Xerox 7655 – Clipping path and objects

The following three images show the ‘raw’ Xerox file, which has no clipping path at the top level and the Preview saved version which shows that Preview added a clipping path. Not also how Illustrator cannot properly deal with the Rotate annotation and shows the document sideways.

Xerox No Clipping Path. Objects Highlighted

Layers expanded no previewAfter rotating the document and saving it, the preview version shows an additional clipping path, just like in the WH LFBC PDF

Xerox Preview saved – Clipping path turned on

clipping pathXerox Preview saved – All objects in the Xerox  file

objects

74 thoughts on “Xerox 7655 – Clipping path and objects

  1. This is after the PDF was saved with Preview. Can you add one more image the file before Preview but with all the objects highlighted? It should not have a clipping mask but all the other objects should be showing.

  2. Thank you. I didn’t mean to assign homework but it clarifies that the 7655 does not add a clipping mask but Preview does.

    The presence of the clipping mask is a big deal to Zullo.

  3. Don’t get me wrong. I appreciate your feedback. And when it helps Zullo to ‘see the light’ it is well worth the little effort, although it is probably wasted on him.

    Anything to clarify and focus on the remaining issues is appreciated and I thank you for your efforts here.

    As to Zullo and his comedy, I tend to spend little time on him, he has nothing…

  4. Do you remember the video that Mark Gillar made about clipping masks? I will try the embed code for the the video but if that doesn’t work I will link it in a separate comment.

  5. Reality Check says:

    August 9, 2013 at 16:17

    You guys should watch Gillar’s video. It is a real hoot especially now that we know how simple it is to explain the clipping mask.

    Not so fast RC the Radio Head !

    It won’t be so simple to explain your Preview Clipping Mask after you read the following:

    Another deficiency of the Xerox 7535 / Preview PDF.

    I am continuing to compare the PDF file “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” with the archived copy of the WH LFCOLB downloaded from here:

    http://web.archive.org/web/20110427171111/http://www.whitehouse.gov/sites/default/files/rss_viewer/birth-certificate-long-form.pdf

    In this posting I will show that the Xerox 7535 scan to PDF removed information from the “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” that is present in the PDF file “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf”.

    This posting will focus on four Clipping path objects (four within each PDF).

    As a point of reference, I have summarized the W and H dimensions of four rectangular objects from the WH LFCOLB PDF. These four objects will then be compared with four rectangular objects from the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf”. Except for the Artboard object, the rectangular boundaries of each rectangle also define a clipping path in Adobe Illustrator. The rectangles are listed from the largest to the smallest within each table.

    Object W(px) H(px) W(in.) H (in.) Pixels/in.

    Background 1276 1652 8.506666 11.013333 150

    Artboard 612 792 8.5 11.0 72

    Clipping Path 1200 1590 8.0 10.6 150

    Mostly Text 1819 1454 6.063333 4.846666 300

    As defined within the PDF file, each image object is assumed have a pixel resolution of 72 PPI. This is the default pixel resolution for PDFs. The corresponding image layer sizes for each image object are larger than the final sizes. Consequently I have replaced the common 72 PPI with the final resolution appropriate for each image’s final dimensions. Thus the object dimensions given in the tables are the final object dimensions as measured in Illustrator.

    The corresponding four rectangular objects from the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf”are as follows:

    Object W(px) H(px) W(in.) H(in.) Pixels/in.

    Background 1280 1664 8.533333 11.093333 150

    Artboard 612 792 8.5 11.0 72

    Clipping Path 1225 1600 8.166666 10.666667 150

    Mostly Text 1749 1403 5.83 4.676666 300

    First, comparing the sizes of the corresponding rectangles we find that the Background rectangle from the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” is larger than the Background rectangle of the PDF “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf”. However, to the contrary, the mostly text rectangle of the Xerox / Preview PDF is considerably smaller than the mostly text rectangle of the WH LFCOLB PDF. I suspect that this difference was caused by the Mixed Raster Compression. Of course this effect would not be present in the WH LFCOLB PDF image because MRC was not applied to this image. Instead, a human forger manually set all of the rectangular object boundaries.

    Also the Clipping Path Rectangle from the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” is larger than the Clipping Path Rectangle from the PDF “ “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf”.

    The Clipping Path from the Xerox / Preview PDF lies between the border of the Green basket-weave background and the Artboard boundary. The Clipping Path from the WH LFCOLB PDF coincides with the outer boundary of the Green basket-weave background.

    The Clipping Path is purported to be added by Preview. The Rectangle dimensions (W = 1225 px; H = 1600 px) are not found by search within the PDF. These dimensions are appropriate for a pixel resolution of 300 PPI x 300 PPI. The corresponding dimension in points for a resolution of 72 PPI x 72 PPI are W = 588 pts x 768 pts. The numbers 588 and 768 are also not found by searching within the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” The Clipping Path is also not included in the object list. However we found a clipping path object with exact dimensions 588 pts x 768 pts by means of Adobe Acrobat Preflight. Preflight also showed the Background Rectangle dimensions after application of the Clipping Path. The Background Rectangle dimensions in Acrobat Preflight were also 588 pts x 768 pts.

    The corresponding Clipping Path from the PDF file “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf” has dimensions of 576 pts x 763.2 pts which also are the dimensions of the outer border of the Green basket-weave safety background.

    It is important to emphasize that the Clipping Paths do not remove the clipped portion of the background image from the PDF files. This was easily verified by removing the Clipping Path in Adobe illustrator.

    When the Clipping Path was turned off in the PDF “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf” two pencil marks near the right side of the background were revealed. Also revealed were additional White-spot anomalies belonging to the White-spot object at the top of the page.

    The pencil markings were not present when the Clipping Path object was turned off in the PDF “wh-lfbc-scanned-xerox-7535-wcpreview.pdf”. There are no White-spot artifacts in the Xerox / Preview image.

    So we have proved that information that is present outside of the border of the Green basket-weave background in the PDF “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf”is not present in the PDF file “wh-lfbc-scanned-xerox-7535-wcpreview.pdf”.

    Importantly, the missing information in the image “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” was not caused by the Clipping Path that was added to the PDF by Preview. Moreover, the methodology used to set the dimensions of this Clipping Path object is unknown.

