Best way to make this manuscript digital again?
December 28, 2018 2:46 PM   Subscribe

I completed NaNoWriMo...... in 2009. However, since then, I appear to have lost the Scrivener file in the shuffle of computers etc. I have a printed manuscript, maybe a couple of hundred pages, with some annotations towards the beginning. How do I turn it back into something I can edit on a computer again?!

I started to retype last year, but it was both slow going and I got tripped up wanting to edit, being mortified by old writing, etc. I think I'll be able to actually focus on the editing if I don't have to retype the whole thing.

What are my options? Get it scanned and run OCR? Hire someone to retype for me? (how the heck does that work?) Or should I just grit my teeth and commit to a few hours a week of typing and telling my critic brain to stfu and let me type?

(Looking back on it, I think it's worth working on, and this looks to be a good year to actually do that. I still like the story and characters, and I've been doing a lot of other writing in the last year and a half. It's time.)
posted by epersonae to Technology (11 answers total) 5 users marked this as a favorite
 
Search for "OCR scanner free" and you'll find many options depending on what kind of phone or scanner you have. You'll be able to save to doc or txt files form there.
posted by ananci at 2:54 PM on December 28, 2018


I vote scan and OCR. It won’t be perfect, but you’re going to rewrite anyway; quicker and easier than retyping it all yourself, cheaper than getting a professional typist. I honestly think retyping it all will be more work than rewriting from scratch with an occasional glance at the original (which actually could be an option if you’ve built up your writing stamina).
posted by Segundus at 2:55 PM on December 28, 2018 [1 favorite]


You could use a transcription program and read the manuscript aloud.
posted by Ursula Hitler at 3:22 PM on December 28, 2018 [1 favorite]


Tesseract is a free open source OCR program I've used before. They recently (October) released a neural net-based OCR implementation, which I haven't tested, but is probably pretty close to the state of the art.
posted by axiom at 3:26 PM on December 28, 2018 [2 favorites]


I like Scannable, a free app, for easy-to-OCR documents. (Anything typed and printed is easy. Hard-to-OCR includes handwritten, heavy with art, and complex forms.) Once you've got images, there are plenty of sites online where you can upload them to get OCR text.

If you run into problems, MeMail me; I have Acrobat Pro and a lot of practice converting scanned pages to editable docs quickly.
posted by ErisLordFreedom at 3:31 PM on December 28, 2018 [1 favorite]


Best answer: Check with your local libraries to see if they already have scanning equipment and OCR software (e.g., it's built into the professional version of Adobe Acrobat). You may even be able to find a scanner or multifunction printer/scanner that has a document feeder so you can scan everything at once (assuming it's printed on flat, regularly-sized paper).
posted by ElKevbo at 5:54 PM on December 28, 2018 [1 favorite]


Best answer: Many modern photocopiers include the option to scan to PDF and save to a USB thumb drive or email the file. That would automate scanning the document, then running the PDF through an OCR program would also be straightforward. Some PDF software also includes the option to save to MS Word doc or similar.

That's my suggested path to get from printed document to editable digital file.

As noted by ElKevbo, your public library might have all these elements already. Otherwise, a copy shop will have it, or your work might, too. If not your work, ask friends and family in your area.

Two final notes: first, a hundred pages might mean something too big to email as a PDF, so saving to USB is probably most reliable. Second, OCR can introduce some odd errors, and OCR'd text to a Word file can add strange formatting, so it's probably good that you're looking to edit this, because the initial output might be usable, but messy.

I've done a bit of this before, both by hand and with more automation, and I'm happy to share more insights or ideas. Good luck!
posted by filthy light thief at 8:16 PM on December 28, 2018


I have had good success with the OCR built into Google Docs. Some tips here.
posted by snoboy at 10:19 PM on December 28, 2018


I can’t imagine working with 100 pages of OCRed docs - the formatting alone will be a huge pain in the butt. You would need to spend hours fixing bad spacing and line breaks.

I’d post this on Fiverr or Mechanical Turk at something like $5/10 pages.
posted by samthemander at 10:42 AM on December 29, 2018


Response by poster: Since I think I have access both to a scanner at the library and actually Acrobat Pro as well, I'm going to give that a shot probably later this week. See how it goes and if I absolutely hate the results I'll look into the other options. (Will update either way. I marked ElKevbo's answer as "best" because I honestly hadn't even considered the library!)
posted by epersonae at 7:39 PM on December 30, 2018


Response by poster: Turns out, the best way to make the manuscript digital again was to find a file in Dropbox called "nano2009-final.doc".

But before that, I was able to scan the printout at the library and had started to make headway in Acrobat Pro, so thank you to everyone who offered suggestions. I think that would have ended up being a good solution if past-me hadn't stashed a backup somewhere I wasn't expecting and hadn't looked.
posted by epersonae at 4:12 PM on January 11, 2019


« Older Exercising better down where it's wetter   |   Gaming Laptop suggestions needed Newer »
This thread is closed to new comments.