Complete Times Roman or Times New Roman?

Although I normally steer clear of the bland, overused fonts, there are rare times when I have to use them. Now is one of those times.

I am digitizing the archives of a programming-language-standards group. Their meetings were run according to Robert's Rules of Order and their documents were meticulously formatted and tracked with version numbers, so this project has to rigorously respect the history.

That means I'm stuck with Times New Roman or a near clone for many of these documents. It doesn't mean I'm stuck with bad typography necessarily, though. I can make non-substantive typographic changes, like shifting acronyms into small caps, introducing old style figures when they're used in the text rather than headers or tables, and so on.

So here's my question, which I've explored before unsuccessfully:

What is the most typographically complete version of Times on the market?

I'm looking for text figures, italic text figures, small caps, and italic small caps at least (we deal with a lot of acronyms, some in book or document titles).

The restriction is that I can't use compatible but recognizably different typefaces. The glyphs have to recognizably be Times.

Any ideas?

clauses's picture

There is also Lido (but no small caps):

Walter Tracy's 'Times Europa' looks awful in the digitized version on the Linotype website, like crazy bad. I haven't studied the actual hot-metal version.

Then there are the 'Times Modern' (2006) headline font, and the 'Times Millennium' (1991) and 'Times Classic' (2001) body-copy fonts. Neither of them are for available for licensing I think.

As a sidenote, the free (gratis) 'Liberation Serif' is metrics compatible with 'Times New Roman'.

Other Times related fonts can be found in overview here: but there are more, e.g. special cuts for German with smaller upper-case letters.

clauses's picture

Oh, I forgot to mention the French version of Times New Roman. On the lead of Maximilien Vox a special french version of TNR was cut, 14 of it's characters was redesigned to be closer to the 'Romain du Roi'. Yes, I kid you not.

dan_reynolds's picture

OMG, I want that French TNR. That would be like a beach party on a page!

philippe_g's picture

Is it the same version as the one shown in John Dreyfus’ article “The evolution of Times New Roman” (Penrose Annual 66, 1973, page 173)?

Maxim Zhukov's picture

The ‘Monotype’ Book for Information (New enlarged edition; London & Salfords: The Monotype Corporation Ltd., 1959) mentions Series 827 on p. 125: ‘Series 827 has some characters redesigned to conform to a style more familiar to French readers.’ It also shows Series 727 with lighter (not smaller) capitals for use in German; 565, 565 and 567 (Greek); 327 and 334 (Cyrillic); 345 (alternate, narrower caps for Series 334 (Bold); 427 (Semi-Bold); 427 (Wide), and 627 (Book, with long descenders). It also lists 5 series of ‘Titlings’.

quadibloc's picture

That sounds like Times 827 is what is being discussed. I knew that the Times for 4-line mathematics had less slanted italics, and a re-designed lower-case v (to reduce confusion with the Greek letter nu) but this comes as a complete surprise to me, as well as Times 727 - presumably, the capitals were designed with thinner strokes because of the German practice of capitalizing all nouns.

philippe_g's picture

Here’s Times 727 from the same source, with normal Times to compare with:

kentlew's picture

Mike Parker did something similar with his approach to Starling. The Book weight has lighter caps to better match the weight of the original lowercase. Conversely, the Roman weight has heavier lowercase to match the weight of the original caps.

clauses's picture

Aha! I was under the impression that the 'german cut' was with smaller capitals, but if the 727 indeed is that 'german cut', and the only german cut too, then I stand corrected.

And it's the first time I've actually seen the 827. Looks hodge-podge. On the other hand I'm all for the lighter capitals in the 727.

dan_reynolds's picture

Dang! I was hoping that closer to the 'Romain du Roi' would have been even closer than that showing of 827 is…

Scott Thatcher's picture

Have you looked into the TeX Gyre project at, specifically their Termes typeface? My understanding is that their small caps, etc., are generated algorithmically, but tweaked with input from a type designer. I've never needed more advanced typographic features for those typefaces, so I haven't taken a close look. It might be worth checking whether they meet your needs.

I couldn't find a project update more recent that 2007--I'd love to know if they're still in progress or considered finished.


