explains the Creator/Producer tags.”

    Balony ! Balogna ! You kept omitting any reference to Adobe Acrobat, repeatedly mentioning only the Adobe Acrobat 9.51 Paper Capture Plug-in as the producer of the 35-1.pdf document.

    [NBC: And what do you think a plugin plugs in to… Geez… It is however important to recognize that it was the paper capture plugin that was responsible]

    It was my repeated posts stating the fact that the Adobe Paper Capture Plug-in has been embedded in Acrobat beginning with version five that forced you to admit your deception.

    [NBC: No deception my friend, just poor comprehension skills. You believed that the document had been created by Acrobat.. What totally amazes me is that you have all the tools that would have allowed you to do Vicklund’s experiment and yet somehow it was Vicklund who had to uncover the evidence you had missed?]

    I also posted a link to a detailed article about the two different ways that Adobe acquires PDF files for OCR scan.

    [NBC: Yes, I understand you can post links but can you understand the actual work flow. You cannot scan directly into Acrobat using the S1500 which lacks a TWAIN interface]

    The differences between the two methods is explained here in detail for Acrobat 9.

    See: “Acrobat OCR: Make your scanned documents searchable”

    http://blogs.adobe.com/acrobat/acrobat_ocr_make_your_scanned/

    The important point (for the present debate) is that the OCR step is done “within” Acrobat for both methods. The action of the Paper Capture Plug-in is entirely internal to Acrobat.

    [NBC: The paper capture plugin defines the workflow. Acrobat is merely the API used to do its biddings.]

    But what you and NBC refuse to address is why would the secretary choose to OCR scan using Acrobat rather than using the far superior OCR capabilities of ABBYY PDF Fine Reader ScanSnap which is bundled with the scanner and which can OCR scan the image “on the fly” during the scan.

    [NBC: Begging the question. She probably did use the ABBYY reader or alternative created a PDF using the built-in Adobe PDF SDK and did not use ABBYY at all. Whatever she did, she used paper capture to import the data into a PDF and any OCR done before was overwritten by the plugin. You too could have done this experiment.]

    It makes no sense that she would use an inferior OCR product which requires her to scan to PDF file and then open the file in Acrobat to apply the OCR scan.

    [NBC: You are still missing the common work flow. Two documents were created and merged. Acrobat is quite a useful tool here. Common sense really… And even if you and I may not understand the exact reasons, the data show clearly what happened. The difference is that you let ignorance cause you to infer forgery when there is a far more logical and reasonable explanation especially since you have failed to explain what and how the forger really did here.]

  40. Strong evidence that this likely was done is the creation of a new, different type of image mask object in the Xerox scanned object list. I have decided to call these objects “skyline text image masks. These are associated with text that adheres to the background image layer. There are two different monochrome colors assigned to each of the four objects of this type that were created by the Xerox scan. One color is near-White and the other is dark Green mottled background pattern. These two colors are both assigned to the sky within these rectangular objects. The switching between the two colors is controlled by a clip group object. Each of the four objects is a rectangular “window” containing the “skyline” of background text in the bottom half of the window and sky above the profile of the text. The color of the sky is switched by the clip group for each of the four objects. When first opened in Adobe Illustrator CS6 the “sky color is near-white. When clip group object is switched off the sky changes color to the dark Green.

    What is happening here? Well, the small-print form text is touching the form lines, creating a dark frame around a small portion of the basketweave pattern. Hoiwever, since this is a scan of a printout of a low-res scan, the color in the framed area is nearly uniform. As a result, the scanner algorithm raises this light-colored area to a monochrome layer, possibly to fill out the layer count. (I notice that similar to the WH LFBC pdf, the number of monochrome layers is a power of two. I had previously predicted this possibility). The layering algorithm doesn’t actually check for text, but rather areas with very similar colors. The dark green mottled area is part of the background image, and is functionally equivalent to the white halos seen where the dark text is lifted to a monochrome layer. It’s a dark halo, if you will, made that color because the area outside the lifted area is very dark.

  41. The layering algorithm doesn’t actually check for text, but rather areas with very similar colors

    That is correct, it will separate as best as it can but it is only an algorithm. It seems that Hermitian is objecting that a print out of a low quality PDF when scanned in again, shows additional artifacts not found in the original PDF. But that’s to be expected. What he totally ignores is how the data show that multiple foreground layers have been created, something which according to some ‘experts’ could not have happened

    Now his argument has moved to: Well, it does but it does not do good enough of a job…

    Fascinating…

  42. Balony ! Balogna ! You kept omitting any reference to Adobe Acrobat, repeatedly mentioning only the Adobe Acrobat 9.51 Paper Capture Plug-in as the producer of the 35-1.pdf document.

    I see that Hermie has now descended into outright lies. I repeatedly stated that I was using Adobe Acrobat.

    It was my repeated posts stating the fact that the Adobe Paper Capture Plug-in has been embedded in Acrobat beginning with version five that forced you to admit your deception.

    No, it was your false declaration that I hadn’t mentioned using Acrobat that prompted me to quote all the times I had mentioned using it.

    Fuck off.

  43. I see that Hermie has now descended into outright lies. I repeatedly stated that I was using Adobe Acrobat.

    He is getting quite desperate and has now descended into the darkness which was to be expected when his worldview was destroyed… Anger towards those who poked so many holes in his claims.

    Makes perfect sense. We should show understanding to Hermitian in his troubled times of having to readjust…

  44. “NBC

    “July 19, 2013 18:43

    “The layering algorithm doesn’t actually check for text, but rather areas with very similar colors

    “That is correct, it will separate as best as it can but it is only an algorithm. It seems that Hermitian is objecting that a print out of a low quality PDF when scanned in again, shows additional artifacts not found in the original PDF.”

    When you’re just making up a story line as you go along like NBC is doing it’s hard to remember all the twists and turns along the way. I suggest that you re-read my “Analysis of Rectangular Objects here:

    http://www.scribd.com/doc/151738307/Analysis-of-Rectangular-Object-Boundaries

    And then explain to your readers how your magic Xerox MRC could set the rectangular object boundaries to always touch text pixels on the bottom and left sides of each rectangle without reading the text?

    Also, I don’t recall your showing that the rectangular object boundaries for your Xerox scan always touch pixels on two sides?

    “But that’s to be expected. What he totally ignores is how the data show that multiple foreground layers have been created, something which according to some ‘experts’ could not have happened”

    And four of the 17 layers (not nine) were a new type of object that is nowhere to be found in the WH LFCOLB PDF image. And your Xerox scan produced zero White speckle layers that were removed from the page 4 LFCOLB image from document 35-1.pdf. And the White Halos that are so prominent on the WH LFCOLB image have also been removed from the page 4 LFCOLB image.

    I guess you thought I wouldn’t notice.

  45. “WKV

    “What is happening here? Well, the small-print form text is touching the form lines, creating a dark frame around a small portion of the basketweave pattern. Hoiwever, since this is a scan of a printout of a low-res scan, the color in the framed area is nearly uniform.”

    So were’re back to the low resolution printout excuse. What’s the matter, can’t NBC’s friend push the 600 dpi button? After all — the text layers of the WH LFCOLB PDF image are 300 PPI.

    Maybe you Obots could explain why all the many versions of Obama’s LFCOLB are cruddy images?

    One would think that if the President of the United States is issuing his personal birth record that he would produce the highest quality image possible. Instead we have a steadily growing number of different images — all of poor quality. And the quality continues to drop.

  46. “[NBC: Begging the question. She probably did use the ABBYY reader or alternative created a PDF using the built-in Adobe PDF SDK and did not use ABBYY at all. Whatever she did, she used paper capture to import the data into a PDF and any OCR done before was overwritten by the plugin. You too could have done this experiment.]”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    I did do the experiment ACE and it was very definitive. I first ran an OCR scan on the WH LFCOLB PDF image with ABBYY PDF Transformer 3. The scan deciphered all of the printed and typed text. However the nine image layers were flattened into one. The text was placed in many separate layers below the flattened image layer. I then opened the scanned image in Adobe Acrobat XI and attempted to OCR scan the image. Acrobat responded with the warning message:

    “Acrobat could not perform recognition (OCR) on this page because This page contains renderable text.”

    So we are left with the secretary intentionally choosing to apply Acrobat OCR which is inferior to ABBYY.

    But that’s not the end of the story…See my next post.

  47. Just a brief clarification concerning the claim of NBC and WKV that the hidden Redaction Rectangles, Strikeouts and lines that I detected in the page 4 LFCOLB PDF image of court Document 35-1.pdf was entirely caused by an OCR scan of the scanned PDF image within Adobe Acrobat.