    We have also proved that the sharp border of the Green basket weave background in the PDF file “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf” resulted from the application of a Clipping Path object.

    To the contrary, the border of the Green basket-weave background in the PDF file “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” is not sharp. In fact the right edge of border is noticeably ragged. Also the width of the White border of the Xerox / Preview PDF image is wider that the width of the White border of the WH LFCOLB PDF image.

    Collectively, these findings are proof that the Xerox / Preview forger just doesn’t have the right stuff.

  6. Hermie, the 7535 image was made from printing out the WH LFBC PDF and then scanning it on a 7535 Xerox WorkCentre. Anything blocked by the clipping mask in the WH LFBC PDF does not print out. Therefore, it will not appear on the 7535 image.

    Also, the clipping mask is not a separate object, as has been explained to you multiple times. It is a command contained in the Obj 6 flate-encoded string.

    Your entire comment completely fails.

  7. Moreover, the methodology used to set the dimensions of this Clipping Path object is unknown.

    Nope… Unknown to you perhaps. But rather than let such ignorance become a foundation for an argument from ignorance, a researcher would attempt to understand how the clipping path object is set.

    Hint: When printing you can select a profile which has the borders defined. This defines the clipping path. When you define a new border setting, you can have this one become the default setting.

    Come on Hermitian, do not let your lack of knowledge become the endpoint of your investigations, that’s where the real fun starts: trying to figure out how it works. This intellectual curiosity is what helps one to uncover knowledge that so far has remained in accessible and enriches ones personal knowledge-base as well as one’s ability to formulate hypotheses.

  8. To the contrary, the border of the Green basket-weave background in the PDF file “wh-lfbc-scanned-xerox-7535-wcpreview.pdf” is not sharp. In fact the right edge of border is noticeably ragged. Also the width of the White border of the Xerox / Preview PDF image is wider that the width of the White border of the WH LFCOLB PDF image.

    Collectively, these findings are proof that the Xerox / Preview forger just doesn’t have the right stuff.

    So you let your ignorance as to how the clipping mask conclude to you that it cannot explain the evidence and yet.. Properly understood, as Vicklund does, it all falls in place quite nicely.

    When the Clipping Path was turned off in the PDF “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf” two pencil marks near the right side of the background were revealed. Also revealed were additional White-spot anomalies belonging to the White-spot object at the top of the page.

    Yes, the clipping mask was wider than the edge-erase. Of course that is somewhat harder to repeat when printing out the document since the wider clipping mask is covering and thus the test results in no additional information being shown.

    Let me help you understand:

    1. Edge erase does not cut off the image but merely adds a band of white pixels 0.10 or 0.12” wide, as set by the scanner
    2. Clipping mask covers up a border set by the user, either globally or temporarily, but does not remove the underlying information, it just covers it up.

    The clipping mask used appears to be 0.20”x0.25” and thus when releasing it, it uncovers “hidden” parts of the image.

    Since both step 1 and 2 are user/admin defined, we can only reverse engineer its settings and then show its impact by repeating the same settings. But we cannot predict it.

    Another meaningless objection by our friend Hermitian. Luckily we have Vicklund to help us out here, as he did the original analysis. It does take some understanding of the processes involved going from the 600 PPI scan to the final document.

  9. Hermie, the 7535 image was made from printing out the WH LFBC PDF and then scanning it on a 7535 Xerox WorkCentre. Anything blocked by the clipping mask in the WH LFBC PDF does not print out. Therefore, it will not appear on the 7535 image.

    Also, the clipping mask is not a separate object, as has been explained to you multiple times. It is a command contained in the Obj 6 flate-encoded string.

    Your entire comment completely fails.

    It surely does… And this is simple logic and a little understanding of how the processes work. So is Hermitian still looking for the clipping mask ‘object’? That is hilarious… He should by now know what a clipping mask looks like and where it would be found and why it is NOT a separate object in a PDF file. Truly fascinating even though we have explained it to him many times.

  10. 1. Edge erase does not cut off the image but merely adds a band of white pixels 0.10 or 0.12” wide, as set by the scanner

    This statement is somewhat ambiguous, so let me clarify. Edge erase changes the outer band of pixels (thickness in inches as set on the scanner) to pure white. This is done to the raw image file, prior to any additional processing, such as adding rows and columns to achieve the mod 8 @ 300 ppi criteria.

    One thing to note – a number of scanners do not have a mod 8 criteria. Both HP and Brother scanners will readily produce jpegs and pdfs with embedded jpegs that are not mod 8 at 300 ppi, or 600 ppi, or even 1200 ppi.

  11. W. Kevin Vicklund says:

    August 18, 2013 at 18:21

    “Hermie, the 7535 image was made from printing out the WH LFBC PDF and then scanning it on a 7535 Xerox WorkCentre. Anything blocked by the clipping mask in the WH LFBC PDF does not print out. Therefore, it will not appear on the 7535 image.

    “Also, the clipping mask is not a separate object, as has been explained to you multiple times. It is a command contained in the Obj 6 flate-encoded string.

    “Your entire comment completely fails.”

    Per two different PDF code parsers, the background image is Im1 and obj {7 0 R}.
    Per Adobe Acrobat XI Pro Preflight the Clipping mask is not associated in any way with either the DCTDecode filter or the FlateDecode filter. Instead, the clipping mask is associated directly (and only) with obj {7 0 R}.

    And to think that you and NBC claim to be the PDF code masters.

    But then unbridled arrogance always proceeds a big fail.

  12. >blockquote>Per two different PDF code parsers, the background image is Im1 and obj {7 0 R}.
    Per Adobe Acrobat XI Pro Preflight the Clipping mask is not associated in any way with either the DCTDecode filter or the FlateDecode filter. Instead, the clipping mask is associated directly (and only) with obj {7 0 R}.

    And to think that you and NBC claim to be the PDF code masters.

    But then unbridled arrogance always proceeds a big fail.

    Funny, I have seen this precede your next fails🙂

    The clipping mask is at the top of the rendering instruction. Anyone can read it in PDF and figure out how they group together.

