A software bug that caused some characters to be substituted for others in scans by some Xerox machines is more serious than previously thought.
The problem came to light last week when German computer science student David Kriesel noticed some figures printed in a small, fine font on a document were incorrectly copied into a PDF file output by the machine. Numbers from one part of the document had been reproduced in another part, apparently due to the way the JBIG2 compression system works.
Xerox originally said the bug only occurred at the lowest quality and highest compression setting and last week advised users to avoid those settings. On Tuesday, the company said setting the quality higher doesn’t totally eliminate the problem.
“After further testing of the scanning function, we’ve now determined the factory default and highest modes do not completely alleviate the problem of substituting characters on ‘stress documents.’ The default and highest modes do substantially reduce the likelihood of character substitution but due to a software bug character substitution is not completely eliminated,” Xerox said in a statement.
Xerox defines “stress documents” as those that combine “small fonts, are hard to read, contain stray pixels and/or have been scanned multiple times.”
The company’s advice to users remains unchanged: Reset the scanner setting to the default and, when it becomes available, apply a software patch. The patch is due in the next few weeks and will eliminate the problem at all scanner settings, Xerox said.
“Regardless of the document condition, we are committed to address any problem even if it is something our customers may never encounter,” the company said.
Xerox said the problem affects machines in four product families: ColorQube models 87XX, 89XX, 92XX and 93XX; WorkCentre models 5030, 5050, 51XX, 56XX, 57XX, 58XX, 6400, 7220, 7225, 75XX, 76XX, 77XX and 78XX; the WorkCentrePro 2XX and the BookMark 40 and 55.
The problem doesn’t occur on copies or faxes, according to Xerox.