    There are three OCR modes provided by Acrobat’s Text Recognition Tool. The user chooses one of these for each application of OCR. The three modes are:

    1. Seachable Image
    2. Searchable Image (Exact)
    3. Clear Scan

    I’m not sure which earlier Acrobat versions had these same OCR scanning options but versions “X Pro” and “XI Pro” both have these same three modes.

    I did OCR trial scans with these two versions of Acrobat for each of the three scanning modes on the WH LFCOLB PDF image. The results indicate that only the first mode (Searchable Image) could possibly have created the added artifacts to the page 4 LFCOLB image. However, the results indicate that this scenario would be very unlikely.

    Mode 2 (Searchable Image (Exact)) returns the same nine image layers as in the WH LFCOLB image (and no more). The file size of the OCR’ed PDF image is 390 KB versus 380 KB for the WH LFCOLB image. The searchable text is invisible when the OCR scanned PDF image file is opened in Adobe Illustrator. The layer stacking sequence is unaltered by the OCR. The deciphered text is searchable and selectable by mouse and cursor in Acrobat. The suspect words can also be highlighted in Acrobat. The deciphered text is also searchable and selectable by mouse and cursor in Adobe Reader XI. All my attempts to make the searchable text visible in Adobe Illustrator CS6 or Adobe Illustrator CC have failed. Additionally, I applied Adobe Acrobat XI Pro File Compare tool to check for differences between the original WH LFCOLB PDF image and the OCR scanned WH LFCOLB PDF image. There were no detectable differences. The file compare report follows:

    Summary
    7/16/2013 5:42:34 PM
    No differences were found between documents.
    New Document:
    birth-certificate-long-form – Acrobat XI OCR Searchable Image (Exact)
    1 pages (389 KB)
    7/16/2013 5:42:31 PM
    Used to display results.
    Old Document:
    birth-certificate-long-form
    1 pages (378 KB)
    7/16/2013 5:42:31 PM
    No changed pages were found
    No pages were deleted
    How to read this report
    Highlight indicates a change.
    Deleted indicates deleted content.
    indicates pages were changed.
    indicates pages were moved.

    Application of mode 3 (ClearScan) to the WH LFCOLB PDF image flattened the nine layers into one and placed the deciphered characters, partial words and whole words in added layers over the flattened LFCOLB image layer. Consequently, we can rule out this method because the page 4 LFCOLB image includes no layers containing characters, partial words or whole words. The file size of the one-page LFCOLB PDF image after application of OCR mode 3 is 729 KB.

    The page 4 LFCOLB PDF image contains an altered and flattened image of the WH LFCOLB on the bottom of the layer stack and 21 added hidden geometric objects in image layers over the flattened image. These geometric objects include two broadline strikeouts, 12 line elements and seven Black Rectangles. There are no layers containing text in the page 4 PDF image.

    Finally, the OCR trials with mode 1 (Selectable Image) yielded a page 4 PDF file that is 753 KB in size as compared to 648 KB for the entire four-page Document 35-1.pdf. However, Black Rectangles and line elements were added above a flattened image of the WH LFCOLB PDF image. There were no broadline strikeouts added. The OCR scan added 16 line elements and 4 Black rectangles. Three of the four Black rectangles were the same width as the Green basket-weave background and together completely covered the entire background.

    The fourth Black Rectangle was a narrow strip with its vertical edges parallel to and straddling the left edge of the Green basket-weave background. The left vertical edge of this rectangle was just outside of the edge of the Green background and the right vertical edge of the rectangle was inboard of this same edge. The top edge of this rectangle is slightly below the midline of the page. The pixel dimensions of this fourth Black rectangle are W = 40 and H = 488.. There are no text or form lines within this narrow rectangle – just the Green basket-weave pattern of the background. This fourth rectangle is entirely covered by the middle full-width rectangle. The height of the middle rectangle is greater than the height of the certificate portion of the page. The pixel dimensions of the three full-width Black Rectangles are:

    Top = 2412 x 788

    Middle = 2412 x 1988

    Bottom = 2408 x 416

    The width of the background is slightly less at the bottom of the page. None of the seven Black Rectangles of the page 4 LFCOLB image are the full width of the Green background.

    Comparing the width of the outer White border between the OCR scanned LFCOLB PDF image and the page 4 LFCOLB reveals that the border of the OCR scanned image is narrower than that of the page 4 LFCOLB image. The page 4 image has more outer margin than the OCR scanned image. Consequently the overall size of the OCR scanned image is slightly greater than the page 4 image.

    The 16 line added line elements includes both horizontal and vertical lines. Each of the lines except for three coincide exactly with the form lines. The three exceptions are two vertical lines which are coincident with the left outer edge of the Green background and one which is coincident with the outer edge of the gutter region. Only one vertical line is coincident with a vertical grid line.

    Entirely missing in these OCR trials are any broad-line strikeouts. There are two hidden broad-line strikeouts in the page 4 LFCOLB PDF image of Document 35-1.pdf. One strikes out the four words “Kapiolani Maternity & Gyneccological”. The other strikes out the one word “August”. These two broad lines become visible by turning off the one Compound Clipping Path object. It is not necessary to first select these broad-line element to make then visible. These two broad-lines and the seven Black Rectangles are made visible by turning off the Compound Clipping Path object.

    The line widths of the other 12 line elements are too small to see the slight color change that occurs when the Compound Clipping Path is turned off. It is necessary to first select each line element and zoom in to see the color change. The effect is difficult to see because all the line elements are congruent with form lines.

  48. Kevin, trolls live to get people to get people angry and respond. Though some of Hermie’s response is out of ignorance and severe bias, he does engage in some trolling behavior. Don’t let him get to you, even on his outright lies

  49. I may not have not have been entirely clear in my question about Hermie’s point. I am not entirely sure why he is going on about the docket files, which are informational.

    That said, I don’t quite understand the necessity of going into the minutiae of the files with Hermie, either. However, if it’s what NBC and Kevin want to do right now, who am I to argue.

  50. Northland10

    I suppose after NBC and Kevin unlocked the secrets of the WH LFBC Henry Blake’s affidavit was the next logical target to go after. Somewhere Scott Tepper and Sam Begley are laughing their asses off at all the attention being paid to a simple process of putting a filing with exhibits together and getting it posted electronically on the court docket. This is something they have probably done hundreds of times.

  51. NBC: I have shown how simple software can do what these ‘experts’ said it would/could never do: create more than the three layers found typically in MCR. As to the level of separation, any expert would realize that the experiment was limited as it relied on a printout of a highly compressed document. Funny how you cannot see the forest through the trees…]

    “And it can never do so without the assistance of a human to manipulate it’s image.

    “[NBC: That is an unsupported statement, just like the claim of forgery… Hilarious]

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    You get no sympathy from me Ace ! You Obots are the only ones with access to Obama’s two sealed certified copies.

    So if you want to do all your trials starting from a crappy copy that some unknown person printed out on his ink jet then you’ve just got to live with the crappy results.

    And of course I will continue to compare your latest crappy results to the WH LFCOLB PDF image and point out to your readers where you have failed again.

    And, by the way, I don’t buy your “compressed document degraded image” excuse.

    You are fully aware that there is a high-resolution AP image that is still posted here on the obamafile.org.

    Kevin Davidson posted an article at obamaconspiracy.org which refers to this image as a copy of a reporter’s copy.

    http://www.obamaconspiracy.org/2011/06/long-form-artifacts-vanish-at-higher-resolution/

    However, because it’s a JPEG copy it would have to be first converted to PDF.

    The first PDF image of the Obama LFCOLB that was released on the internet was created by Scott Applewhite of the AP from an unsourced paper copy. He created his first PDF image at 9:00:38 AM on 04/27/2011, approximately 12-1/2 minutes after the start of the early morning press gaggle at the White House. This PDF was published the same day by the Iowa Muscatine Journal.

    See: http://muscatinejournal.com/pdf_6a633f26-70d9-11e0-8729-001cc4c002e0.html

    Applewhite then created another copy at 9:28:48 AM, also on 04/27/2011 that ABC published here:

    http://abcnews.go.com/images/Politics/ap_obama_certificate_dm_110427.pdf

    This one has been referred to as the “AP” copy.

    Each of these images has a uniform pixel size. Consequently, the background image is higher resolution than the WH LFCOLB PDF image.

    So what’s you next excuse?