    Relying on Illustrator again ignores looking at what Illustrator read and it forces you to understand how Illustrator makes sense of this.

    The clipping mask is active when all objects are ‘printed’, but it does require you to understand PDF.

    Have you found the clipping mask yet?

    Check out this posting and perhaps you understand. If not, no worries, we can help

  13. W. Kevin Vicklund says:

    August 18, 2013 at 20:07

    “”Moreover, the methodology used to set the dimensions of this Clipping Path object is unknown.””

    “The dimensions of the clipping path are set by printable area of the current printer of the computer.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    The point is Dude that the Clipping Path is set way outside the border of the Green basket-weave background. And somehow you still end up with a broader White band around the page.

  14. The point is Dude that the Clipping Path is set way outside the border of the Green basket-weave background. And somehow you still end up with a broader White band around the page.

    Yes.. I am sure you understand the origin of this… It requires a bit of thinking as to what was scanned…

    Remember:

    WH LFBC had no white borders
    Xerox started with white borders of the WH LFBC

    Duh…

  15. NBC

    “”Hint: When printing you can select a profile which has the borders defined. This defines the clipping path. When you define a new border setting, you can have this one become the default setting.””

    HHHHHHHHHHHHHHHHHHHHHHHHHH
    NBC if this is the best that you can do then (as far as I am concerned) you can leave off the hints.

    There’s no there there !

    Two sentences = two degrees of freedom.

    NBC just can’t help himself.

    We’ll never get a straight answer out of this slippery guy. His motto is “fuzz it up”.

  16. The point is Dude that the Clipping Path is set way outside the border of the Green basket-weave background. And somehow you still end up with a broader White band around the page.

    The most annoying of default Adobe settings: Fit to Page for a printer with a print area smaller than the standard page. The whole image has been scaled down. If past experience is any judge, to 96%.

  17. Two sentences = two degrees of freedom.

    You wanted to know how it works. Since you seem to be having trouble educating yourself here, I provided you with the necessary information.

    I guess it may be information overload? But regardless, you have yet to show that any of my findings are wrong.

    Although, I do admire your efforts, it’s just that so far I am not very impressed by the quality of the arguments.

    Have you figured out the 8 bit alignment yet?

    Or the YCbCr comment

    Or the… or the…

    You’re funny my friend, really funny. And it is a pleasure to have you contribute on my blog.

  18. So we have proved that information that is present outside of the border of the Green basket-weave background in the PDF “birth-certificate-long-form WayBack Machine 04-27-2011_17-11-11.pdf”is not present in the PDF file “wh-lfbc-scanned-xerox-7535-wcpreview.pdf”.

    Yes, since the Xerox scans were based on the WH LFBC and since the clipping mask hides the information, you do not see it on the PDF.

    Duh…

    Is that it? A bit of thought would have saved you from wasting all this time on a non issue

  19. The most annoying of default Adobe settings: Fit to Page for a printer with a print area smaller than the standard page. The whole image has been scaled down. If past experience is any judge, to 96%.

    Yes, the scaling aspect does complicate matters even further. I believe you are right with the 96%. Is this what is confusing our friend?

    For tabloid, 11×17 I found 97%. Too bad our friend does not have a Mac so he could have studied this himself.

  20. If past experience is any judge, to 96%.

    Same in preview… Annoying 96% for the WH LFBC if I print to PDF.

    I can set it to 100% if that helps Hermitian.

    Another non issue, what a waste of effort… Why does he not focus on something relevant? Or is this the only thing he can object to?

    Fascinating.. With the recent Museum PDF we have once again confirmation of many of the artifacts.

    And Hermitian has done nothing to show evidence of a forger, other than assert it. Oh the argument from ignorance…

  21. Based on the size of the clipping mask in the various Preview files NBC has generated, I predict that his default printer scales an 8.5×11 pdf to 96% for Fit to Page.

  22. I would say on general principle you should print at full size for future documents. Fortunately, the printer at the WH has margins almost as bad as the HP 7500 (1″ margins on a 11×17″! I hate that thing – been trying for years to get my boss to replace it, but it’s the only thing that scans 11×17 in the office. Sorta scans 11×17.)

  23. I would say on general principle you should print at full size for future documents.

    Doing this already🙂 Did not want to mess with the most likely work flow but I see how this may have caused confusion to Hermitian.

  24. NBC

    “1. Edge erase does not cut off the image but merely adds a band of white pixels 0.10 or 0.12” wide, as set by the scanner
    ” 2. Clipping mask covers up a border set by the user, either globally or temporarily, but does not remove the underlying information, it just covers it up.

    HHHHHHHHHHHHHHHHHHHHHHHHHHH
    So the Xerox edge erase doesn’t explain the wide White border because it’s at most 0.12 wide. Likewise the Clipping Mask also doesn’t explain the wide white border because the distance between the Clipping Path and the outer edges of the background page is uniform and only 0.1667 in.

    The measured width between the rectangular boundary of the Green basket-weave safety background and the outer rectangle edges of the background page varies between 0.35 in. to 0.43 in. for the Xerox 7535 / Preview background image.

    So, within your proposed workflow, what’s the cause of this wide White border ?

    The White border is 0.25 in. wide along the long edges and 0.20 in. wide for the short edges of the WH LFCOLB background image.

  25. NBC

    “Have you figured out the 8 bit alignment yet?”

    For the WH LFCOLB, the alignment with the 8 x 8 blocks @ 300 PPI x 300 PPI is satisfied for the upper-left origin or for the upper-right origin. For the 16 x 16 blocks @ 300 PPI x 300 PPI the alignment is satisfied for the upper-left origin but not for the upper right origin. The alignment to the 8 x 8 blocks @ 150 PPI x 150 PPI is also satisfied for the upper-left origin but not for the upper-right origin.

  26. W. Kevin Vicklund says:

    August 19, 2013 at 01:22

    “Based on the size of the clipping mask in the various Preview files NBC has generated, I predict that his default printer scales an 8.5×11 pdf to 96% for Fit to Page.”

    But the size of the background page of the Xerox 7535 / Preview PDF image is larger than the background page of the WH LFCOLB. So the Xerox WC increases the page size but shrinks the image ?

