This deserves a 2nd look.
I played with stegdetect ages ago, and confirm that it false positived some photos that I took myself and knew to be clean of actual stego.
I'm curious if this has false positives, too.
How certain are we of that signature?
I'm not arguing about how the pattern is detected (i see you know unix too)
What I'm asking about is often this signature appears in the wild inside images that aren't actually marked with sneaky bits.