  52. “NBC

    “[NBC: PDF 1.3 supports DCTDecode and thus nothing is double compressed. Yes, Xerox uses a flatedecode to compress a JPG which is a bit of an overkill but anyone knows that FlateDecode is lossless compression. Preview uncompresses the JPEG and then does not recompress it. The content remains the same. But these are good questions an interested researcher could have resolved himself, just like I did.]”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    It’s amazing how a little probing by me causes the Obot rats to toss a few new facts my way.

    Now we learn from NBC that this Xerox forger used FlateDecode to compress the background image rather than DCTDecode which was used to compress the background image of the WH LFCOLB.

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

    “Preview uncompresses the JPEG and then does not recompress it. The content remains the same.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    So you are claiming that Preview can decompress but it can’t compress ?? Wouldn’t that increase the file size ?

    The file wh-lfbc-scanned-xerox-7535-wc.pdf was created and produced by

    Xerox WorkCentre 7535

    Xerox WorkCentre 7535

    Therefore it was not produced by Preview.

    However, the code indicates that the background image was compressed/encoded by

    ZIP/Flate compression (FlateDecode)

    JPEG compression (DCTDecode)

    The other 16 objects were compressed by

    JBIG2 compression (JBIG2Decode)

    Thus we have the possibility of double compression of the JPEG image before any use of Preview.

    Time to get out the operator’s manual for the Xerox Workcenter 7535.

    P.S. I found four more objects that are associated with the background only. That makes a total of eight which are background objects. The WH LFCOLB has none.

  53. “NBC

    July 19, 2013 17:14

    “”We know that as a fact because we have proved that the new objects are associated only with the text that remains on the background image.””

    “You are still not making sense. You are saying that these objects had to be manually moved?… Fascinating how you mind does not allow for anything else than a forger even though I continue to show you to be wrong.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    There are actually eight objects that are associated with the background image. There are two different types — neither of which are found in the WH LFCOLB PDF image.
    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

    “It would help if you could identify these objects more accurately so that we can help you understand their origins.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    I obviously don’t need any help because I found the eight objects immediately after opening the Xerox scanned file and you still haven’t found any of them. I then confirmed that there are 17 objects total including the background JPEG listed in the PDF code.

    So if you search through the 16 non-background objects the probability that the first one picked is a background object is 50%. One would think that 1:2 would be adequate odds for the guy who produced the 17 images from his own PDF file.

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

    All the homework we have to do for you I guess that’s what happens when one lacks the proper tools for analysis.

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    Wrong Ace! That would be homework that I would be doing for you because it’s your work and you don’t have the tools to find your own eight objects.

    Just look for stuff that you have never seen in the WH LFCOLB.

  54. It’s amazing how a little probing by me causes the Obot rats to toss a few new facts my way.

    I am sorry that they are new to you. Anyone familiar with PDF should be familiar with these facts.

    No?…

    So far Hermitian has found nothing of relevance to counter these findings… The embedded string in the JPEG, the fact that Xerox Workcentre creates more than three layers…

    Yes, as I showed, the JPG from Xerox is encoded in DCTDecode (JPG) and then in lossless FlateDecode. Somewhat of an overkill.
    When you open the document in Preview and print as pdf, you will find that Preview cleans up the double compression.

    Simple really… It also renames many of the objects such as the bitmaps and reorders the objects.

    All things you missed by relying on Illustrator.

    You may want to realize that these objects related to the background are in fact objects that a MRC algorithm may recognize.

    It’s so trivially simple so why are you still so desperately hanging on to a conclusion of forgery where so far, none have been shown?

    We all predicted that MRC would explain the artifacts, and finally we identified the culprit…

    That’s great predictive and detective work… How are you doing on your ‘forger’

    No worries though, the judge in MS is likely going to ignore your affidavit as irrelevant and rule accordingly. It seems that the case is back on track for judgment now that Begley informed the court he wants to submit additional information, the court has woken up.

  55. I suppose after NBC and Kevin unlocked the secrets of the WH LFBC Henry Blake’s affidavit was the next logical target to go after.

    And the fact that Hermitian is Henry Blake adds quite a bit of fun. The poor soul has been embarrassing himself at Dr C’s long enough…

    Where people let their ignorance conclude a ‘designer’ is fair game to me. Whether it be creationism or accusations of a forger when the ‘researcher’ cannot fully explain the data, we see how an argument from ignorance becomes quickly vulnerable to real data. It’s a forger of the gaps argument where the gaps have become smaller and smaller.

    I still cannot believe how these ‘researchers’ relied on high level Acrobat tools to draw their conclusions, basically ignoring all the relevant information that can be found in the raw PDF.

    And there is a wealth of information to be found there… And in many cases, these data cannot be observed from Illustrator.

  56. That said, I don’t quite understand the necessity of going into the minutiae of the files with Hermie, either. However, if it’s what NBC and Kevin want to do right now, who am I to argue.

    It’s fun to educate people… Especially when they let ignorance rather than evidence guide them. Statements like “software never could do ‘x'” therefor it should have been the work of a forger are just to tempting…

  57. “RC

    “I knew it as soon as I read it. Why would Scott Tepper and Sam Begley create a new version of anything to put in a filing in MS?”

    RC the radio head had Scott Tepper as a guest on his show — Sam Begley was also a scheduled guest but was a “no show”.

    I’ve asked RC this question before…

    RC why didn’t you ask Tepper why he switched the page 4 LFCOLB for the WH LFCOLB?

  58. “NBC

    “No worries though, the judge in MS is likely going to ignore your affidavit as irrelevant and rule accordingly. It seems that the case is back on track for judgment now that Begley informed the court he wants to submit additional information, the court has woken up”

    Or else Begley has a snitch in the courthouse and he got wind that Judge Wingate’s ruling was imminent so he called him up to plead for more time.

    And Judge Wingate gave him a few days to “go through the motions” (pun intended).

    [NBC: I love it how blind Hermitian is to the actual data and evidence… Instead he prefers made up stories that lack any supporting evidence… I found much of the same trends in his research about the PDFs]

  59. Here’s an old post I found about print to PDF compression in MAC OS.

    See: https://discussions.apple.com/thread/3048661?start=0&tstart=0

    The screen shots don’t load but it’s pretty straightforward.

    ——–

    “j.ve

    “May 9, 2011 11:47 PM

    “The below tip applies to Mac OS X 10.5 and 10.6 (Leopard and Snow Leopard). I suspect it will also work with older OS X versions, and hopefully also with future updates. Please post comments if you have more insights into this.

    “As most users know, the majority (if not all) Mac OS X of applications support Print to PDF functionality. However, when producing PDF this way (ie by printing to Mac OS X’s native PDF spooling function), there is no user control over the image compression settings in the resulting PDF file (nor over any other file-size-related features). This can lead to very big, full-resolution PDF files.

    “The below describes how you can compress PDF files on a Mac without using any third party software (like Adobe Acrobat).

    “A. The Basics

    “1. Open the PDF you would like to compress/reduce in size in Preview

    “2. In Preview, select “Save as…” from the File menu.

    “3. In the resulting dialog window, select the Quartz Filter “Reduce File Size,” then save the file.

    “Note: You probably want to save this file under a different name or in a different location, unless you are comfortable with overwriting the original, full-resolution file.

    “That is all for the basics. The result will be a (very small) PDF file, with average JPEG compression, which has been reduced to at least 50% of its original resolution and no bigger than 512×128 pixels. This is very small indeed.”

    ———-

    So now we know that the Xerox Workcenter 7535 first compressed the JPEG with FlateDecode and then with DCTDecode. Then the scanned PDF was printed to PDF in MAC OS Preview. This step applied another compression to the PDF and reduced both the image resolution and the file size.

    Let’s hope that today’s Preview and MAC OS (Print to PDF compression algorithm) has been improved. Otherwise it doesn’t matter what the resolution of the original paper copy was.

    The reader should note that the file-size of the Xerox scanned PDF is only 253 KB whereas the size of the WH LFCOLB is 380 KB. And the Xerox PDF has 17 objects and the WH LFCOLB has only nine. And this PDF has not yet been printed to PDF from Preview.

    See: wh-lfbc-scanned-xerox-7535-wc.pdf

    Now this is what I would call sloppy research !

  60. Does anyone really thing that the CCP as part of their research would have overlooked which scanning software that departmental copiers/scanners would have most likely been used to create this document. I mean you think even a bad expert in computer graphics would not know that scanners can create PDFs in many different forms. I would think that as part of their research they tried to replicate the LFBC using every piece of office equipment they could get their hands on. Especially any that were known to produce layers.