  27. So, within your proposed workflow, what’s the cause of this wide White border ?

    We just spent what, 10 comments going over that? Fit to Page.

    You can verify that the image has been scaled yourself. Two easy methods:

    1) Count the repeats of the basketweave pattern. If they are the same, the image has been scaled. Requires no tools other than a viewer

    2) Measure a large object on the page, such as one of the horizontal form lines. If they are different, the image has been scaled. You can calculate the scale factor by dividing the dimension from the 7535 file by the dimension from the WH LFBC file.

  28. But the size of the background page of the Xerox 7535 / Preview PDF image is larger than the background page of the WH LFCOLB. So the Xerox WC increases the page size but shrinks the image ?

    No, whatever printer was used shrunk the image when it printed it to an 8.5×11 sheet of paper. The increase in overall image dimension is due to that sheet of paper being skewed by about half a degree when it was scanned by the Xerox (skew verified by measurement). Are you really that stupid that you can’t separate the act of printing from the act of scanning?

  29. “Yes, since the Xerox scans were based on the WH LFBC and since the clipping mask hides the information, you do not see it on the PDF.”

    The Clipping Mask of the WH LFCOLB hides at most 0.25 in. on the long sides and 0.2 in. on the short sides.

    The White border on the Xerox / Preview image is nearly twice the width produced by the Clipping mask of the WH LFCOLB. And at the same time the page size of the Xerox / Preview background image is larger than the page size of the WH LFCOLB background page. So you have to explain why the page size increases but the image shrinks with your Xerox / Preview workflow.

  30. I’ve already explained it. But I’ll hold your hand and step you through the math.

    Many printers have an unprintable margin of around .15-.2 inches on all sides – sometimes more, sometimes none at all. For the ease of calculation, let us use .15 inches [note that the actual value probably is different]. So what happens when you try to print with Fit to Page selected?

    The printer will scale the image so that the outer edge of the page (Artboard in Illustrator, Page in PDF format) fits within the printer margins, while maintaining the aspect ratio. For our test printer, what is that scale factor? First, let’s look at the width. In order to fit, the width must be 8.5-(2*.15) &lte; 8.2, or &lte; 96.5%. Similarly, the height must be 11-(2*.15) &lte; 10.7, or &lte; 97.3%. Since the width required the most scaling, that is the scaling that must also apply to the height. Reversing the equation, we find that Fit to Page adds margins of .15 inches to the long sides, and .194 inches to the short sides.

    But don’t forget about the margin from the clipping mask! That margin still remains, though scaled. We have to add the scaled margin to the printer margins to find the final margin. So the long side margins are .2*.965 + .15 = .343, while the short side margins are .25*.965 + .194 = .435 inches.

    Which happens to be almost exactly what Hermie measured! (and how’s that for eyeballing the scale)

    Remember, this all printed on a sheet of paper nominally 8.5″x11″. So when you place this sheet of paper slightly askew on the Xerox scanner, the raw scanned image will include the edge of the sheet of paper – which being slightly askew, requires a bigger rectangle. The edge is later obliterated by the edge erase, of course.

  31. Hermitian: For the WH LFCOLB, the alignment with the 8 x 8 blocks @ 300 PPI x 300 PPI is satisfied for the upper-left origin or for the upper-right origin. For the 16 x 16 blocks @ 300 PPI x 300 PPI the alignment is satisfied for the upper-left origin but not for the upper right origin. The alignment to the 8 x 8 blocks @ 150 PPI x 150 PPI is also satisfied for the upper-left origin but not for the upper-right origin.

    Sigh.. but you claimed that it did not work for the 7355 documents. Where you wrong?… Very likely… But will you admit to it? Very unlikely…

  32. My goodness, how clueless can one be… Thanks Kevin for making the effort. This is just getting too surreal for me…

  33. NBC

    “Xerox 7655 – Clipping path and objects

    Posted on August 9, 2013 by NBC

    “The following three images show the ‘raw’ Xerox file, which has no clipping path at the top level and the Preview saved version which shows that Preview added a clipping path. Not also how Illustrator cannot properly deal with the Rotate annotation and shows the document sideways.”

    So now you are claiming that your Xerox forged image is so special that it can only be viewed in Preview ?

    NBC still thinks that Adobe is some fly-by-night software house.

  34. So now you are claiming that your Xerox forged image is so special that it can only be viewed in Preview ?

    Well, only in Preview, and Adobe Reader, and Adobe Acrobat, and Corel Draw, and…

    …pretty much every pdf viewer except Illustrator.

    This is a major problem for Illustrator. Just Google the following:

    Illustrator “/Rotate” pdf

    and see all the bitching about how pdfs are rotated 90 degrees in Illustrator.

    As it turns out, Illustrator does not know how to parse the /Rotate command for /Page objects (generally called Artboards in Illustrator). When Preview saves a pdf, it gets rid of the /Rotate command and instead rotates all the images individually, which Illustrator can parse. In fact, this may be part of why Preview modifies the structure of the pdf: to make sure that Illustrator can properly view the file.

    We’ve been explaining this to you for several months, Hermie.

  35. So now you are claiming that your Xerox forged image is so special that it can only be viewed in Preview ?

    Nope, other readers work equally well, it’s just that it is a known issue in illustrator.

  36. WKV

    “This is a major problem for Illustrator. Just Google the following:

    “Illustrator “/Rotate” pdf

    “and see all the bitching about how pdfs are rotated 90 degrees in Illustrator.

    “As it turns out, Illustrator does not know how to parse the /Rotate command for /Page objects (generally called Artboards in Illustrator). When Preview saves a pdf, it gets rid of the /Rotate command and instead rotates all the images individually, which Illustrator can parse. In fact, this may be part of why Preview modifies the structure of the pdf: to make sure that Illustrator can properly view the file.”

    “We’ve been explaining this to you for several months, Hermie.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    Illustrator is a vector graphics program designed to work with layered PDFs. So if you send it a PDF with layers then it looks for the rotation matrix for each layer.

    It’s really a deficiency on your part because you sent it a rotate command that’s appropriate for a flattened PDF image.

    Besides, there is zero evidence that your problem with the Xerox workflow and image rotations occurred at any step in the creation of the WH LFCOLB PDF. The WH PDF loads fine in Illustrator.

