What Font to Use for Body Text? Serif or Sans Serif?

About 15 years ago, in fact probably even 10 years ago, there was no doubt about which font type to use for body text. The answer would always have been a serif font. This was because whatever the documentation was that was being created it would have been consumed, which means ‘read’, as hardcopy (i.e., printed on paper).

There was no doubt in any copy editor’s mind about what font type to use for textual content that was going to be printed and read.

The answer every time would be a serif font. They would not even have to think about it.

San serif fonts would only ever be used for headings, call outs, footings, end-notes, and narrow ‘highlight’ columns of text.

But then computer screens came along. Computer screens have two huge problems when it comes to reading text on them. The first is that they have a much lower resolution than printed text. Printed text in a glossy magazine is typically set at 1,200 dpi. Text in a quality printed novel is typically set at 600 dpi (as is almost all inkjet or laser printed output that you might encounter at your work). Newspapers are effectively printed at the equivalent of 300 dpi (some may go as low as 220 dpi).

But computer screens are around 70 dpi which is about a quarter of the resolution of the print in a newspaper, an eighth of the resolution of the print in a novel, and a seventeenth of the resolution of print in a high quality magazine. So compared to printed text the text on a computer screen is downright crappy.

The second problem with computer screens, and more so with LCD screens than with the older CRT screens, is that they have a problem creating good contrast. Their blacks are really a very dark grey, and their whites are generally not that white. Also many users do not have their screens correctly adjusted for the working environment, which makes the contrast (difference between black and white) even worse.

There is a third problem with computer screens that I will just quickly note but I won’t go into the details—computer screens are backlit whereas reading from paper you are using front-lighting.

So, due to a combination of these issues, with computer screens came this problem. Sans serif fonts ‘looked’ more solid and crisp on a lower-contrast lower-resolution back-lit computer screen. This was simply because they don’t have all the little curly wispy serifs or hints that every letter in a serif font has.

Because of the low resolution of the computer screen the serif hints on serif fonts make the font look un-crisp. Serif fonts do not look as solid on a computer screen. This was especially so before Microsoft implemented Clear Type technology in Windows 7 and Office 2012. But even with Clear Type enabled serif fonts still don’t appear as solid as sans serif fonts when viewed on a computer screen.

The above effect is shown in the following image. The top line is a sans serif font (Arial) and the lower line is a serif font (Schoolbook). To show the problem I have enlarged the capture by 2x and Clear Type is enabled (for anyone wondering).

Font01

See how the lower serif font line looks less solid than the sans serif line above it. But this only occurs on the low resolution computer screen due to pixel averaging and anti-aliasing that has to be done to present the fonts on the screen, and because the serif font has a lot more going on (so to speak) due to the serifs a lot more pixel averaging and anti-aliasing needs to be applied to this font. For example look how complex the serif “E” is compared to the san serif “E” which is basically just four straight lines.

However, when you print on a 600 dpi laser printer, which is how almost all documents in business are printed, then all this requirement for pixel averaging and anti-aliasing goes away and things look quite a bit different.

Font02Scan2

[The above image comes from the text from the first example, which was done in Corel Draw, being printed on a Samsung 2010 600 dpi laser printer and then scanned using a 1200 dpi Epson Perfection 1650. Photoshop Elements was then used to set the white balance and size the scan to fit my posting width. No enhancement or sharpening or other touch-up editing was done.]

Now—obviously—both lines look a lot more crisp as they have been printed at 600 dpi on a laser printer and then scanned. But the point here is that now there is much less difference in solidness and crispness between the two. In fact, due to the way serif fonts are set out and the use of the hints, to most people—after looking at this for 10 seconds or so—the bottom serif font is going to appear darker and more solid. This is the very reason serif fonts were designed in the first place.

When you factor in back-lit vs. front-lit, which I cannot demonstrate here, the difference gets even bigger.

Back when the printing press was invented (in the mid-1400s I think it was) it would have been a gazillion times easier for them to make all the itty bitty metal letters as sans serif without all those wisps and hints on every letter. Much less work. But they went with serif fonts for a reason. They are actually easier to read—well when it is printed text on paper at relatively high resolution they are anyway.

So this brings us back to the starting point problem. Whether to use a serif font or a sans serif font.

The answer is both obvious and confusing. Basically if the content you are creating is going to be consumed as hardcopy (i.e., printed on paper) then a serif font is the way to go, but, due to the very low resolution of computer screens (compared to print on paper), the lower contrast, and the effect of back-lighting, if your content is going to be consumed directly off a computer screen then a sans serif font is most likely a better choice.

Many people creating content that is going to be consumed as printed output on paper make the mistake of using a sans serif font because they spend 95 percent of their time seeing the content on a screen as they create it; and they prefer seeing in as sans serif. They overlook the key point that, once the work is finished, the intended audience will be consuming the copy as a printed work—which, as I have pointed out above—is a completely different scenario.

But then there is the confusing bit I mentioned earlier—much content that is created with the intent that it be consumed as printed hardcopy may end up being consumed on a screen. For example a manual produced as a PDF intended to be printed and read may often be referred to by the intended audience using a PDF reader directly from the screen.

Then what do you do?

Well I don’t really have a useful answer to this but at least now you might have a bit more to think about when choosing which font to use.

BarryMark