  61. “NBC

    “And there is a wealth of information to be found there… And in many cases, these data cannot be observed from Illustrator”

    Who said that I only use Adobe Illustrator? I have real-time access to all the Adobe products on Adobe’s Creative Cloud and the full suite of Tracker Software products. I also use ABBYY PDF Transformer 3 for file compression and file type conversion. Corel Fusion is very useful for assembly of large PDF documents from many smaller documents with different formats.

    [NBC: And none of these offer any insight into the raw document. You rely on high level tools which do not provide you with the tell tale signs and the full understanding of the impact of the various steps. Using other high level editors is not very helpful when it comes to detecting forgeries or tracking workflow. We have seen countless examples of this by now.]

  62. Hermie; “This step applied another compression to the PDF and reduced both the image resolution and the file size.”

    When I used Preview on the 7535 pdf, the file size increased not decreased.

    7535 – 252.07 KB (258,116 Bytes)
    Preview 7535 – 295.83 KB (302,932 Bytes)

  63. “NBC

    “So far Hermitian has found nothing of relevance to counter these findings… The embedded string in the JPEG, the fact that Xerox Workcentre creates more than three layers…”

    Yes it created 17 layers whereas the WH LFCOLB has only nine layers.

    And the eight extra layers (consisting of two different types) are all associated with the background layer. Neither type is found in the WH LFCOLB PDF image.

    So do the math !

    17 – 8 = 9

    So where did the nine new extra layers go ?

    Hmmmmm… Maybe moved to the foreground layers?

    [NBC: You mean background layer… Sigh.. So confused. But yes they remained in the background layer as the MRC algorithm is not perfect. We already have seen this in the WH PDF where it failed to detect quite a few words.]

  64. Hermie: “Or else Begley has a snitch in the courthouse and he got wind that Judge Wingate’s ruling was imminent so he called him up to plead for more time.”

    Nope, the order from Judge Wingate is clear. The defense asked to submit additional authority and the judge allowed it. We also know exactly what Begley and Tepper submitted. It was a notice about the recent ruling in the Grinols case.

    From the order by Judge Wingate – “Recently, Samuel Begley, counsel for defendants, telephoned this office asking permission to submit additional authority.”

    From Begley’s letter to judge Wingate – “Pursuant to the Court’s recent text order in the above captioned cause we bring to your attention the Memorandum and Order of Dismissal entered in Grinols v. Electoral College, 2013 WL 2294885 (E.D. Cal. May 23, 2013), a copy of which is attached.”

  65. Hermie: “Yes it created 17 layers whereas the WH LFCOLB has only nine layers.”

    And the PDF created on the 7655 only had 5 layers.

    Two of the layers on the WH LFCOLB consisted of small white specks which barely visible on the computer screen and most likely invisible on a printout of the White House PDF.

  66. “Gorefan

    “July 20, 2013 21:34

    “Hermie; “This step applied another compression to the PDF and reduced both the image resolution and the file size.”

    “When I used Preview on the 7535 pdf, the file size increased not decreased.

    “7535 – 252.07 KB (258,116 Bytes)
    Preview 7535 – 295.83 KB (302,932 Bytes)”

    Replay for Gorefan

    ——–

    “j.ve

    “May 9, 2011 11:47 PM

    “The below tip applies to Mac OS X 10.5 and 10.6 (Leopard and Snow Leopard). I suspect it will also work with older OS X versions, and hopefully also with future updates. Please post comments if you have more insights into this.

    “As most users know, the majority (if not all) Mac OS X of applications support Print to PDF functionality. However, when producing PDF this way (ie by printing to Mac OS X’s native PDF spooling function), there is no user control over the image compression settings in the resulting PDF file (nor over any other file-size-related features). This can lead to very big, full-resolution PDF files.

    “The below describes how you can compress PDF files on a Mac without using any third party software (like Adobe Acrobat).

    “A. The Basics

    “1. Open the PDF you would like to compress/reduce in size in Preview

    “2. In Preview, select “Save as…” from the File menu.

    “3. In the resulting dialog window, select the Quartz Filter “Reduce File Size,” then save the file.

    “Note: You probably want to save this file under a different name or in a different location, unless you are comfortable with overwriting the original, full-resolution file.

    “That is all for the basics. The result will be a (very small) PDF file, with average JPEG compression, which has been reduced to at least 50% of its original resolution and no bigger than 512×128 pixels. This is very small indeed.”

    ———-

    Next time do step 3 once you have completed steps 1 and 2.

    [NBC: Why perform step 3… There is no evidence that this step was used… Duh… And how come you cannot do the experiment yourself… Such a researcher.]

  67. Hermie: “Yes it created 17 layers whereas the WH LFCOLB has only nine layers.”

    And the PDF created by the 7655 had only five layers.

    On the WH LFCOLB, two of the layers consist of small (1 or two pixels in width) white specks. They are barely visible on the computer screen and probably invisible on a print out of the PDF.

    The 7655 created layers that are very similar to the WH LFCOLB including the three date stamps, the registrar signature stamp.

    The 7535 also created layers of the date stamps, and parts of the registrar stamp.

    The fact that the three PDFs (WH LFCOLB, 7655 and 7535) were all created on different scanners and with different source documents could play a role.

  68. “gorefan

    “July 20, 2013 21:46

    “Hermie: “Yes it created 17 layers whereas the WH LFCOLB has only nine layers.”

    “And the PDF created on the 7655 only had 5 layers.

    Nope ! There are actually exactly 17.

    [NBC: learn to read… sigh]

    See: https://nativeborncitizen.wordpress.com/2013/07/01/pdf-scanned-on-xerox-workcentre-7535-part-2/

    Count them up !

    Maybe you can pick out the eight objects which are different from any of the nine objects of the WH LFCOLB.

    Apparently, NBC can’t find them.

    [NBC: They are clearly visible but they hardly disprove anything, as anyone familair with logic would understand. Of course different copies will result in different artifacts. Even between scans of the same document there may be differences in OCR accuracy. What Hermitian has failed to comprehend is that I have shown how there is software which creates multiple bitmap layers, something which some ‘experts’ had insisted could not be done and why therefor the PDF was a forgery.
    Not a very solid argument]

  69. Hermies: “Count them up !”

    The XEROX 7535 – 17 layers

    The XEROX 7655 – 5 layers

    Oh, and the “Save as ..”/ “Reduce File Size” creates a file with much of the text too blurry to read.

  70. “gorefan

    “July 20, 2013 22:58

    “Hermies: “Count them up !”

    “The XEROX 7535 – 17 layers

    “The XEROX 7655 – 5 layers

    Oh, and the “Save as ..”/ “Reduce File Size” creates a file with much of the text too blurry to read.

    Oops! Sorry — I didn’t notice the different Xerox model number in your earlier post. I don’t remember any results from a Xerox 7655 ???

    Also much of the small form text from the Xerox 7535 scan is too blurry to read.

    But I lost interest in the image quality way back when NBC reported that the first compression with the FlateDecode filter drops the image resolution to 72 PPI.

    Why do all the Obots only make crappy copies of the Obama LFCOLB?

    NBD: Why is it that the CCP and other ‘experts’ rely on crappy copies of the LFCOLB when making their claims of forgery? Remember that the original file size for an uncompressed 300 dpi document would be around 25Mbytes, compared the the few hundred kilo bytes for the final PDF. The forensic tools suggest that the quality setting of the JPEG compression was set to about 50. This is a significant reduction in quality, and can be easily observed by looking at the clearly visible JPEG artifacts. I am sure that you can repeat the experiment with a better copy? Looking forward to you doing some real research…

  71. Hermie: “I don’t remember any results from a Xerox 7655 ???”

    https://nativeborncitizen.wordpress.com/2013/06/24/the-proof-forgery-done-by-xerox-workcentre-7655/

    Hermie: “Also much of the small form text from the Xerox 7535 scan is too blurry to read.”

    Not surprisingly, it looks like all of the text and vertical and horizontal lines on the 8-bit background layer.

    Hermie: “But I lost interest in the image quality”

    The resulting image has to at least legible.

    Hermie: “Why do all the Obots only make crappy copies of the Obama LFCOLB?”

    Tim Selaty Jr. in his report to the Cold Case Posse said that using the reduce file size options resulted in unreadable text.

    It appears that in Preview only the “File/Print/Save as PDF” option creates a file that is legible.

  72. “Gorefan

    “Two of the layers on the WH LFCOLB consisted of small white specks which barely visible on the computer screen and most likely invisible on a printout of the White House PDF.

    There are four “skyline text image mask” objects and four Green-splotch image mask objects found within the objects of the Xerox 7535 scanned LFCOLB image. These are new type objects that are not found in the WH LFCOLB PDF image. Both types are associated with the background image and not to the foreground text layers.

