Hacker News Reader

I skimmed the article and can confidently say this is garbage science. The microscopic parameter changes inherent in every print (across printer types) means that any sort of unique ID (the type of ID which would be required in a court of law) on a single printer would be impossible. Using the word "fingerprint", which is essentially a UUID for humans, as a stand-in for "identifying a class of printers" is hugely misleading

View on HN

"The microscopic parameter changes [are] inherent in every print" seems like a weak argument for claiming that the article is garbage science. No two fingerprints are pixel-perfect, and DNA samples can get damaged, but this doesn't mean that such methods are invalid - in fact, they can be valid even in forensic contexts.

Why is it garbage science? We can match an unaltered digital image to a camera because of minor differences in the CCD/CMOS sensors why is such a stretch that it would be possible to find a forensic finger print to tie a print to a printer?

I know a few hardcore 3D printer enthusiasts that can identify with a fairly decent accuracy the printer a print came from even pretty wrk estimate the settings and even the software used for slicing based on how the layers are set.

Reply on news.ycombinator.com