It appears that the answer may be when you use a Xerox scanner to produce the copy! According to a story reported today, the “normal” scan settings on at least two models of Xerox scanner actually change the characters that appear on the copy that is generated. The compression software used by the machine – JBIG2 – is tuned to produce smaller files at the expense of image quality. Although OCR functionality is apparently not enabled, the machine can misinterpret certain characters under these conditions and replace them with others. A test completed by German researcher Dr. Kriesel showed a 6 replaced by an 8, and vice versa. Currently, it is not known if this issue is present in other scanners using the same compression software.
I think most of us are aware of the potential issues when using file compression software but I’m guessing that we all assume that a scanner or photocopier will accurately copy the content from an original document to the copy that is being generated, albeit perhaps at a lower resolution. This has significant repercussions for the life sciences industry who rely on data from preclinical and clinical studies being accurately reported – even when those reports are scanned or photocopied. At the moment this problem is not considered to be widespread but perhaps we should not always take so much for granted the accuracy of the technology that we are increasingly reliant upon.