    Curiously, the four green-splotch background image mask objects (in the Xerox scan) are all located near the top edge of the page. These appear to correct defects in the Green basket-weave pattern of the background. However, only missing green bars are patched or replaced.

    The White speckle layers in the WH LFCOLB image are above the background layer in the layer stacking sequence. Hence, the two speckle objects are on top of the background image. This can be easily verified by selecting the object box of one of the White Speckle objects and dragging it to a new position on the background. The speckles move with the box. However, the white speckles are not part of the background image. Therefore the White speckle objects are associated only with the White speckles contained within the box.

    The four green-splotch objects are also located above the background image in the layer stack. The Green bars also move with the box. However, the difference is that the green bars are part of the background image. Hence when the object box is moved, the green bars move away from their initial positions that are consistent with the pattern to relocated positions that are not consistent with the pattern. It follows that, the green-splotch objects are associated with specific Green bars within basket-weave pattern. If the Green bars are relocated to other image layers, then they must be carefully placed over their initial positions.

    These subtle differences cannot be detected by looking at lines of code with a parser. These image mask objects can only be properly classified within a sophisticated vector graphics program such as Adobe Illustrator or InkScape.

  73. And, by the way, I don’t buy your “compressed document degraded image” excuse.

    It’s not an excuse, it is a trivial observation to anyone familiar with JPEG compression. The estimated compression in quality is about 50%. Which is not surprising as you can observe many jpeg artifacts in the picture.

    It’s not what you are willing to ‘buy’, it’s what the facts tell us. And I assume that you are familiar with how one may estimate the level of compression used in JPEG?

    Common sense…

  74. Why do all the Obots only make crappy copies of the Obama LFCOLB?

    NBD: Why is it that the CCP and other ‘experts’ rely on crappy copies of the LFCOLB when making their claims of forgery? Remember that the original file size for an uncompressed 300 dpi document would be around 25Mbytes, compared the the few hundred kilo bytes for the final PDF. The forensic tools suggest that the quality setting of the JPEG compression was set to about 50. This is a significant reduction in quality, and can be easily observed by looking at the clearly visible JPEG artifacts. I am sure that you can repeat the experiment with a better copy? Looking forward to you doing some real research…

    The simple facts are that this quick experiment showed that contrary to claims by so called ‘experts’ there are copiers which can produce multiple layers of bitmaps. Thus it destroys a major argument. Of course, Hermitian has now to point to some layers in the Xerox copy not found in the WH PDF as if that somehow undermines the facts.

    Anyone familiar with copiers and scanning would realize that the actual breakdown by MRC software is sensitive to the quality of the original.

    And Hermitian still has failed to explain the embedded comment in the jpeg.

  75. When I used Preview on the 7535 pdf, the file size increased not decreased.

    Of course… Anyone with a bit of logic could have predicted this. No JBIG2 compression and no Flatedecode of DCTDecode. Although the latter should not make much of a difference.

    Why is it that people find it so hard to do these simple research steps?

  76. . I would think that as part of their research they tried to replicate the LFBC using every piece of office equipment they could get their hands on. Especially any that were known to produce layers.

    What research :=) … They were informed early on about MRC compression but decided to downplay and consequently ignore it.

  77. There are four “skyline text image mask” objects and four Green-splotch image mask objects found within the objects of the Xerox 7535 scanned LFCOLB image. These are new type objects that are not found in the WH LFCOLB PDF image. Both types are associated with the background image and not to the foreground text layers.

    That does not make sense. All objects are in the background image layer until they are separated by the MRC algorithm. The MRC algorithm does not separate text perse but anything that is significantly different from the background itself and relies on many factors to make such a determination. For instance, it seems that when an object is touching an object that has remained in the background layer, it too will remain in said layer.
    That the document shows additional objects not found in the Original pdf does not mean that they existed in the original PDF. That is just a somewhat foolish position. Remember that these parts are scanned in half the resolution for the background layer, making it impossible to move them to the text layer, similarly moving the text layer to the background layer would be detectable in the background jpeg. No evidence of such was found.

    So the most likely explanation is that the document I used was not a perfect imitation of the WH Document, which of course makes sense. That document was likely scanned at 600 or 1200 DPI and then had MRC applied and was subsequently downsampled and the text was decoded in JBIG2 and the background in lossy JPEG. So the background not only had significant downsampling applied but alos JPEG compression which reduced the quality to 50 % as estimated by algorithms that can be used to estimate the level of compression.

    Hermetian still refuses to accept that I have shown how contrary to ‘experts’ opinion, there can be multiple text layers as a result of MRC compression and thus I have disproven a major foundation. Furthermore, our friend is still ignoring the embedded comment in the jpegs which I have found consistently in Xerox Workcenters but nowhere else (so far).

  78. “I knew it as soon as I read it. Why would Scott Tepper and Sam Begley create a new version of anything to put in a filing in MS?”

    Let’s look at the history and you too may understand it.

    1. Orly submitted, in Orly fashion, an illegible copy of the Long Form Birth Certificate
    2. Begley submitted a long form birth certificate (15.1_
    3. The document was subsequently printed out and sent to Hawaii to have them certify the document
    4. Hawaii responded that they looked at the WH PDF and the 15-1 PDF which had been printed out and verified the information
    5. The letter and the documents were sent to the chambers of the Magistrate Judge and Orly et al
    6. The document was subsequently scanned into the court records.

    A logical work flow but somehow Hermitian let his lack of imagination conclude ‘forgery’ because he could not understand why someone may follow such a workflow. Note that there was no evidence of manipulation found in the PDF beyond the simple process of scanning it using a Fujitsu #S1500 and importing it into Adobe using Paper Capture plugin.

    As we have shown, the process introduces similar artifacts to the ones observed in document 35-1. No need for a forger.

  79. So were’re back to the low resolution printout excuse. What’s the matter, can’t NBC’s friend push the 600 dpi button? After all — the text layers of the WH LFCOLB PDF image are 300 PPI.

    Incorrect. The foreground layers are 300 DPI, the background is 150 DPI and highly compressed JPEG which can be seen from the various artifacts created.

    Come on Hermitian, get some of your facts right… Printing a 600 DPI when the original is 300 and 150 DPI is not going to make any difference.

    Geez…

  80. One would think that if the President of the United States is issuing his personal birth record that he would produce the highest quality image possible. Instead we have a steadily growing number of different images — all of poor quality. And the quality continues to drop.

    Not if you want to put it up online. I am sure that you would appreciate what the size of an uncompressed TIFF file would be? Instead the staffer used a common workflow on the scanner which generated the document.
    Again, the common workflow explains what happened.

    Stop blaming others for your own follies.

  81. Hermitian: And then explain to your readers how your magic Xerox MRC could set the rectangular object boundaries to always touch text pixels on the bottom and left sides of each rectangle without reading the text?

    Quite simple, it does not read the text, it tracks the pixels. And yes, I document this fact in quite some detail when I showed how this is easily explained by a simple algorithm and not by a forger.

    Did you somehow forget?

    Let me educate my friend a bit more about how a scanner works. The picture is scanned by capturing scan lines while an optical device moves over it. The scan direction creates individual scan lines and they are stacked to form a picture.

    After scanning the MCR algorithm finds blocks of 8×8 that appear to be separate from the background and places the monochrome pixels into a bitmap. Once i realizes that no more data is to be included, it has remembered the last scanline and offset where it found a matching monochrome pixel and cuts it off at this one.
    Thus at the top and the right, the images align at 8 bit boundaries and at the bottom and the left they align at the edges.

    It’s a simple algorithm and you too could have looked it up to understand how it may work. Remember: the goal is to compress the document by separating foreground from background, allowing them to be compressed differently.

    Thus blocks are optimized by aligning them with the bottom and left hand side.

    No need to ‘read the text’ just look at the pixels of the scanline.

    This is not rocket science.

  82. “NBC

    “July 21, 2013 01:49

    “”So were’re back to the low resolution printout excuse. What’s the matter, can’t NBC’s friend push the 600 dpi button? After all — the text layers of the WH LFCOLB PDF image are 300 PPI.””

    Incorrect. The foreground layers are 300 DPI, the background is 150 DPI and highly compressed JPEG which can be seen from the various artifacts created.

    Come on Hermitian, get some of your facts right… Printing a 600 DPI when the original is 300 and 150 DPI is not going to make any difference.

    Wrong again Doofus ! The resolution of the layers of the WH LFCOLB PDF image are 150 x 150 pixels/inch for the background layer and 300 x 300 pixels/inch for the non-background layers. These values are as measured with Adobe Illustrator CS6 with the grid set at 150 lpi or 300 lpi.