    It’s your Xerox forger that’s the problem.

  37. It’s really a deficiency on your part because you sent it a rotate command that’s appropriate for a flattened PDF image.

    I will note that Illustrate can’t handle the /Rotate command on a flattened PDF image, either. So your argument is basically a deficiency on Adobe’s part because it made a command available in the PDF format that Illustrator can’t handle.

    Well, I guess I can’t argue with that.

  38. NBC says:

    August 19, 2013 at 00:51

    “”>blockquote>Per two different PDF code parsers, the background image is Im1 and obj {7 0 R}.
    Per Adobe Acrobat XI Pro Preflight the Clipping mask is not associated in any way with either the DCTDecode filter or the FlateDecode filter. Instead, the clipping mask is associated directly (and only) with obj {7 0 R}.””

    It’s nice to see that NBC is having trouble with the blockquote on this wreck of a blog site ! But fortunately my comment remains untouched.

    “”And to think that you and NBC claim to be the PDF code masters.””

    “”But then unbridled arrogance always proceeds a big fail.””

    “Funny, I have seen this precede your next fails

    “The clipping mask is at the top of the rendering instruction. Anyone can read it in PDF and figure out how they group together.

    You obviously know nothing about Adobe Acrobat and even less than nothing about Adobe Acrobat Preflight. Preflight looks at the PDF at code level. Whatever it shows you can take to the bank because it is fully verified. And Preflight displays the DCTDecoded and FlateDecoded images from the stream data which is something that your freetoy PDF parser cannot do.

    [NBC: I am impressed, so you pay hundreds of dollars for something you can look at for free… And of course, since Adobe interprets the PDF, it may cause confusion to those looking through this filter..]

    “Relying on Illustrator again ignores looking at what Illustrator read and it forces you to understand how Illustrator makes sense of this.

    The clipping mask is active when all objects are ‘printed’, but it does require you to understand PDF.

    My findings were completely independent of Illustrator. I used two other tools both of which are code-level PDF parsers.

    Have you found the clipping mask yet?

    I found it Just where it was when I reported it the first time.

    The Clipping Mask and the Graphic-state are are separate branches in the PDF tree structure. They are at the same level in the tree and are connected to a common node. The common node connects directly to obj [7 0 R]. The Color Space and Image Stream are together in a separate branch which also connects directly to obj [7 0 R].

    And that’s just the way it is.

    Check out this posting and perhaps you understand. If not, no worries, we can help

    No worries Mate ! I don’t need your help. However, you obviously need mine.

  39. WKV

    “As it turns out, Illustrator does not know how to parse the /Rotate command for /Page objects (generally called Artboards in Illustrator). When Preview saves a pdf, it gets rid of the /Rotate command and instead rotates all the images individually, which Illustrator can parse. In fact, this may be part of why Preview modifies the structure of the pdf: to make sure that Illustrator can properly view the file.”

    According to the Xerox / Preview workflow, Illustrator never sees the Rotate 270 command because Preview removes it. Instead what Illustrator sees is the print to PDF file that Preview created. So if there is any problem with the image as seen in Illustrator it’s Preview’s fault and not Xerox’s fault. I am not at all surprised that Preview does not produce a PDF that is compatible with Illustrator. Preview is to Illustrator as Adobe Reader is to Illustrator.

    So the bottom line is that the Xerox “Rotate 270″ may create a problem for Preview but it can’t affect the PDF image as seen in Adobe Illustrator.

    So your frequent posts on the Rotate 270” problem in Illustrator are just diversions because your workflow says so.

    Moreover rather than scanning to E-Mail, if instead the Xerox scanned directly to Adobe Acrobat then compatibility with Illustrator would be a slam dunk.

  40. “”And to think that you and NBC claim to be the PDF code masters.””

    Just relative to others my friend, just relative to others. To be true PDF code masters may take years of effort, but to understand some of the basic principles, well..

    The Clipping Mask and the Graphic-state are are separate branches in the PDF tree structure. They are at the same level in the tree and are connected to a common node. The common node connects directly to obj [7 0 R]. The Color Space and Image Stream are together in a separate branch which also connects directly to obj [7 0 R].

    And that’s just the way it is.

    Obj 7 is the DCTDecode encoded JPEG. Now that you have finally located the clipping mask and come to realize that it is not a real object in the PDF document, you may come to understand better the structure of the document.

    Since you are new to all this, let me walk you through the structure in small steps.

    The trailer indicates that the root of the PDF document can be found in obj 31, which is the Catalog object. It refers to Obj 3 the Pages object which contains the structure for the pages of the PDF. It contains 1 page, and a child Obj 2.

    Obj 2 is the real interesting object which defines the resources (color space and image objects) as well as the content object (obj 4)

    Obj 4 contains a stream of PDF commands which define the layout of the page

    It starts by defining a clipping mask. Under the clipping mask it defines 9 image ‘layers’, each containing an image and a mask defining where it is to be rendered. All image objects are at the same level.

    Nothing earth shattering really as we find the same structure in other Xerox created documents that were saved using Preview.

    Thanks for supporting my findings. Always helpful as ever…

  41. My findings were completely independent of Illustrator. I used two other tools both of which are code-level PDF parsers.

    So it is operator error then?

  42. According to the Xerox / Preview workflow, Illustrator never sees the Rotate 270 command because Preview removes it. Instead what Illustrator sees is the print to PDF file that Preview created. So if there is any problem with the image as seen in Illustrator it’s Preview’s fault and not Xerox’s fault. I am not at all surprised that Preview does not produce a PDF that is compatible with Illustrator. Preview is to Illustrator as Adobe Reader is to Illustrator.

    Illustrator is unable to handle the Rotate 270 command, which is why it shows up in landscape. A good thing that Preview rewrites the layout to make it more compatible with Adobe…🙂

    So the bottom line is that the Xerox “Rotate 270” may create a problem for Preview but it can’t affect the PDF image as seen in Adobe Illustrator.

    And yet, Preview displays the document correctly and Illustrator does not..

    So your frequent posts on the Rotate 270″ problem in Illustrator are just diversions because your workflow says so.

