Someone could probably create a printer driver that does that automatically. It probably would need some calibration to work, though.
There's an additional challenge that some more recent printers often use a different watermarking technique whose details have never been documented.
it doesn't work well at all, but every now and then it produces some pretty amusing results.
Early versions of photoshop would put a big black box over me in some of the photos.
fun for the whole family.
There's a Gibson novel, I think, where the protagonist at some point obtains t-shirts with a pattern which is recognized by the universal DRM-enforcement code in video camera chipsets, making them invisible to surveillance.
But how would they add the watermark? GIMP I guess...
Digital cameras can be pretty sensitive to IR, but there is typically an IR filter over the sensor to prevent detection.
Your best bet is to wear one of these:
Given that it's -10 C outside today... nobody will even think you're weird.
And here's a thread about it: https://news.ycombinator.com/item?id=7006848
Here's a thread about a BBC article: https://news.ycombinator.com/item?id=9776985
Also, slightly unrelated, but this for some reason reminded me of a certain issue some copiers had a while back, where they would swap certain numbers/letters for some reason when copying. Don't really remember what the cause of that was, though.
With simple technology, at first, you might get grainy or blurry images, or poor frequency response in audio. Those are easy to perceive as artifacts of the medium. Then you get things like JPEG fringes and context-dependent color shifts, which are a bit harder to learn to see past, since they are interacting with the image in sophisticated ways. JBIG (from the article) detects and compresses repeated textural elements or multiply-used glyphs, but if it applies that behavior to something that is actually information-bearing, it'll produce convincingly misleading artifacts.
There's a sequence in _A Fire upon the Deep_ where two people are conversing over a link with rapidly deteriorating bandwidth— their communicators are compressing more and more heavily to compensate, until at the end the character is "interacting" with a sophisticated Markov-like model of their interlocutor with only a few bits per second of actual entropy from the far end. Or possibly none at all.
I suspect this patent might have something to do with it, but the math is beyond my comprehension: http://www.google.com/patents/US7720249
This patent was referenced by the above patent and seems even more applicable:
http://www.google.com/patents/US6449377 (references patents 5583614, 4723149, 5633952, 5640467, and 5424807 which may be good further reading).
I also remember someone mentioning that they reverse engineered the currency detector code in Photoshop(?) and got something like "fourier mellin transform, followed by checking specific values in an array".