    [NBC: That’s what I said.. 150 DPI for the background layer and 300 DPI for the foreground. Who is the real dufus now. And printing a 150 DPI document at 600 or 1200 DPI is not going to improve the document. Geez my friend this is basic stuff]

    A decent quality print of the eight non-background images would be printed at 1200 dpi.

    So 600 dpi would be about right for the background image.

    [NBC: You are hilarious in your ignorance. A highly compressed JPEG is not going to look better at a higher DPI, as it will be interpolated…]

    Then there would be 16 dots/pixel in the printouts. And the colors would be close to the original.

    See: “DPI and PPI Explained”

    http://www.andrewdaceyphotography.com/articles/dpi/

    Be sure and read the last paragraph.

    Then call up your anonymous friend and tell him to print out the WH LFCOLB PDF image at 1200 dpi before scanning the printout with the Xerox Workcenter.

    You badly need a good PDF editor with an accurate grid.

    [NBC: Printing out a 150 DPI document at 1200 DPI still leaves it at a 150 DPI document. Duh…]

  83. “NBC

    “July 21, 2013 01:28

    “”When I used Preview on the 7535 pdf, the file size increased not decreased.””

    “Of course… Anyone with a bit of logic could have predicted this. No JBIG2 compression and no Flatedecode of DCTDecode. Although the latter should not make much of a difference.”

    The post you were responding was Gorefan’s — Not mine !

    NBC: Yes, I know, he was surprised by your findings… Of course a document that was compressed using JBIG2 would be larger when saved as a PDF 1.3 compliant document. That’s self evident

    I verified that the compression filters applied by the Xerox Workcenter 7535 were first FlateDecode and then DCTDecode. So the image was already doubly compressed before it was opened in Preview to strip out the METADATA.

    NBC: Yes, I showed the same data from my raw data dump. I have seen quite often a double compression, DCTDecode is a lossy compression, FlateDecode is a non-lossy compression. Since JPEG is already quite compressed, additional compression may not be significant. But these are dumb algorithms.

    Preview will uncompress and display the JPEG and when printing to PDF it will only output the JPEG using DCTDecode.

    You first said this about what Preview does:

    “[NBC: PDF 1.3 supports DCTDecode and thus nothing is double compressed. Yes, Xerox uses a flatedecode to compress a JPG which is a bit of an overkill but anyone knows that FlateDecode is lossless compression. Preview uncompresses the JPEG and then does not recompress it. The content remains the same. But these are good questions an interested researcher could have resolved himself, just like I did.]”

    Next you posted this:

    “Nothing, because only the background image is a JPEG, and thus wasn’t modified by Preview. No double compression required.”

    NBC It did not modify the jpeg, it merely uncompressed the FlateDecode object and rendered the jpeg. When it saved it, it did not modify the background image. you seem to be confused about the image versus how it is stored. That’s like saying that a tar.gz contains different resources than at tar file itself. When decompressed both lead to the same end product.

    And then:
    Preview has no reason to ‘warn the user’ that the document is recompressed, it is merely doing what the user asked.”

    And next you bloviated:

    “[NBC: PDF 1.3 supports DCTDecode and thus nothing is double compressed. Yes, Xerox uses a flatedecode to compress a JPG which is a bit of an overkill but anyone knows that FlateDecode is lossless compression. Preview uncompresses the JPEG and then does not recompress it. The content remains the same.

    And finally:

    “Yes, as I showed, the JPG from Xerox is encoded in DCTDecode (JPG) and then in lossless FlateDecode. Somewhat of an overkill.
    When you open the document in Preview and print as pdf, you will find that Preview cleans up the double compression.”

    Of course if you were right then the DCTDecode would reverse the compression of the FlateDecode and the file size would not be 253 KB but much larger. But I have verified that the DCTDecode was applied after the FlateDecode.

    Nope, DCTdecode is applied before FlateDecode. Which is of course simple to understand since the DCTDecode data contains the JPEG and the FlateDecode is a lossless compression.

    According to you then Preview would have to uncompress the DCTDecode compression because it is the last compression step applied by the Xerox.

    NBC Preview will just ‘render DCTDecode into a bitmap format but when it saves it, it takes the JPEG/DCTDecode data and just stores it in the PDF without applying FlateDecode

    However that doesn’t agree with the DCTDecode compression call out in the WH LFCOLB PDF for the background JPEG. No PDF translator that I know of would place the DCTDecode compression label for an actual decompression applied by the translator.

    NBC Huh, could you translate your previous paragraph into proper English. Let me help you understand what happened:

    MRC creates a JPEG encoded layer which is saved by Xerox using FlateDecode. A bit of an overkill as JPEG itself does not lend itself to ZIP-like based compression. The PDF is then read into preview which uncompressed the FlateDecode object and finds a DCTDecode object which it renders using the JPEG renderer. When it is printed to PDF, it takes the JPG/DCTDecoded object and just stores it as PDF object without re-zipping it using FlateDecode. This is how Preview renders JPEG, no need for another compression step that does little other than add to the rendering CPU usage and time.

  84. You forgot — I wrote the book on the placement of the object boundaries.

    ROTFL… You are so funny… It was gsgs who ‘wrote the book’ and you added little relevance to the analysis.

    For the longest time you could not do the proper analysis yourself, remember…

    The likelihood that a forger would carefully align these blocks at 8 bit boundaries is quite small. An algorithm makes far more sense. But we already know that the document shows all the signs of having been ‘forged’ by Xeroc work center. The embedded comment, the fact that it generates multiple foreground layers, the fact that it does JBIG2 compression, all align with the proposed work flow which was simple and straightforward.

    That some people are still worrying about minor details is fascinating as they allow their ignorance to lead to a conclusion of ‘forgery’ without making any logical case for it. That my friend is a classic argument from ignorance: I do not know or understand how this could have happened therefor I will conclude that it was ‘designed’/’forged’ where forgery now becomes a placeholder for your ignorance.

    In order to show that your forgery ‘explanation’ is a better one you need to show a logical workflow that explains all the facts. So far, noone has explained why the forger did what he did… Imperfect separation of the background into the text layers, 150 DPI background layers versus 300 DPI foreground, JBIG2 like compression leading to identical characters, the generation of the halos and the fact that the background leaks into the white space behind the letters.

    All this is trivially explained by realizing the simple work flow: Scan on a Xerox workcenter, email using standard profile settings, open in preview, and print as PDF.

    This creates all the known artifacts with a minor exception of the two areas of white speckles which of course have NO explanation by a forgery either. The fact that one area overlaps with the area of the seal and the fact the there is despeckling done by the Xerox work center proposes a workable hypothesis, which is far more than Hermitian or anyone else has been able to present.

    Poor Hermitian has failed to understand how the raw PDF contains a wealth of information that remained hidden from him and complicated a proper analysis of the data, leading him to conclude that since he could not explain it, it must have been a forgery. The latter conclusion is supported by the fact that Hermitian has failed to present any workflow or hypothesis that explains all these artifacts in light of a ‘forgery’ hypothesis. Since in real science one compares competing hypotheses, the absence of a testable hypothesis on the part of Hermitian has rendered his analysis completely irrelevant.

    He at best managed to point out some interesting features which have now been carefully explained through a single workflow. Sure, some mysteries may continue as we do not have access to the original document for proper experimentation. Of course, the mere claim that one would get a better document by printing it out on a resolution higher than the original is just preposterous.

    Fascinating how we continue to have to educate our friend. At least I hope that he has found the time to install proper forensic analysis tools that allow him a low level insight into the document structure. There is just too much data otherwise lost.

    As I pointed out many times now, the mere fact that these so-called ‘experts’ have failed to provide any analysis of the raw data is troubling as this is how a forensic examiner would go about analysing a document properly. Especially when claims of fraud are involved.

  85. Poor Hermitian is still confused about the ‘double compression’

    I already showed that the original PDF stores the JPEG object as a FlateDecode/DCTDecode stream. Which means that the original jpeg was further ‘compressed’ using FlateDecode.

    I also showed that Preview does not rezip the JPEG but just stores it in a PDF Object.

    Needless to say, none of these steps do anything to the JPEG itself which remains rendered properly. This is what allowed me to find the YCbCr comment embedded in the JPEG and tracked it back to Xerox software.

    The alternative compression: First flatedecode and then DCTDecode makes no sense at all. Anyone familiar with PDF rendering would understand this.

