JPEG – The Gory Details – Part 4 – A Xerox PDF with layers

I found a great document [Work In Progress]

Found another one with JPEG background and text monochrome bitmaps

The document shows how a PDF scanned by a Xerox Workstation is ‘broken down’ into a background JPEG and multiple monochrome bitmaps containing mostly textual information.

So let’s look at the document

Note:

1. So far no evidence of background being at half the resolution of the foreground bitmaps

2. I have not yet looked at the alignment of the layers at 8 bit boundaries

3. No JBIG2 compression

4. The document is in A4 format 210 mm wide and 297 mm high,  or 8.3” x 11.7”

Page one has the following  XObjects

obj 6 0
 Type: 
 Referencing: 11 0 R, 17 0 R, 15 0 R, 29 0 R, 25 0 R, 19 0 R, 27 0 R, 7 0 R, 21 0 R, 9 0 R, 13 0 R, 23 0 R

  <<
    /ProcSet [ /PDF /ImageB /ImageC /ImageI ]
    /XObject
      <<
        /Im3  11 0 R
        /Im6  17 0 R
        /Im5  15 0 R
        /Im12 29 0 R
        /Im10 25 0 R
        /Im7  19 0 R
        /Im11 27 0 R
        /Im1   7 0 R
        /Im8  21 0 R
        /Im2   9 0 R
        /Im4  13 0 R
        /Im9  23 0 R
      >>
  >>

The JPEG background

obj 7 0
 Type: /XObject
 Referencing: 8 0 R, 31 0 R
 Contains stream
<<
 /Length           8 0 R
 /Type             /XObject
 /Subtype          /Image
 /Width            1170
 /Height            827
 /ColorSpace       31 0 R
 /BitsPerComponent 8
 /Filter           /DCTDecode
 >>

The JPEG is 827 by 1170

The document is 8.27 × 11.7 inches 100dpi

And our familiar Obj 4

obj 4 0
 Type: 
 Referencing: 5 0 R
 Contains stream
  <<
    /Length 5 0 R
    /Filter /FlateDecode
  >>f

 'q Q q 
0 0 595 842 re W n 
q 0 842.4 -595.44 0 596 0 cm /Im1 Do Q
0.0431 sc 
q 0 223.2 -451.44 0 535.52 351.36 cm /Im2 Do Q 
0.0941 sc 
q 0 111.96 -129.96 0 545.24 668.16 cm /Im3 Do Q 
0.051 sc 
q 0 120.6 -108.72 0 189.92 195.84 cm /Im4 Do Q 
0.3059 sc 
q 0 34.2 -146.16 0 221.6 763.2 cm /Im5 Do Q 
0.1961 sc 
q 0 40.32 -136.8 0 223.76 673.92 cm /Im6 Do Q 
0.5137 sc 
q 0 139.68 -14.04 0 95.24 408.96 cm /Im7 Do Q 
0.0392 sc 
q 0 24.48 -65.88 0 461 604.8 cm /Im8 Do Q 
0.0471 sc 
q 0 24.84 -43.2 0 337.52 604.8 cm /Im9 Do Q 
0.0392 sc 
q 0 22.68 -51.48 0 135.56 607.68 cm /Im10 Do Q 
0.0275 sc 
q 0 7.92 -38.88 0 229.52 622.08 cm /Im11 Do Q 
0.898 sc 
q 0 119.16 -48.96 0 539.12 610.56 cm /Im12 DoQ 
Q'

Let’s first look at page 1

Bygningsforsikring - police/Im1

obj 7 0
 Type: /XObject
 Referencing: 8 0 R, 31 0 R
 Contains stream
  <<
    /Length           8 0 R
    /Type             /XObject
    /Subtype          /Image
    /Width            1170
    /Height            827
    /ColorSpace       31 0 R
    /BitsPerComponent 8
    /Filter           /DCTDecode
  >>

Not very interesting but it is a JPEG. No those are not smears on your screen…

danish-000

/Im2

danish-001

/Im3

danish-002

/Im4

danish-003

/Im5

danish-004/Im6

danish-006/Im7

danish-005/Im8

danish-007/Im9

danish-008/Im10

danish-009

/Im11

 danish-010

/Im12

danish-011

6 thoughts on “JPEG – The Gory Details – Part 4 – A Xerox PDF with layers

  1. IM12 is part of a graphic design or watermark that is on the right side corner of the first page. Half of the design was put on its own layer (IM12) and the rest was put on the background layer.

    Did the software rotate these images 90 degrees?

  2. Sorry that should read “Im1 (the smears) contains the rest of the design that goes with Im12.”

    There are other stuff on Im1

  3. My one concern with these documents is that the bitmap layers are all within the grayscale. The entire document is within the grayscale. The Obama Longform BC’s bitmap layers are not in the grayscale spectrum. Can you find any documents that do that too? A colored background would help. The Obama BC still has those two top layers that have background colors behind it. I’d like to see anything similar. Good work though!

  4. Some spot calculations show that these are being rotated counter-clockwise.

  5. Yes, there are still some issues to address. However, finding a document that is segmented into a jpeg background layer and several monochrome bitmaps already places quite a bit of doubt on the work by the Cold Case Posse.
    Add to this the embedded comment in the jpeg which was found in similar documents from a Xerox Workcentre and you already have a strong case. Still some questions remain, such as the masking layer etc. I have found at least one with a color background but not as colorful by any standard as the one in the whitehouse PDF.

    I will have to do some experiments here. Or rely on others…

Comments are closed.