    Nope they are not diversions, they are facts

    Moreover rather than scanning to E-Mail, if instead the Xerox scanned directly to Adobe Acrobat then compatibility with Illustrator would be a slam dunk.

    If the moon were made of cheese, astronauts would have carried lots of wine with them. Perhaps our friend is not familiar with network connected devices… But in the end, the facts show that Xerox documents are misinterpreted by Illustrator, which is a well known problem.

    Sorry that it led you to be confused about some of your ‘findings’. As I said, Illustrator is not a very good tool when it comes to understanding.

  43. According to the Xerox / Preview workflow, Illustrator never sees the Rotate 270 command because Preview removes it. Instead what Illustrator sees is the print to PDF file that Preview created. So if there is any problem with the image as seen in Illustrator it’s Preview’s fault and not Xerox’s fault. I am not at all surprised that Preview does not produce a PDF that is compatible with Illustrator. Preview is to Illustrator as Adobe Reader is to Illustrator.

    Let’s review the workflow yet again:

    1) Sheet is scanned by a Xerox
    2) Original PDF file created by Xerox has a /Rotate command (other than 0)
    3) Original PDF file, when opened in Illustrator, does not rotate properly
    4) Original PDF file is opened in Preview and rotates properly
    5) Preview Prints to PDF to create new PDF file
    6) New PDF file created by Preview does not have /Rotate command
    7) New PDF file rotates properly in Illustrator.

    So your claim is completely backwards – Preview fixes the Xerox PDF so that Illustrator can display it properly.

  44. So your claim is completely backwards – Preview fixes the Xerox PDF so that Illustrator can display it properly.

    Leave it up to Hermitian to get confused about all this… But perhaps he feels that he has been betrayed by his expensive toy🙂

  45. NBC says:

    August 19, 2013 at 05:18

    “”Hermitian: For the WH LFCOLB, the alignment with the 8 x 8 blocks @ 300 PPI x 300 PPI is satisfied for the upper-left origin or for the upper-right origin. For the 16 x 16 blocks @ 300 PPI x 300 PPI the alignment is satisfied for the upper-left origin but not for the upper right origin. The alignment to the 8 x 8 blocks @ 150 PPI x 150 PPI is also satisfied for the upper-left origin but not for the upper-right origin.””

    “Sigh.. but you claimed that it did not work for the 7355 documents. Where you wrong?… Very likely… But will you admit to it? Very unlikely…”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    As I previously stated, when I posted my earlier finding, that because you had not responded to my questions about your placement of the origin point, I would then assume that you used the default position. The default position in Illustrator is the upper-left corner of the artboard. The image objects do not align with the 8 x 8 blocks for this position of the origin point.

    My subsequent analysis placed the global origin point at the upper-left corner of the background page.

  46. As I previously stated, when I posted my earlier finding, that because you had not responded to my questions about your placement of the origin point, I would then assume that you used the default position. The default position in Illustrator is the upper-left corner of the artboard. The image objects do not align with the 8 x 8 blocks for this position of the origin point.

    Yawn.. Still confused… We are talking about relative distances to the edges and I have shown how they meet the 8×8 bit alignment, not just by showing this in Illustrator but also by showing this using direct calculations of the PDF information obtained from the cm transformation matrix.

    You need to understand what you are doing and it seems that you are still struggling but you have confirmed the 8 bit alignment as far as I can tell.

    I thank you for that my friend.

    Furthermore, you have to be careful with the alignment as you have to use the correct reference point in the image, and remember how it is originally stored. In other words, when you see a rotated version, you need to do it relative to the top left corner in the unrotated version…

    I thought that this was self evident… Sigh… This is how you can also determine the orientation of the document when scanned. Think about it and let me know if you need more hand holding.

  47. As I previously stated, when I posted my earlier finding, that because you had not responded to my questions about your placement of the origin point, I would then assume that you used the default position. The default position in Illustrator is the upper-left corner of the artboard. The image objects do not align with the 8 x 8 blocks for this position of the origin point.

    Except you were claiming that they didn’t work even before you asked for his origin point (which he actually told you several times, but you refused to listen because he used a different method than you wanted to use). So your failure dates to before that.

  48. NBC”

    “And yet, Preview displays the document correctly and Illustrator does not.”.

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    Don’t you mean that Preview displays the image correctly after the operator rotates it by 180 degrees ?

    So without this manually applied rotation the image is upside-down. In illustrator, without a 90 degree counterclockwise rotation the image is in landscape orientation.

    So I think you meant to say that your Xerox forger is not compatible with either Preview or Illustrator.

    However, because in your workflow, Illustrator never sees the Xerox scan to PDF file, your harping on the Illustrator problem with the Xerox Rotate 270 command is just a BIG RED HERRING !

  49. Except you were claiming that they didn’t work even before you asked for his origin point (which he actually told you several times, but you refused to listen because he used a different method than you wanted to use). So your failure dates to before that.

    I see it that way as well

  50. So I think you meant to say that your Xerox forger is not compatible with either Preview or Illustrator.

    Nope, it works in preview.

    However, because in your workflow, Illustrator never sees the Xerox scan to PDF file, your harping on the Illustrator problem with the Xerox Rotate 270 command is just a BIG RED HERRING !

    And yet you raised it as an objection before you apparently understood the workflow outlined to you. So the red herring is purely a creation of your own my friend. Your confusion not mine…

  51. W. Kevin Vicklund says:

    August 22, 2013 at 18:42

    “”As I previously stated, when I posted my earlier findings, that because you had not responded to my questions about your placement of the origin point, I would then assume that you used the default position. The default position in Illustrator is the upper-left corner of the artboard. The image objects do not align with the 8 x 8 blocks for this position of the origin point.””

    “Except you were claiming that they didn’t work even before you asked for his origin point (which he actually told you several times, but you refused to listen because he used a different method than you wanted to use). So your failure dates to before that.”

    HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
    So you must be referring to your earlier comments about NBC’s method not requiring an origin point. You said that his method avoided measuring absolute distances (you even included the equations). Of course anyone with the most remedial knowledge of Cartesian Coordinate systems would know that your comments were total balony (bologna). My graphical proof nails it.