  86. “Onlooker permalink

    “July 20, 2013 20:09

    “Does anyone really thing that the CCP as part of their research would have overlooked which scanning software that departmental copiers/scanners would have most likely been used to create this document. I mean you think even a bad expert in computer graphics would not know that scanners can create PDFs in many different forms. I would think that as part of their research they tried to replicate the LFBC using every piece of office equipment they could get their hands on. Especially any that were known to produce layers.”

    That’s certainly what any rational person would think. But then Obots aren’t rational.

  87. Hermitian: That’s certainly what any rational person would think. But then Obots aren’t rational.

    So we agree that this is rational. So why did the CCP miss testing the Xerox WorkCentre? Why did they not report on the low level PDF format which contained the comment in the JPEG? Why did they miss so many helpful and telltale signs?

    They failed at so many steps, that it would be a pleasure to rip their claims to shred in a court. But given the unlikeliness of this, I have to just take pleasure in ripping their findings apart on the internet.

    If they had done their homework they would have found how the Whitehouse released various scans using a Xerox WorkCentre and they would have taken the diligence to analyze the output of such a device.

    If they had done so, they would not have made such foolish claims that MRC compression never generates multiple bitmap layers, which was one of the shaky foundations on which they built a claim of fraud.

    Poor tools, poor diligent research all may have contributed to their failures.

  88. “NBC

    “Let me also point out to you that the embedded jpeg string found in Xerox Workcentre documents such as Obama’s WH LFBC is so far only found in document created by a similar workcentre.
    None of the many other jpegs I have collected show any sign of this string

    YCbCr

    The smoking gun that totally destroys the myth.”

    ROTFL ! Some smoking gun !

    The YCbCr color space is the standard color space for JPEG files that meet the
    JPEG File Interchange Format. You can read all about it in this article:

    JPEG File Interchange Format
    Version 1.02
    September 1, 1992

    See: http://www.jpeg.org/public/jfif.pdf

    And: JPEG File Interchange Format

    See: http://en.wikipedia.org/wiki/JPEG_File_Interchange_Format#Color_space

    The 1992 issue is the latest issue of the standard. Work on the JPEG format is dead. But if you want to exchange JPEGS across platforms then JFIF reigns.

    From the standard 1992 edition:

    “Standard color space

    “The color space to be used is YCbCr as defined by CCIR 601 (256 levels). The RGB
    components calculated by linear conversion from YCbCr shall not be gamma corrected
    (gamma = 1.0). If only one component is used, that component shall be Y.”

    You would expect Xerox to be hanging on to the past. But you can bet that there are tons of other JPEGS out there with YCbCr.

    I hope NBC didn’t pay a lot for his smoking gun !!!

  89. NBC

    “July 21, 2013 21:47

    “You forgot — I wrote the book on the placement of the object boundaries.

    “ROTFL… You are so funny… It was gsgs who ‘wrote the book’ and you added little relevance to the analysis.

    “For the longest time you could not do the proper analysis yourself, remember…

    Nope ! I don’t remember that. What I do remember is my trying to tell GSGS where he went wrong. He let the machine pick the x-y coordinates instead of the forger.

    [NBC: ROTFL. There is no evidence of a forger. Yes you tried a lot but GSGS’s findings remained valid.]

    I also remember that both YOU and GSGS stated that the rectangular boundaries do not satisfy 16 x 16 blocks. But then neither one of you knows a hoot about graphics. So I wasn’t surprised.

    [NBC: Says the one who did not even notice an embedded comment in a jpeg.. Really my friend, for an ‘expert’ you know remarkably little of the inner workings of JPEG or PDF beyond editing tools :-)]

    Read my treatise on the placement of the rectangular object boundaries again. And pay close attention to the relationship between the 16 x 16 blocks at 300 PPI x 300 PPI and the 8 x 8 blocks at 150 PPI x 150 PPI. Then come back and tell the readers why that relationship is essential to the composite image of the WH LFCOLB.

    [NBC: I have read your thoughts on this and nothing really of relevance pops out. Of course, anyone realizes that the text is compressed at twice that of the background and thus 8×8 in the background coincides with 16×16 blocks in the foreground. This is an inevitable effect of MRC as I and others have pointed out many many months if not years ago.]

    See: http://www.scribd.com/doc/151738307/Analysis-of-Rectangular-Object-Boundaries

    GSGS’s work was a good start but, like you, he assumed that a machine placed the rectangles rather than a person. That’s why he never solved the puzzle.

    [NBC: He solved the puzzle because he made no assumptions at all. He merely observed how the alignment worked and observed how unlikely a forger would do this.]

    The WH LFCOLB PDF image requires two congruent grids because the background and foreground are different resolutions yet text characters from both are used within the same word.

    [NBC: Duh… that was well know many years ago. The alignment of the foreground layers to the 8×8 bit background was not well described until gsgs figured it out.]

    That’s why both modulo conditions must be satisfied. Take a close look at the “R” in “BARACK” and the binary characters on either side. The “R” is on the background layer and thus is at 150 PPI x 150 PPI. The binary characters are at 300 PPI x 300 PPI.

    [NBC: Your point? I still fail to see any relevance to this other than that gsgs showed how all these objects aligned to 8×8 bit boundaries, which is quite well understood from a perspective of an algorithm but there is no reason why a forger would perform such a careful placement. Geez Hermitian you keep raising objections where your own findings support gsgs. The conclusions are irrefutable: An algorithm explains these boxes quite well, a forger… well, not so much…. Of course, so far you have failed to show anything that is better explained by a forger… Your ‘expert testimony’, while it would never be admitted in court, would not survive the first minutes of examination. You should thank us for having you spared from any further embarrassments ]

  90. The YCbCr color space is the standard color space for JPEG files that meet the

    I know what it stands for and I am happy to hear that you did some research to familiarize yourself.

    You still do not understand the significance. There is no need for this comment to be in the jpeg and other than Xerox WorkCentre related PDF’s I have not found the comment in other JPEG’s

    Why is it that you have such trouble with simple logic?

    Show me other JPEGs that have YCbCr encoded. And yes, the encoding is best used because of the way our eye works. JFIF is a newer format to embed information but the good old JPEG comment still exists.

    And of course, we know that Xerox software does embed it so I am looking forward to you presenting to us these tons of other JPEGs.

    Well Hermitian….

  91. “NBC

    “July 22, 2013 04:33

    “The YCbCr color space is the standard color space for JPEG files that meet the

    “I know what it stands for and I am happy to hear that you did some research to familiarize yourself.

    “You still do not understand the significance. There is no need for this comment to be in the jpeg and other than Xerox WorkCentre related PDF’s I have not found the comment in other JPEG’s

    “Why is it that you have such trouble with simple logic?

    “Show me other JPEGs that have YCbCr encoded. And yes, the encoding is best used because of the way our eye works. JFIF is a newer format to embed information but the good old JPEG comment still exists.

    “And of course, we know that Xerox software does embed it so I am looking forward to you presenting to us these tons of other JPEGs.

    “Well Hermitian….

    Funny you didn’t share with your readers that YCbCr is the ONLY permitted color space that meets the JPEG File Interchange Format (JFIF) and therefore these JPEGS are as common as dirt.

    [NBC: I presumed that people were familiar with JPEG and it is not relevant to my findings. There is no need to have this comment embedded and yet, for some reason I have shown the Xerox does embed it…]

    You even chopped off “JPEG File Interchange Format” from my comment.

    ROTFL !!!!

    You are the guy that’s claiming exclusivity to Xerox.

    [NBC: then show me an example of a jpeg with said embedded comment…]

    So you are the guy who has to prove exclusivity to Xerox.

    So how many JPEGS have you personally looked at?

    And how many JPEGS do you suppose are out there?

    So why don’t you ask Xerox if they will stand behind your ridiculous claim?

    I rest my case. Good luck Dude…

    [NBC: As expected, Hermitian was given a scientific opportunity to prove me wrong and he failed]

  92. “[NBC: ROTFL. There is no evidence of a forger. Yes you tried a lot but GSGS’s findings remained valid.]”

    So are you denying that both YOU and GSGS posted that the Rectangular object boundaries of the WH LFCOLB PDF image do NOT satisfy the 16 MOD 0 ?

    [NBC: Which is correct. As I have shown, these boundaries do not always satisfy 16 MOD 0]

    Because If that’s what you are saying then you are lying.

    [NBC: Hilarious… ]

    Of course the most important word in both your posts about the 16 x 16 blocks is the word “NOT” !

    When I saw your posts I knew immediately that you were wrong.

    It took me awhile to prove it both Mathematically and Graphically.

    [NBC: Anyone familiar with math realizes that when you divide something by 8 and are left with an odd number, it will not be integer-divisible by 16 … Duh.. Perhaps the math is a bit too challenging? ]

    You know the tragedy here is that you still don’t get it dude !

