13 em(s) ?

sebsan's picture

Hi,

I am reading James Felici's The Complete Manual of Typography and at one point he talks about the missuse of Times. He goes into comparing the length of different lines of type made up of 26 letters —from A to Z— to show that Times is a very condensed type. What I don't understand is that he says that the average length of a line of type should be around 13 em(s). I though em(s) was (were) a relative mesure.

Norbert Florendo's picture

Yes, an em is relative to the point-body, so when a typeface is set at 10 points the em is 10 pts. wide.

Times is (was) essentially a newspaper face, therefore the reference of a suggested average of 13 ems to the line actually refers to newspaper columns and not book setting.

Consider when Times is set at 12 pts. the width of the line is 13 picas wide (12 points = 1 pica) or approximately 2 3/16 inches wide.

A more common reference used for optimal reading widths is character count and not ems.

sebsan's picture

Does that mean that the characters of Times are relatively narrow (condensed) compared to the width of it's em? —the width of 26 characters in Palatino seem to correspond better with 13 em than when set in Times— I hope this is not giberish to you.

Norbert Florendo's picture

Though I'm not exactly sure what you are trying to determine about how well 26 characters of a typeface fit to 13 ems, but this GIF will show Times New Roman setting narrower than Palatino.

William Berkson's picture

Sebastien, I think you have misread Felici. On p. 130 he says "the lower case of a *standard* text face is 13 ems long." He doesn't say it "should be", and doesn't say that narrower or wider faces are "wrong". He cites Times as an example of a narrower face.

He explains, "Once the alphabet length of a typeface diverges more than 5 percent from the 13-em standard, you will probably have to adjust your point size and measure." (Meaning, adjust from his rule of thumb that the maximum measure in picas is three times the point size.) He says that "27 characters is the minimum line length, 40 the optimum, and 70 the maximum."

Bringhust has a somewhat different view, explaining that 45 to 75 characters is satisfactory for a single column of a seriffed face in a text size, with 66 ideal. If you have discontinuous text with extra leading, you can go longer.

For multiple column Bringhurst recommends 40 to 50 characters. (Bringhurst has a convenient table showing the relationships between alphabet width and measure.)

In any case, Felici makes clear that there is an interaction between the width of the typeface, a good measure and the leading. These have to be considered together, and a good match found for the page size you are setting on. What is best will be influenced by the width of the typeface, the x-height, whether there are serifs or not, and other factors. So in practice you need to experiment with a particular typeface and page size until you find a good solution. And finding a good solution on size, measure, and leading is, I think, the most important aid to achieving a readable text, in addition to having a decently readable typeface.

sebsan's picture

Hi,

I am not that far in the book yet. I was referencing the page 69. The captions for the figure 5.2 clearly reads —The “standard” width for the lowercase alphabet of a text is 13 ems.— What I don't get, and I am probably slow, is why the ems are used as a fixed and universal value in this case when Felici writes on page 24 —The fundamental relative unit in typography is the em.— and just before that— Relative units have no fixed value in terms of absolute units such as picas, inches, or millimeters.— I'd like to understand before I continue with the book.

sebsan's picture

Ok, now I've got it. It took me a while. It finaly clicked into place when I was writing the previous post. I hope I didn't loose my right to asked more stupid questions.

Syndicate content Syndicate content