1
$\begingroup$

I am somewhat new to this. Forgive if my description is not clear.

I have two code blocks. I know that each block, in part, codes for a serial number (which I know). Part of the code probably relates to a model number (which I know).

Each block is a 9x9 array. However, I believe the close two dots at bottom right are anchors and orient the code as seen in the attached images. Therefore, I believe the coding array to be 7x8 extending up and left from this location. The columns could be two 4 bit or one 8 bit?

It is possible that the serial comes from a (proprietary) lookup table and will not be directly gleaned from the code blocks.

Image 1 has encoded serial P9640300510 for an Aficio CL7000 Ricoh. Image 1

Image 2 has encoded serial C747LA00011 for MPC5504ex Ricoh. Image 2

Does anyone have ideas on extracting these codes?

edit to add: Image 1 is the upper left. Image 2 is the lower right.
edit to respond to comments: Hi @Lanny Strack , Yes, those are good papers and I have them both. That's right, I believe you sum up the papers well - the coding portion is the 7 column by 8 row grid. The two close dots at bottom right are to orient the code. @retzler has placed the coding into an array. If we look at Peter Buck's paper, (https://www.researchgate.net/publication/325976319_Reverse_Engineering_the_Machine_Identification_Code ) we can, perhaps obtain some clues on how to proceed with a possible decode. However, as I said in the OP, there may be a lookup table involved (which would dead end us).

$\endgroup$
  • 2
    $\begingroup$ Is it a dot-to-dot puzzle? :'D $\endgroup$ – William Pennanti May 16 at 8:31
  • 1
    $\begingroup$ I believe this refers to the hidden dot patterns used by color copiers to leak date and serial number. $\endgroup$ – retzler May 23 at 13:34
  • 1
    $\begingroup$ I found two documents online which discuss the Ricoh dot pattern: a scholarly article from Egypt (ejfs.springeropen.com/track/pdf/10.1186/s41935-019-0140-8), and a document from Duke University in the US (people.duke.edu/~ng46/collections/…). Both conclude that the dots appear in a 7x8 grid, with two additional alignment dots in one corner of that grid. The Egyptian paper concluded that the device serial number is encoded (but no timetstamp), but do not provide specific evidence of that in the above paper, nor any kind of decoding method. $\endgroup$ – Lanny Strack May 23 at 18:20
  • 1
    $\begingroup$ OP: (1 of 2) I looked at the Peter Buck paper you linked to, and from there located and downloaded the EFF dataset, which includes printouts from Ricoh s/n P9640300510, the printer depicted in your image 1. In fact, if you look at a full-sheet printout, and pay attention to the various grids printed across the page, they are similar to each other, but NOT identical, as is noted in the Duke paper (the grids on the Ricoh print shown in the Duke paper vary from 17 to 22 dots). $\endgroup$ – Lanny Strack May 25 at 2:56
  • 1
    $\begingroup$ (2 of 2) If you want to make a serious attempt at deciphering, I think you would need to note all the various grids printed on a page, and pay attention to the parts they have in common, and the parts that differ. Regardless, with data from only a few different printers, I think your chances of being able to decipher the code are slim. $\endgroup$ – Lanny Strack May 25 at 2:57
1
$\begingroup$

(Community evidence locker— feel free to add and revise)

A possible calibration of the mystery codes:

Successful identification (copied with permission from a comment by retzler):

I believe this refers to the hidden dot patterns used by color copiers to leak date and serial number. Cf. this Wikipedia Article.


WɪᴋɪᴘᴇᴅɪA


Richter, Timo; Escher, Stephan; Schönfeld, Dagmar; Strufe, Thorsten (June 14, 2018). “Forensic Analysis and Anonymisation of Printed Documents.” Proceedings of the 6th ACM Workshop on Information Hiding and Multimedia Security. ACM: 127–138. doi:10.1145/3206004.3206019. ISBN 9781450356251.

Published decoding attempts (copied with permission from a comment by Lanny Strack):

I found two documents online which discuss the Ricoh dot pattern: a scholarly article from Egypt, and a document from Duke University in the US. Both conclude that the dots appear in a 7x8 grid, with two additional alignment dots in one corner of that grid. The Egyptian paper concluded that the device serial number is encoded (but no timestamp), but do not provide specific evidence of that in the above paper, nor any kind of decoding method.


Ahmad S. Salim and Asmaa A. Abdalla. “The determination of identity and uniqueness of color laser printouts of Ricoh® brand by Adobe® CreativeCloud Photoshop® 2018.” Egyptian Journal of Forensic Sciences. (2019) 9:40.


“Steganography - Color Laser Copier Dot Codes.” Duke University.

Ricoh Aficio MPC 4501 / Code: D540-94 / SN: W2008404434

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.