    How many graphics classes have you sat in on dude? Ever had any course work in drafting? How about descriptive geometry? Ever produce a drawing using Autocad, Pro Engineer or Turbocad?

    [NBC: Keep digging yourself that hole…]

    Ever use a 3D Finite Element mesh generator dude? Ever produce an FEP model with millions of nodes and elements? Ever develop from scratch the graphics program to analyze FEP results in 3D color?

    [NBC: Ever bothered to look at the raw PDF data Now that was a real oversight]

    So to make it easy on you —

    Just remember that GSGS’s 16 x 16 blocks don’t work but mine do.

    [NBC: Therefore your 16×16 blocks are not gsgs’s blocks. Dude … So we were right after all]

    There is only one coordinate system within which the WL LFCOLB could have been constructed and it is not the user space coordinate system of the PDF but rather the coordinate system of the forgers vector graphics program. And forgers use the most popular vector graphics software.

    [NBC: What forger? The relevant coordinate system shows that the graphics were stored rotated.. Why would anyone work on rotated graphics… It does not make sense… And yet from our experiments we know that Xerox does exactly this… Again an algorithm beats this ‘forger’ non-explanation of yours.]

    The forger’s screen is where the relationship between the page boundaries and the coordinate grids for each image is established. Each of the nine images that make the WH LFCOLB PDF image is constructed on the forger’s screen at a screen resolution of 1/72 in. x 1/72 in.

    [NBC: ROTFL… ]

    The transformation of the device space of his monitor to the PDF user space coordinates is specified by the CTM (Current Transformation Matrix) of the device. This matrix is an element of the PDF graphics state.

    [NBC: Throwing out random terms you have just learned is not very impressive. Yes the CTM can be observed in action in the raw data. How come you failed to look at them?]

  93. “NBC

    “The embedded comment, the fact that it generates multiple foreground layers, the fact that it does JBIG2 compression, all align with the proposed work flow which was simple and straightforward.”

    These connections exist only in your mind Dude.

  94. “NBC

    “He at best managed to point out some interesting features which have now been carefully explained through a single workflow. Sure, some mysteries may continue as we do not have access to the original document for proper experimentation. Of course, the mere claim that one would get a better document by printing it out on a resolution higher than the original is just preposterous.”

    If you mean a 600 dpi print of a PDF document with a bitmap image of 300 PPI resolution is preposterous then you’re preposterous.

    You just don’t get it Dude! You sure have a mental block for anything to do with graphics.

    I posted a link that explains all about dpi vs PPI and you still didn’t get it.

    No wonder that you prefer to stare at lines of code with your freeware parser.

    [NBC: You can buy someone an expensive tool but when they do not know how to use it properly or use it inappropriately, their conclusions will be limited. This freeware parser provides you with an in-depth insight into the guts of the PDF document. i can understand you may find it too challenging but as I have shown, it prevents you from finding the real tell-tale signs… If you really want to do a forensic examination, you cannot avoid these tools. ]

  95. “NBC

    “Show me other JPEGs that have YCbCr encoded. And yes, the encoding is best used because of the way our eye works. JFIF is a newer format to embed information but the good old JPEG comment still exists.

    The last JFIF standard was issued in 1992. So that’s new to you? I wonder how many JPEGS were issued under JFIF since 1992?

    “And of course, we know that Xerox software does embed it so I am looking forward to you presenting to us these tons of other JPEGs.

    “Well Hermitian….

    Your challenge is a fool’s errand Dude…

    1. A reader accepts NBC’s challenge

    2. He quickly finds another JPEG with the YCbCr embedded comment and posts a link to it for NBC

    3. NBC verifies that the reader’s JPEG indeed has the YCbCr

    4. NBC informs the reader that his JPEG was created by a Xerox

    Get it ?

    [NBC: Oh I get it, you have accepted my findings and are now making up excuses…]

  96. Funny you didn’t share with your readers that YCbCr is the ONLY permitted color space that meets the JPEG File Interchange Format (JFIF) and therefore these JPEGS are as common as dirt.

    Totally irrelevant. The fact that the JPEG contains the embedded comment and the same one is found in Xerox WorkCenter provides additional evidence.

    So far I have not found any JPEG that contained this comment other than those from Xerox WorkCenters.

    It should therefor be an easy job for you to disprove me…
    All you have to do is find one…

    But I doubt you are up for it…

    Please explain to me what is so ridiculous about my claim

    Poor confused Hermetian… He cannot even explain why finding an embedded comment tag in a JPEG is not relevant when it so far has been found in other Xerox work centers only.

    Poor Hermetian still does not udnerstand the scientific method,

  97. So are you denying that both YOU and GSGS posted that the Rectangular object boundaries of the WH LFCOLB PDF image do NOT satisfy the 16 MOD 0 ?

    They don’t. Anyone can easily observe this.

    My goodness sakes …

    I am glad that you care learning more about PDF and coordinate transformation though, although quite irrelevant to our present discussion.

    gsgs resolved a minor mystery that showed, once again, that it is best explained by an algorithm, not a forger.

  98. “The embedded comment, the fact that it generates multiple foreground layers, the fact that it does JBIG2 compression, all align with the proposed work flow which was simple and straightforward.”

    Like your forger claim, with the additional benefit that I can actually show evidence of both JBIG2 compression as well as the embedded comment?

    Poor soul…

  99. No wonder that you prefer to stare at lines of code with your freeware parser.

    I am sorry to hear that you still do not understand the low level coding of PDF’s.

    PPI, DPI… It was a preposterous statement of yours that we should print out the low resolution document at a higher DPI.

    Anyone who is familiar with graphics would understand this.

  100. I am done reading anything Hermie writes. He has been told many times the strings of HHHHH’s are distracting and to use the blockquote function yet this hard headed fool thinks he needs to use them because it is easier for him to find his own nonsense using the search function. Henry is further proof that like old dogs you cannot teach Birthers anything.

  101. 1. A reader accepts NBC’s challenge

    2. He quickly finds another JPEG with the YCbCr embedded comment and posts a link to it for NBC

    3. NBC verifies that the reader’s JPEG indeed has the YCbCr

    4. NBC informs the reader that his JPEG was created by a Xerox

    Get it ?

    Well, it’s a good assumption based on the evidence so far that Xerox workcenters create the comment and that no other jpegs have been found.

    But anyone familiar with the scientific method would know that there are ways to exclude the Xerox workcenter as the creator.

    I am glad to hear to that you accept so far that the comment is created by the Xerox work center and that no other examples have been found.

    Somehow, ‘researchers’ missed this tell tale sign as they focused on high level tools that hid from their sight the relevant information.

    It’s fine Hermitian, you tried and failed… Nothing wrong about finally admitting to that.

    We now have two viable work flows explaining the WH PDF and the 35-1 document and Hermitian has nothing more than ‘well yes but it must have been a forger’.

    Such arguments to not survive the laugh test, let alone peer review, and not to mention cross examination in a court of law.

    Count yourself lucky dude that Vicklund and others helped you realize that your ‘arguments’ were based on an unfamiliarity with the tools.

    Did you not have paper capture yourself and it never occurred to you to do the experiment to see if it would create these artifacts?

    Really?…

  102. The last JFIF standard was issued in 1992. So that’s new to you? I wonder how many JPEGS were issued under JFIF since 1992?

    You are still being confused about the argument.

    I am not arguing that the use of YCbCr colorspace is unique to Xerox but rather the fact that they embedded the colorspace into a comment. A comment plays no role in how the jpeg is rendered.

    Hilarious… I have shown how a fingerprint was found and all Hermitian has to offer is: well, there may be others with the same fingerprint out there.

    But he fails to provide any relevant examples…

    Remember my hypothesis is extremely vulnerable as it has to overcome the examination of every jpeg while Hermitian’s task is far simpler.

    But even there he faulters.

  103. Hermitian won’t tell…

    Of course, his failure to explain this becomes his albatross🙂

    I can explain the embedded comment through a simple workflow on a device we know is owned by the White House and used to scan in documents (tax forms for example)

  104. Reality CheckHe has been told many times the strings of HHHHH’s are distracting and to use the blockquote function yet this hard headed fool thinks he needs to use them because it is easier for him to find his own nonsense using the search function.

    It is even harder to accept anything Hermie states about PDF construction when he will not even use a simple HTML code such as blockquote. He does not even need to add the reference back to the original comment, which is not that hard either.

    Lucky for me, I stopped reading Hermie’s comments, except the occasional brief scan.

Comments are closed.