Xerox 7655 Additional Experiments

I received documents captured on a Xerox 7655 WorkCentre which involved the normal workflow. The original document was scanned Right Side Up (RSUP) and Up Side Down (USD). I extracted the pictures and observed that

  1. RSUP images are rotated 90 degrees clockwise
  2. USD images are rotated 90 degrees counter clockwise
  3. USD images when the PDF is rotated in Preview 180 degrees are still rotated 90 degrees counter clockwise
  4. The Quantization Matrices do not match the WH LFBC, and the WH Tax matrices but this can be explained by realizing that the scanner is not using the up-to-date firmware as it scans to PDF 1.4 not PDF 1.5 as found in the WH Tax form PDF.
  5. The RSUP contains 5 layers, the USD 4 layers
  6. They all contain the YCbCr comment tag
  7. The Xerox WorkCentre Firmware in question creates PDF version 1.4 documents

Relevant Data

#File: RSUPWH7655-000.jpg

Quantization table
  Precision=0; Table index=0 (luminance)
       9  12  12  15  15  15  18  18
      18  18  23  22  21  22  23  28
      27  25  25  27  28  36  33  31
      30  31  33  36  49  42  39  38
      38  39  42  49  55  47  44  43
      44  47  55  60  52  50  50  52
      60  67  59  59  59  67  76  69
      69  76  90  82  90 107 107 146
  Estimated quality level = 54.60%

Quantization table
  Precision=0; Table index=1 (chrominance)
      10  13  13  16  16  16  19  19
      19  19  24  23  22  23  24  30
      28  27  27  28  30  38  36  33
      32  33  36  38  53  46  42  40
      40  42  46  53  60  50  48  47
      48  50  60  65  56  54  54  56
      65  73  64  64  64  73  83  76
      76  83  98  90  98 118 118 162
  Estimated quality level = 50.84%
Average quality: 55.78% (56%)
JPEG Information

USDWH7655-000.jpg        1656 x 1280 24bit n/a N 307502 "YCbCr"
USDWH7655Rotated-000.jpg 1656 x 1280 24bit n/a N 307502 "YCbCr"
RSUPWH7655-000.jpg       1656 x 1280 24bit n/a N 306274 "YCbCr"

9 thoughts on “Xerox 7655 Additional Experiments

  1. 1. – 2.

    -90 – (+ 90) = – 180

    3.

    So now you are claiming that the Xerox 7535 is device independent with respect to the MAC OS Preview ?

    4.

    It’s past time for you to post a link to the Preview PDF file created from the Xerox PDF file that you posted before.

  2. What are the Quantization Matrices tables for the Norwegian Police PDF? It was run on a WorkCentre 7328 and the PDF is Version 1.3 created by on a Mac

  3. It’s past time for you to post a link to the Preview PDF file created from the Xerox PDF file that you posted before.

    ROTFL.. Does our ‘researcher’ who examined a preview created PDF not have access to a Mac?…

    Come on Hermitian, show us that you can do some independent work…

  4. So now you are claiming that the Xerox 7535 is device independent with respect to the MAC OS Preview ?

    I never made the claim that it wasn’t. I pointed out that Illustrator fails to properly render some documents…

  5. NBC says:

    July 26, 2013 at 15:59

    “”So now you are claiming that the Xerox 7535 is device independent with respect to the MAC OS Preview ?””

    “I never made the claim that it wasn’t. I pointed out that Illustrator fails to properly render some documents… ”

    And I pointed out in great detail how Adobe Illustrator CS6 and Adobe Illustrator CC both “properly render” all nine images that comprise the WH LFOLB PDF image. Also the WH LFCOLB PDF composite image loads in the correct portrait orientation in both Illustrator versions. Additionally, the WH LFCOLB PDF composite image loads in the correct orientation in Adobe Acrobat XI Pro, Adobe Reader XI, PDFXChange Viewer Pro, PDF XChange Editor Pro, InkScape, Corel Fusion, Tracker PDF Tools 4.0 Pro and PDF Transformer 3.0 Pro.

    I have additional results from anaylsis of the Xerox 7535PDF wh-lfbc-scanned-xerox-7535-wc.pdf that I obtained by means of Tracker PDF Tools 4.0 Pro. I will post these confirming results as soon as I have completed the new analyses. The different analyses of this same file are in agreement with earlier results obtained with Adobe Illustrator CS6 and Adobe Illustrator CC that I have already posted.

  6. And I pointed out in great detail how Adobe Illustrator CS6 and Adobe Illustrator CC both “properly render” all nine images that comprise the WH LFOLB PDF image.

    Still missing the point. Yes, the WH LFBC and the Xerox preview saved PDF both render ‘correctly’ and both require rotations of the images from their landscape internal positions into the portrait orientation. You were confused about the Xerox PDF (without preview saving) opening up in landscape, and I showed why and how this is a real issue for Illustrator.

    If only you spent time reading what has been written, then you would not be so confused.
    Why you are spending time analyzing a Xerox document and not the preview saved one is beyond me. I am sure you will find the same as I have found.

