Private Printers

Is there such thing? I watched a video recently about how printers all print tiny codes on each paper that can be uniquely traced back to you, send info to the government, etc Are there any browsers that dont spy on your every moment? >inb4 just use digital not an option

>>15927 printers* not browsers my mind was wondering

If you are worried about microprint, IIRC the EFF used to have a list of printer models that do not print microprint. Public computers and printers provide some ability to hide who you are if you can access them without being observed. What kind of printer use are you looking at doing? inb4 printing out pohl memes to put up around your campus

if i remember rightly if you can get your hands on a good printer from the early 90's they don't have this issue

>>15930 What about ink? can you refill cartridges?

>>15929 I am working for a company in russia remotely in the USA and they have requested it

>>15927 >line printer >dot matrix >teletype printer pick one Modern stuff is cyberhumped and so is the EFF ( >>15929 )

this one's a little too schizo for me to reasonably believe without evidence

>>15937 https://www.eff.org/issues/printers

What about greyscale laser printers?

>>15927 Be realistic most of these old printers cannot even run on ancient hardware (10+ years of age) >>15935 >No colour >Loud >Slow

Buy a pen plotter, it draws only what color you give it and only what you have programmed into it.

>>15930 >>15933 >>15931 >>15937 It is real, but can be defeated. I hope you guys realize that there is software which is FOSS that can scramble the identifying dots.

>>15975 Which software? I didn't find much when looking it up that wasn't simply more articles about the tracking.

>>15978 OK so I dont remember it off the top of my head, but here's what I do know: >Here's a github link for something I just found: https://github.com/DerEchteKay/InvisibleTrackingDots >IOTBW threads (It's OK to be white) on pohl had a link to some better software. You can look them up in the archives, they get posted all around October. The threads have links because anons didn't want people to find out who they were >You can use lower-tech ways to defeat printer dot recognition by adding random noise to the printed page. You can also add the dots manually yourself, especially if you match the color or nearly match it. You can use software such as GIMP to do so. >If you take a page and copy it and then scan it and copy it, this effectively defeats the invisible tracking dot system Any other questions? Your answer is in the archives on the IOTBW threads.

I dont remember invisible dots but i do remember reading about printer changing letters slightly for the same purpose. Dude noticed it as an issue and wanted to fix the bug until he found out why... Tldr its not just dots.

>>15933 a...anon...that's not company, is KGB...

>>15994 This is easy. Send a bitmap file instead of postscript.

>>15997 its inbuild basic ocr, so format doesnt matter, it scans it like an image and changes where it thinks it found a letter it should change. Same reason got it spotted - dude is foreigner and his native character set triggered it where it shouldn't. Manufacturer likely only tested for local market which pwned them.

>>15998 Still no proofs. You sure it wasn't proprietary software on computer? Running OCR on every bitmap going through printer is an computationally intensive task.

>>15927 There you go. Hope that helps https://www.eff.org/pages/list-printers-which-do-or-do-not-display-tracking-dots

>>16003 Do you understand german? If yes i will dig it out.

>>16003 It runs on the hardware lol

I remember printers have firmware level detection like at some point in high school my idiot teacher tasked us to pick a foreign country and print their currency in a bond paper (back when sending digital documents weren't as prevalent) and to my surprise the printer rejected it a bunch of times, personnel had to edit and in the end the result was it had these vertical lines like the printer was out of ink and some kind of weird dots/blot somewhere in the paper could be literally seen although it wasn't exactly running out of ink since I tried printing other bunch of stuff. Stories aside, I've heard about places in Africa selling used printers for whatever purpose you need it be.

>>16005 I'll use a translator, always provide evidence without asking dumb questions. >>16033 Can it possibly recognize all world's currency, even unreleased bank notes?

Instead of looking up soykaf yourself you complain and complain. This took me 2 min to find https://www.dkriesel.com/en/blog/2013/0802_xerox-workcentres_are_switching_written_numbers_when_scanning

>>16037 >use Enterprise Grade Shitware™ with built-in mystery meat OCR last updated in 2005 so basically useless even for it's purpose >complain when millions of lines of enterprise-grade firmware code make it malfunction >gotta make a conference talk out of it >b-but we definitely switched OCR fuction off! (background laughter track)

if the secret dots are printed at the top or bottom you could always print on legal paper, which I believe is longer than the normal stuff, then cut the dots off.

>>16185 Considering it's a broader problem with JBIG2 that was simply first demonstrated at scale on Xerox machines, and not exclusive to Xerox machines, your post qualifies as little more than a low-effort soykafpost.