JanekZ's picture

last ver. oct. 2009

clauses's picture

I quick glance of Termes reveals that is does not have small caps, but rather petit caps. It has however what looks like support for Vietnamese and Greek, but the Greek lower case glyphs are of a very very bad quality.

toad42's picture

On Nick's recommendation I picked up Starling and am impressed. It is a surreal experience to be able to set text properly in what is essentially Times New Roman with text figures, small caps, sloped small caps, and ligatures. Being able to tame all the figures and acronyms - some of which are in book titles and so require italics - restores the spotty, shouty mess Times New Roman makes of our technical documents to some semblance of a calm woven texture. What a relief! Starling has solved my typographic problem.

Four observation on Starling from my casual study of it:

First, Starling has another advantage over Times New Roman, though I didn't make use of it in this project; Times New Roman only comes in two weights, but Starling offers six, which gives the typographer much greater control over the text's color on the page.

Second, on the flip side, Times New Roman has a much more extensive collection of glyphs, including extended Latin, Arabic, Hebrew, Greek, and Cyrillic. One does expect a ubiquitous face like TNR to support more languages than most. I didn't need those characters for this project, but others will.

I considered whether the two weights of Times New Roman might be used to fill in those gaps in Starling. TNR regular weight may or may not match Starling book weight closely enough to blend in, and I couldn't tell in my skim whether bold TNR is close enough to heavy Starling to match. Even if these weights are close enough, the lack of small caps in TNR means this approach to supplementing Starling would leave the extended Latin set incomplete, since there would be no extended Latin small caps to match TNR's extended Latin upper case. Further, the other four weights of Starling would then be left without equivalent characters.

While eyeballing the weight in Font Book I noticed that the axis of TNR's O and Q don't match Starling; TNR rationalized their axis, even though the lower-case letters (most visibly the o) retain a humanist axis that doesn't match. There may be other differences as well, but that's enough reason alone not to try to use TNR to extend Starling when additional characters are needed.

So, for projects requiring those additional characters, the ideal solution would be to add them to Starling, which would be a significant undertaking.

Third, neither Starling nor TNR has superscript and subscript figures, which is surprising considering TNR's widespread use for academic and scientific manuscripts. Calculating them as most word-processors do produces results as satisfactory as calculating small caps - that is, better than nothing but by no means good.

Although both faces include half and quarters as preformed fractions, TNR also includes thirds and eighths. Most word-processors can't calculate these at all, so their inclusion avoids a choice between using typographically awkward manual fractions or modifying the fonts to add them (assuming numerator and denominator figures are included to make that a trivial edit). Neither Starling nor TNR includes numerator figures 5 through 0 (1 through 4 are included in Starling, 1 through 3 in TNR), or any denominator figures, making repairing the fonts difficult.

In technical work, superscripts, subscripts, and fractions come up a lot, even though most fonts provide little support for them. Adding these forty-two characters to each of Starling's twenty-four fonts would give it another edge over TNR.

Fourth, a helpful but relatively easy change for the font designer to make is adding tabular (monospaced) figures (both lining and text). This makes the face better suited to financial use, since financial text tables become easier to compose. Neither Starling nor TNR includes them yet, despite TNR's widespread use in offices.

Overall, I was surprised by how much typographical support is missing from Times New Roman. Although I knew about its basic typographical deficiencies (the modernist exclusion of text figures and small caps), I still expected it to have a greater advantage over Starling in these other areas, given TNR's huge office-user base and its much longer history of active development. Instead, with a little work Starling's developers could easily extend it to exceed TNR's numerical support, though meeting or exceeding TNR's linguistic support will take more work.

Nevertheless, even without these features, just being able to set most Western European text properly makes Starling a huge improvement over Times New Roman.

quadibloc's picture

And, considering the subject of another thread, Starling even has a circle-P character matching its copyright symbol. There also is a character that looks like a picture of a sailing boat...

Incidentally, I noticed on the Font Bureau page describing Starling that William Burgess Starling designed it in 1904. As that was before Times Roman even existed, I was a bit puzzled; how could it be a modified version of Times Roman.

So I searched, and found this on the web:

which seems to go into quite a bit of detail on the story.

quadibloc's picture

There's one detail that is being overlooked, though.

Times Roman does strongly resemble Plantin. The Monotype revival of Plantin, however, dates from 1913, which is also after 1904. Even if Stanley Morison was a "character", there's no reason to suspect the origin of the Plantin revival.

This may not be an issue, since Burgess could have used the original Plantin typeface as a source of inspiration, of course.

Syndicate content Syndicate content