    Grownups are not going to sit around waiting on you two to get your act together. Especially when the grownups know that you are just blowing smoke.

  52. Don’t you mean that Preview displays the image correctly after the operator rotates it by 180 degrees ?

    Only if it was scanned upside down and didn’t have large amounts of graphics (such as a basket-weave background). But in any case, Preview accurately renders the instructions in the Xerox PDF, which Illustrator doesn’t do.

    However, because in your workflow, Illustrator never sees the Xerox scan to PDF file, your harping on the Illustrator problem with the Xerox Rotate 270 command is just a BIG RED HERRING !

    You are the one harping on the Illustrator problem. But I’m glad to see you’ve finally admitted that you were dragging a red herring.

  53. So you must be referring to your earlier comments about NBC’s method not requiring an origin point. You said that his method avoided measuring absolute distances (you even included the equations). Of course anyone with the most remedial knowledge of Cartesian Coordinate systems would know that your comments were total balony (bologna). My graphical proof nails it.

    ROTFL… Hermetian is still confused. His complaints of where the origin was located had no relevance. All that matters if the distance obtained is divisible by 8 or not. And as I have shown quite convincingly, it does.

    And Hermitian has been quite helpful in supporting my findings, although he is still struggling with the idea as the cause of the 8 bit alignment and how to measure it properly.

    Note that there are two sides that align and two that typically do not align. The location of the sides depends on how the image was originally captured and how it is stored inside the document.

    I am starting to think that Hermitian still has not solved my earlier riddle, but yes, you can determine the scanning direction by looking at which sides show the 8×8 bit alignment.

    If Hermitian still needs more information here, I am more than willing and able to do so, as long as I understand the sources of his confusions.

    Your graphical proof was after my graphical proof ‘nailed it’ ROTFL. Until then you appear to have been struggling with properly following the step to interpolate to 300 ppi before attempting to align the grid.

  54. So you must be referring to your earlier comments about NBC’s method not requiring an origin point. You said that his method avoided measuring absolute distances (you even included the equations).

    Once again, Hermie lies about what I’ve said. At this point, I have to assume it’s deliberate. What I said was that you didn’t need to know the origin point, because NBC said his method was measuring absolute distances based on coordinates with an arbitrary origin. So long as the origin point doesn’t change, all you need to know is the point you’re measuring from, which NBC stated. All you proved with your graphics was that if you choose as your origin the point you’re measuring from, the grids will line up with the block boundaries (provided you’ve set everything else up properly, including moving everything!)

  55. NBC

    “ROTFL… Hermetian is still confused. His complaints of where the origin was located had no relevance. All that matters if the distance obtained is divisible by 8 or not. And as I have shown quite convincingly, it does.”

    Ok Hotshot! Here’s a challenge for you. It also will require hardly any effort on your part. Load up the Xerox 7535 scan to PDF in Illustrator and using the default origin point read off the x, y and W, H from the info panel for each object and then check for the 8 MOD 0 condition. And report back your findings.

  56. Ok Hotshot! Here’s a challenge for you. It also will require hardly any effort on your part. Load up the Xerox 7535 scan to PDF in Illustrator and using the default origin point read off the x, y and W, H from the info panel for each object and then check for the 8 MOD 0 condition. And report back your findings.

    with or without preview?

    I have done the former some time ago…

    https://nativeborncitizen.wordpress.com/2013/08/11/wh-7535-8-bit-alignments/

    The part you may have missed is the offset

    Note that the jpeg extends -3.36 and -1.2 outside the clipping mask or 14 pixels by 5 pixels in a 300 PPI frame.

    It can happen to the best of us…

  57. W. Kevin Vicklund says:

    August 22, 2013 at 19:15

    “”So you must be referring to your earlier comments about NBC’s method not requiring an origin point. You said that his method avoided measuring absolute distances (you even included the equations).””

    “”Once again, Hermie lies about what I’ve said. At this point, I have to assume it’s deliberate. What I said was that you didn’t need to know the origin point, because NBC said his method was measuring absolute distances based on coordinates with an arbitrary origin. So long as the origin point doesn’t change, all you need to know is the point you’re measuring from, which NBC stated. All you proved with your graphics was that if you choose as your origin the point you’re measuring from, the grids will line up with the block boundaries (provided you’ve set everything else up properly, including moving everything!)””

    Actually my previous comment pretty much captured your last comment.

    “What I said was that you didn’t need to know the origin point, because NBC said his method was measuring absolute distances based on coordinates with an arbitrary origin.”

    The generally accepted physical meaning of the term “absolute distances” is when the origin point is set at one of the distant fixed stars. Other coordinate system such a Sun centric and Geocentric are not absolute systems.

    NBC: That is so meaningless.

    So one never gets absolute distance from an arbitrary origin point.

    So the proper way to test the modulo hypothesis is as follows. Given a fixed grid of 8 x 8 or 16 x 16 blocks is the appropriate modulo condition satisfied for a particular rectangular object ? The answer depends on the relative position of each object relative to the fixed grid. The modulo criteria are satisfied if the sides of the rectangular object are congruent with the major grid lines.

    The test works because only one grid of 8 x 8 blocks and one grid of 16 x 16 blocks can be congruent to at least two perpendicular sides of each fixed rectangular object such that the included corner is an intersection point of two major grid lines.

    See: http://www.scribd.com/doc/161641778/Xerox-7535-Wc-Block-Alignments

    NBC: Thanks for confirming that the items align at 8 bit boundaries

  58. So one never gets absolute distance from an arbitrary origin point.

    No you get distance from two points, one is the corner of interest, the other the coordinate of the parallel side.

    Sigh… The origin has no impact on this… This is basic stuff my friend.

    Did you forget to add the offset by any chance? Is that it?

  59. NBC says:

    August 22, 2013 at 21:14

    “Ok Hotshot! Here’s a challenge for you. It also will require hardly any effort on your part. Load up the Xerox 7535 scan to PDF in Illustrator and using the default origin point read off the x, y and W, H from the info panel for each object and then check for the 8 MOD 0 condition. And report back your findings.””

    “with or without preview?”

    Either one. Take your pick.
    Then come back to report your failure to satisfy either modulo condition.