    But it’s apples and oranges when you want to understand the WH LFBC which is a Xerox WorkCentre created document saved as PDF in preview.

    Is this really that hard to understand or accept that the two documents are virtually identical in all aspects, including the rotations, the separations, the embedded comment, the quantization matrices, the scaling factors, the differences in fore and background resolution and so on and so on.

  7. YCbCr Label “Smoking Gun” or “Slight of Hand?

    NBC

    “JPEG Information

    “USDWH7655-000.jpg 1656 x 1280 24bit n/a N 307502 “YCbCr”
    “USDWH7655Rotated-000.jpg 1656 x 1280 24bit n/a N 307502 “YCbCr”
    “RSUPWH7655-000.jpg 1656 x 1280 24bit n/a N 306274 “YCbCr””

    NBC has claimed a connection between his Xerox 7655 forger and the WH LFCOLB. But does his “so called smoking gun” originate with the Xerox or not ?

    Surprisingly the evidence suggests that it does not.

    First we have established beyond doubt that the Xerox 7655 Workcenter does not support the JPEG image format. We know that because the detailed specification document for the Xerox 7655 says so. Rather Xerox utilizes JPEG compression to compress JFIF formatted files. However, there is nothing within the specification document that indicates that the JPEG format is supported.

    I have also established a very odd result that adds further uncertainty to NBC’s claim of a smoking gun.

    I searched for the YCbCr label in NBC’s Xerox 7535 PDF and the search returned zero hits. However, I had previously searched the WH LFCOLB PDF for the YCbCr label and found one occurrence. Consequently, the PDF file that NBC claims is the source of the smoking gun label does not contain the label but the PDF file that he claims was the recipient of the label from the Xerox is the same format as the purported source file. This is a very strange result.

    So where does the YCbCr label first turn up in NBC’s convoluted workflow? As we see from NBC’s latest post, the label first turns up in his JPEGs which he extracts from the Xerox scan to e-mail PDFs with his freetoy extractor. But we now know that the Xerox 7655 does not support the JPEG format. And we also know that NBC’s extractor tool only extracts JPEG formatted files. So what does his extractor tool return when the PDF file does not contain a JPEG formatted file but instead contains a JIFF formated file. I don’t really know but my best guess is that his dumb extractor tools just does what it is programmed to do; hence it cobbles together the best JPEG formated file where one did not previously exist.

    As further proof that NBC may have pulled a fast one, I have recreated the Image for the background layer using state-of-the-art extraction tools that do not presuppose the image file format. In one case the image file was JFIF and in the other the image file was a Photoshop PSD image file. Neither of these files contain the YCbCr label. Additionally, the JFIF file contains all the HEX encoded and ASCII labels that are required by the JIF and JFIF standards.


    JFIF segment format [edit source]
    Field Size (bytes) Description
    APPO marker 2 Always equals OxFFEO
    Length 2 Length of segment excluding APPO marker
    Identifier 5 Always equals “JFIF” (with zero following) (0x4A46494600)
    Version 2 First byte is major version {currently 0x01), Second byte is minor version (currently 0x02)
    Density units 1 Units for pixel density fields • 0 – No units, aspect ratio only specified • 1 – Pixels per inch • 2 – Pixels per centimetre
    X density 2 Integer horizontal pixel density
    Y density 2 Integer vertical pixel density
    Thumbnail width (tw) 1 Horizontal size of embedded JFIF thumbnail in pixels
    Thumbnail height {th) 1 Vertical size of embedded JFIF thumbnail in pixels
    Thumbnail data 3 x hv* th Uncompressed 24 bit RGB raster thumbnail

    Consequently, the bottom line question is where did the YCbCr label that appears once in the WH LFCOLB really come from ?

    Was NBC’s JPEG extraction tool used in the creation of the WH LFCOLB ?

    Considering, that NBC has refused to release any of his proof, we will just have to figure this out for ourselves.

  8. NBC

    NBC, you keep dodging and dancing but you haven’t realeased any proof of your claims. You have only posted JPEG extracted results to support your claim of the YCbCr “smoking gun label”.

    And while you have continued to stonewall for weeks and weeks you and your Obot horde continue to lam-bast Zullo and his Posse for not releasing every scrap of evidence of the crime of forgery of an identity document.

    And all the time that you have waffled over your “simple” workflow, changing the physical rotation of the original paper document along with dubious manipulations applied by the Preview operator, my detailed results are all available on Scribd and the archived copy of birth-certificate-long-form.pdf that I have analyzed is readily available to anybody.

    And you haven’t even posted the proof that your Xerox MRC creation actually places the rectangular object boundaries where they are in the WH LFCOLB. You could have done this without releasing any of the Xerox files. And all the while my detailed tables which prove that the WH LFCOLB has the right stuff are available on Scribd.

    And you have absolutely no answer to the fact that the WH LFCOLB does not exhibit the landscape orientation problem in Illustrator which is present in all of the Xerox/Preview files. What is your logic for blaming Illustrator for your problem?

Comments are closed.