
jim said:
You might do better by putting a colored transparent text overlay over a colored text base, such that one can spot the color change where the (additive please not subtractive) colors "don't mix."
For example a green base text layer with an additive red transparent text OCR overlay layer give you a yellow letter where the colors mix -- "hit", but green or red where there is a "miss".
big kudos to jim for his improvement upon the well-known standard suggestion of this tactic... it's very clever. the problem is that it's _too_ clever, by half... you guys are treating this task as if it is some visual-discrimination type of chore, which is the mistake someone would make if they had rarely -- or perhaps _never_ -- done the job... but for anyone who has actually _tracked_ the type of errors that are common in o.c.r. today, a visual approach to this task is the last resort. once again, the _most_ common type of error is spacey punctuation. you don't need some fancy "color overlay" to find spacey punctuation, folks! indeed, you don't even need to _look_ for it at all! you can merely instruct your text-editor to find it. when d.p. started, a decade ago, maybe then you needed to verify every line, word-for-word -- but today that's a ridiculous waste of time and energy. unless/until y'all get that through your thick skulls, you are chasing your tails around the wrong bush... and your efforts will never begin to scale as needed. -bowerbird p.s. the smartest digitizers, like nicholas hodson, have known for _years_ that text-to-speech is the _best_ strategy for finding well-disguised errors... your computers and ipads can _speak_! use that!