    “I have done the former some time ago…”

    “https://nativeborncitizen.wordpress.com/2013/08/11/wh-7535-8-bit-alignments/

    “The part you may have missed is the offset

    Note that the jpeg extends -3.36 and -1.2 outside the clipping mask or 14 pixels by 5 pixels in a 300 PPI frame.

    If you include the offsets, then you are not using the default origin point at the upper-left corner of the artboard to set the grid. Try it again with the rulers origin point in the default position and with two perpendicular grid lines coincident with the default origin point.

    So you actually didn’t work my challenge Dude. You worked a different grid. But the worse part is that you still don’t know that you worked a different grid than the one specified in my challenge.

  60. If you include the offsets, then you are not using the default origin point at the upper-left corner of the artboard to set the grid. Try it again with the rulers origin point in the default position and with two perpendicular grid lines coincident with the default origin point.

    You are sooo clueless… Hilarious… And no it is not the distance to the origin that I am interested in.

    Sigh… Did you miss all my 8 bit alignment postings which even showed pictures🙂

    So so clueless…
    The 8 bit alignment stands, it’s that simple, even you seem to agree.

  61. NBC says:

    August 22, 2013 at 22:11

    “So one never gets absolute distance from an arbitrary origin point”.

    “No you get distance from two points, one is the corner of interest, the other the coordinate of the parallel side”.

    “Sigh… The origin has no impact on this… This is basic stuff my friend.

    “Did you forget to add the offset by any chance? Is that it?

    NBC is a dense one for sure!

    You can always fool yourself in believing anything if all you care about is your storyline.

    But the facts are that every Cartesian grid has by definition an origin point. If you place the 8 x 8 grid origin point at the upper corner of the artboard then the grid is in the default position and the rectangular objects do not align with the major grid lines. However if you place the 8 x 8 grid origin point at he global position which is the upper-left corner of the outermost rectangular object then the top and left sides of the Xerox scan to PDF objects do align with the major grid lines.

    So it’s all about the position of each object with respect to each fixed grid.
    So any old grid origin point does not do the trick.

    End of Story !

  62. But the facts are that every Cartesian grid has by definition an origin point.

    Who cares, I am interested in distance to the edge of the image, not some origin.

    Clueless.

  63. So it’s all about the position of each object with respect to each fixed grid.
    So any old grid origin point does not do the trick.

    Only if you want to visually align them, but that is the same as adding the offset…

    ROTFL… slow learner..

  64. The easiest method for re-positioning the rulers origin point in Illustrator is to place the selection tool at the intersection point of the screen rulers and drag the rulers intersection point to the desired new location for the origin point.

    Hence to move the origin point from the default position (upper left corner of the artboard) to the upper-left corner of the background page you drag the intersection point of the rulers from the upper-left corner of the screen to the corner of the background page.

    Notice that you do not drag the origin point from the default position (at the upper-left corner of the artboard) to the new global position at the upper-left corner of the background page.

    It really isn’t complicated — after you know how to do it.

  65. NBC says:

    “August 22, 2013 at 23:36

    “”But the facts are that every Cartesian grid has by definition an origin point.””

    “Who cares, I am interested in distance to the edge of the image, not some origin.”

    “Clueless.”

    So if you select an arbitrary point to measure from and then you add the x0 and y0 offsets between the arbitrary point and the corner of the background page then you get the total distance from the corner of the background page to the measured point.

    But what you have really done is move the grid origin from your arbitrary reference point to the corner of the background page.

    It’s that graphics part that you don’t get.

    But fortunately your ignorance doesn’t alter the facts.

    I wonder if all Obots are spatially deprived like NBC ?

  66. NBC says:

    August 22, 2013 at 23:30

    “Check out the work I have done in this area. You may want to take notice”

    “We are way ahead of you, as usual.”

    I’ll check out your work as soon as you post the PDF files Dude. Every day that you sit on your results and continue to belly ache about Zullo and the Posse holding their cards tight you lose a little more credibility.

    I believe in trust but verify.

    Remember Cold Fusion. I know the guy who was at the center of that fiasco. He still believes that he produced a fusion reaction in a liquid at near room temperature.

  67. NBC says:

    August 22, 2013 at 23:21

    “”If you include the offsets, then you are not using the default origin point at the upper-left corner of the artboard to set the grid. Try it again with the rulers origin point in the default position and with two perpendicular grid lines coincident with the default origin point.””

    You are sooo clueless… Hilarious… And no it is not the distance to the origin that I am interested in.

    What you should be interested is my question concerning the overhang of your Xerox 7535 / Preview PDF image beyond the artboard extents. The overhang of your image is five pixels on each long side and 14 pixels on the short sides.

    As a point of reference the overhang of the WH LFCOLB is one pixel on the long sides and two pixels on the short sides. This is the minimum overhang that is both larger than the 8.5 in x 11.0 in. artboard and satisfies both the 8 MOD 0 and the 16 MOD 0 for the top and right edges of the rectangular boundary of the mostly text layer.

    Sigh… Did you miss all my 8 bit alignment postings which even showed pictures

    You must be referring to those screen shots with the red rectangles that were shot at maybe 200% Zoom ratio. Those shots are worthless. For pixel resolution of 300 PPI x 300 PPI the major grid lines are not accurately displayed below 800% Zoom ratio. The minor grid lines are not accurately displayed below 3200% Zoom ratio.

    See: http://www.scribd.com/doc/161641778/Xerox-7535-Wc-Block-Alignments

    “So so clueless…”

    ” The 8 bit alignment stands, it’s that simple, even you seem to agree.”

    Yes I agree that the alignment to 8 x 8 blocks is restricted to one grid.

  68. NBC says:

    August 22, 2013 at 18:59

    “”So I think you meant to say that your Xerox forger is not compatible with either Preview or Illustrator.””

    “Nope, it works in preview.”

    Then that means that if it doesn’t load correctly in Illustrator then Preview is at fault.

    Also you never answered the obvious question.

    Which is —

    How does the Xerox scanner detect that the paper original was placed upside-down on the glass and then act to add the Rotate 270 command to correct the incorrect orientation?

Comments are closed.