I remember vividly a discussion about if a scanner is an essential part of the “translator’s workstation”. Back then, the answer was – “sure, why not, but don’t expect to much”. The idea was: if I scan a hard copy, then I don’t have to spend time re-creating the layout and I can work by over typing the word document (or can do it with my CAT tool).
Problem was and still is the conversion into Word. The magic abbreviation is OCR. Looks like there are now lots of PDF2word conversion tools around – and they all have two major problems: recognising the source language and creating a word document that is somewhat flexible when it comes to editing. Why is that a problem? Well – the OCR is a problem because you don't want to save time just to end up to triple check that every single character has been recognised (or how else can you make sure that the “I” hasn't become a ”1” or the “!” a “I”?). And the layout of the automatically created Word document has to be flexible enough to do with changed word, line and paragraph lengths – no place here of paragraph marks at the end of each line (and even manual line breaks are a pain).
So how much additional time do you spend on cleaning and checking your converted file – and isn’t it sometimes just easier to do the good old translation as we used to in the olden days where source texts came as hard copy?