Number of Character Speech Bubbles

[Onyavar]

Were thought bubbles counted? And what about hypothetical/imagined scenarios like Roy and Tarquin in Elan's imagination?

Thought bubbles (e.g. Haleys or Nales internal dialogues) counted, yes. In the translation I did something to distinguish them from spoken language. And the imagined scenarios of course counted, too: It's a panel with a regular speech bubble, nothing extraordinary. Same applies for Belkars hippie dream quest thing with Shojo.

...do you have a job or go to school or something? Not that I'm admiring your accomplishment here, but...really?

No, I'm not making fun of them, but you still have go go through every single character and get all their speech bubbles. Even with the document in word, that's a lot of work.

Uh huh. In case you didn't notice, the german translation thread has been around since 2006. It's by far the oldest thread that is still current and has never been archived. I started participating there in 2009, around comic 450. That project was a free time thing and now, after six years (!!), I managed to put all the effort of 100 contributors into a single text file... Yes, it took a large bit of my free time over the years. But it was fun.

If only there were some sort of electrical computing device that could do the counting in some sort of automated fashion...

OK, I'm curious as to how the Search function in Microsoft Word could be used to count the speech bubbles in a comic strip. I just spent like 10 minutes looking over the help file for it and I just don't see it. I can definitely see it working if you had a transcript, but with an image file?

Since I have the transcript and this in a nicely formatted fashion, it IS sort of automated. I simply input the search term "Roy: " or "Nale: ".
Now how to get a transcript automatically...

Spoiler

Show

If you have enough time and devotion, I can tell you how to do it. First you would need to download all comics. Next you would program a perfect science fiction-like OCR that can recognize white bubbles and put the text of each bubble into a single line. Afterwards, you need to do the same with black speech bubbles (IFCC/Xykon/Qarr) and blue (Roy/Celia) and so on. Then you would need to go over those results and sort out which character said which bubble. Next, you need to let the OCR recognize all text which wasn't put into a speech bubble and then sort out which one is text said by a character (noises and inscripts don't count, remember).
Having done all of that your next step is to program OCR-bots to fool every captcha in the net and become bilionaire.

Or you could simply pay around 50 students to do the job of typing down a trancsript, 17 comics for each and you're done.
Wait. NOT paying 50 devoted OotS fans to type it will also work.

Well, I have no idea how this would work out in MS Word. I'm using notepad++ to handle the text file, and said file comprises of 44470 Lines (as of #850) and apparently 2 Million charakters/letters. The file size is almost 2 MB - I can imagine it would be at least 10 times bigger if you start with Word.