>>16185 >With industry meat OCR fuarrrking lost it, 10/10

>>15975 >>15981 I would have imagined the spy dots would have been implemented on the printer's firmware level. Clearly they're not yet there in whatever file you want to print (since the file doesn't know what printer you're going to print from), how would adding anything in GIMP help? >If you take a page and copy it and then scan it and copy it, this effectively defeats the invisible tracking dot system If standalone printers have microdots, wouldn't printers in copy machines have them as well?

>>17106 >how would adding anything in GIMP help? Simple. Add "tracking dots" to the page at random. When the alphabet men go to look at the tracking dots to figure out whose printer the manifesto you sent to the press came out of the real tracking information is obscured by visually indistinguishable added noise. The idea would also work by copying/scanning on another printer if the scanner is capable of picking up the tracking dot. Scanners probably aren't able to capture the dot but if they were the result would be a mash of 2 sets of tracking data overlaid which can't really be untangled.

>>15948 >What about greyscale laser printers? These could be fine. >>15935 These as well.

>>15937 >this one's a little too schizo for me to reasonably believe without evidence Even a cursory search of Google would demonstrate to you this is not "schizo" and it has a history going back over a century to franking devices. To comment that such steganographic watermarking methods mentioned in this thread are a mental illness is obtuse at best, and wholly insulting otherwise. Currently the most anonymous solution to this problem is a pen plotter. (Don't forget to add a fuzzing algorithm to consistently add an error and delete after printing) Whilst the methods mentioned >>15935 are valid they can also leave characteristic signatures on the printout. >>15954 >Be realistic most of these old printers cannot even run on ancient hardware Be realistic you are lacking the skill necessary to get most of these old printers running on ancient hardware >No colour Irrelevant to the desired objective of an anonymous print >Loud Irrelevant to the desired objective of an anonymous print >Slow Irrelevant to the desired objective of an anonymous print

>>15948 >What about greyscale laser printers? >>17132 >>What about greyscale laser printers? >These could be fine. From investigations by anons (a number of years ago in threads on this topic) we detected a form of frequency transform at glyph borders for b/w laser printers. TL;DR No they are not. Consider all except pen plotters compromised.

>>17135 >we detected a form of frequency transform at glyph borders Evidence or didn't happen. Things like these should be documented and published, why haven't you wrote to eff or something?

>>15929 >printing out pohl memes to put up around your campus Hey at least it got people to mind their privacy more so hey

>>17135 Could you please link any leads to this investigation for my purposes? I want to know if this is true

>>17123 >can't really be untangled To the contrary, given that the appropriate dot matrix is repeated several times, I imagine parsing away randomly added dots or distinguishing between original printer vs copier dots would be fairly simple with some analysis. I thought the dots were something smaller than the printer normally does, but appears that's not the case. So now to fool the glowies you'd have to add identical noise to every copy of the matrix that's printed. And also make sure that your dots cannot be distinguished from the tracking dots (exactly appropriate size, color, placement).

>>17170 >I imagine parsing away randomly added dots or distinguishing between original printer vs copier dots would be fairly simple with some analysis True for random dots, but if you know something about the format of the dots you should be able to add dots in the same format or close enough to prevent the decoder from telling the real data apart from the noise. I'm not sure if anybody's reverse engineered the format. Maybe even just scanning a printout from a different printer and then adding those dots as a background would confuse the decoder.

Has anyone seen this? Printing dots to cover the dots. https://github.com/dfd-tud/deda

Relevant site that explains these tracking dots; http://seeingyellow.com/ This tool allows you to upload a scanned document and identifies these dots; https://forensicdots.de/

I'm more concerned about printer memory. That's a big security risk Id imagine. >>17452 So the dots are yellow? Then couldn't someone just keep an empty yellow cartridge so it fails to print these dots?

>>17551 1. Color printer refuses to print without yellow cartridge. 2. Some Russian hacker probably wrote a firmware patch to skip checking yellow(or any) paint container volume for popular line of printers long time ago.

>>17561 Don't even necessarily need a firmware patch. The cartridges have a hidden reset if I recall correctly. Basically it tells the printer it's full after reset which is how people refill them or use up last few drops of ink. The printer doesnt check real ink levels just some sensor that detects how much that cartridge was used. I could be wrong though and this may not be on every ink cartridge brand.

>>15927 you can buy one used also put it on it's own vlan, and block wan access

>>15978 https://github.com/dfd-tud/deda also use a needle printer for really private stuff

>>16035 >Can it possibly recognize all world's currency, even unreleased bank notes? every modern bank note contains an astronomic constellation that the printers detect but you could add it to a document to make it uncopyable