Reading Papers with Unclear Typesetting

Question

From time to time, I have to read a paper with unclear typesetting. By this, I mean:

• very small fonts; sometimes 8pt or smaller.
• very bad scan;
• Using bitmap fonts instead of vector ones.
• too old formatting (usually dating back to 1970s or the beginning of 1980s.)

In these cases, one might prefer to read a more clear version submitted to the authors' home page. Unfortunately, the author version is not always available; and even in that case, they might be very lengthy (the so-called full version of the paper).

In the case of bitmap fonts, I found a good strategy: Find the postscript (PS) version of the paper, and convert the bitmap fonts to their vector counterparts using pkfix or a combination of pkfix-helper + pkfix, whichever applies (read the documentation).

However, I still can't find a workaround for other problems. Specially, small-font papers bother me a lot. My best bet is to use a desk lamp: I don't know why, but magically the words seem larger under its light!

How do you read a paper whose typesetting is unclear? Specially, what do you do when the font is too small to read?

PS: Some months ago, I found an application which "splited" a PDF into several segments, so that they were easy to read on an iPhone or the like. (Sorry, I don't recall the name of that app.) However, the output was downgraded to fit the iPhone small screen, and therefore it wasn't apt for computer screen or printer.

Answer 1

If you work on a university campus, there's likely a large building nearby full of study carrels and magazine racks called a "LIE-BURY". Deep in the bowels of such buildings, one can often find images of old research papers inscribed onto thin sheets of cellulose fiber extracted from wood and cotton, called "paper", and then bound into larger brick-sized objects called "books". (At least according to Wikipedia, this is the reason we call our research reports "papers". Who knew?!)

Surprisingly, despite the crudeness of the technology that produced them, these so-called "books" are often significantly easier to read than their more modern electronic counterparts.

One serious disadvantage of this approach is that one must actually walk all the way across campus, but we all must occasionally suffer for our art.

Answer 2

Not a perfect answer to my question, but you can give Scan Tailor a shot. It does magic with poorly scanned documents.

I also use Adobe Acrobat's ClearScan OCR, after processing the document with Scan Tailor. It vectorizes the text, and therefore enhances the readability.

Here's an example of converting a "photographed" page to one that looks like a "scanned" one. It is provided to give an impression of the quality enhancement provided by Scan Tailor.