Hmm, I missed the 'Calibri Kerning' thread in the Design forum so it seems that DTL OTMaster Light and TTX might be better than the MS/Typograf tools for giving accurate kerning pairs numbers. Am I right?

Still not sure if they may help me verify the originality of the fonts...

1. If the font contains class kerning, you probably need an application like FontLab, if you want to see, how many kerning pairs the font includes. But the number of kerning pairs does not say anything about the quality of the spacing, because the pairs may be wrong or they correct wrong plain metrics or they are simply (almost) never used (kerning between small capitals and regular small letters for example).

2. For me the use of unlicensed fonts is okay, if you don’t earn money with them in any way, neither with selling them nor with using them in a project, for which you get money. If you license a font, you will get the original and there is no need to check a font for authenticity. It is a bit boldfaced to ask your question on Typophile. I mean, what you want to have, is support for piracy from the developers of fonts.

Arno, the second question wasn't anything to do with ‘use of unlicensed fonts’, it was about checking whether a font has been modified. Plenty of licenses permit modifying fonts so long as the derivative versions are not distributed.

There's no easy way to check whether a font has been modified without having the original against which to do a diff.
____

Re. the kerning question: there are two kinds of kerning that a font may have: kern table pair kerning, and GPOS table kerning (which may involve classes or, indeed, even funky things like contextual kerning, or kerning implemented as single glyph adjustments instead of pairs). When you examine a font using the MS Font Properties Extension, the 'Features' tab will indicate how many ‘standard kern pairs’ a font contains; this is the number of kern pairs in the kern table. If a font is not reported as having any standard kern pairs, this does not mean that it has no kerning: it may have GPOS kerning (check the OpenType Layout feature tree in the same tab).

Font editing or examining tools do different things when looking at fonts with GPOS kerning. Some will parse the GPOS kern feature and report the information back in terms of pairs, but there are various ways in which this can be misleading. For instance, a tool may or may not expand class kerning, and since GPOS kerning, like all OTL features, may be script and language specific what data is actually being looked at?

As noted in the Calibri kerning thread, tools also do different things when looking at kern table kerning. The MS font properties extension and DTL OTMaster will give you accurate pair counts (as will Ascender's FontChef tool). Other tools, including TTX, may not, depending on the size of the table.

Note, however, that the number of pairs in the kern table and even a list of those pairs won't necessarily reflect the actual kerning that you're going to get in different applications. Some applications can't handle very large numbers of kern pairs, and some have weird bugs based on things like 8-bit codepage contents (e.g. MS Word kerning only monotonic Greek diacritics, not polytonic).

Thanks for your answer Arno but I think you misunderstood my questions, maybe my poor english...

1. I know that the number of kerning pairs does not say anything about the quality of the spacing. It's a really subjective matter for what I know. I'm just looking for some nice tools because I'm a geek and I was surprised to see different numbers with the two applications mentionned. The problem seems indeed related to class kerning so thanks for that. i will read some papers about that and will investigate.

2. I suppose my question was poorly asked but let me say that I DO NOT support fonts/software piracy. On the contrary, my job as IT specialist is to make sure that we only work with licensed software (and that can be a pain for intranet applications...)!
In fact, I wanted to know if it was possible that the font may have been modified by the previous designer, maybe with a program like Fontlab or whatever, and if it was possible to check that in the font properties. I'm not used to that kind of technical details coming from a database and network engineer background. I guess I'll re-install everything from scratch on her computer and will see what happens.

Thanks anyway,
F

Thanks for your detailed answer John, it's exactly what I wanted to know ;-)

About the font originality, I'll try to find out next week and will re-install the original package. It's a pitty we can't find checksums in the technical details published by most foundries...

Have a nice week-end,
F

@ Fontology

I guess I'll re-install everything from scratch on her computer and will see what happens.

Everything? Do you mean the fonts? Then you have access to the originals? Somehow I don’t get it. Checking, whether anything was changed in a font, is much more complicated than reinstalling the original. If you want to ensure, that everything is licensed on a computer, (formatting the harddisk and) reinstalling everything (that is licensed) is much easier than checking everything. In case of fonts there is not much to install. Naturally, if you use a font manager with a data base, it might be more complicated, but I would prefer to organize the fonts by myself as designer. And the same is true for the other software. I would like to have a fresh computer with regard to the installed software. (I even would not like it, if anybody else than me installs anything on the computer.)

Arno: If you want to ensure that everything is licensed on a computer...

That's not what he's trying to do, Arno. He never mentioned licensing in his questions and his explanation make quite clear what he is actually trying to check: whether the kerning in a particular font has been modified or is as provided by the foundry.

Arno,

In my company, we manage about 12000 workstations so we have some automated process to install most of the software (not the fonts and some specialized apps like CAD/Publishing) but I won't get into details. I have more than 15 years of IT experience so I know a little about computers and whenever I get the chance to learn new technical stuff I'm happy. I never had this kind of problems before (we only have a few people working with CAD and Publishing software) so it was a good opportunity to ask the question here and get enlightened but of course the easier for me is to send one of our technician re-install everything. We keep track of all software installed on a user computer so everything will be fine next week, and yes, we have access to all originals but fonts CD's are stored on a remote site and are not accessible trough the network so I could'nt make a checksum to compare with the original... So if I was able to check rapidly the originality of the font, I would not have to send a technician get the CD to the remote site, re-install, and bring the CD back. Sometimes simple things may be really complicated in a big organization!

Cheers,
F

Fontology,

A first thing to check to see if a font has been modified or not, is to check its version number against an original. On Mac OSX, this is as simple as Get Info -- for Windows, you need Microsoft's own Font Properties extension.

If the version number is different, that is almost certainly the reason. If the version number is the same but the modification date/time and file size is different, the font might have been edited -- most foundries are careful to maintain accurate version numbering, precisely for this reason.

Charles — If your designer swears that the fitting is different than she’s used to and you’re able to confirm that the font files are “original” and unmodified, then check to see whether InDesign’s Optical Metrics feature has been applied.

This feature very subtly, but very definitely, mucks up the spacing of a font — ignoring and overriding the font’s metrics and kerning.

I just had a conversation with a colleague about a circumstance where the first proofs of a magazine redesign using a custom typeface design, which had been meticulously spaced and kerned, came back looking unbelievably atrocious, and my friend thought he was going crazy and began to question everything he thought he knew about spacing and H&Js.

It turns out someone had ignorantly applied Optical Metrics. Upon turning this off and reproofing, everyone agreed that the composition and design suddenly looked infinitely better and miraculously fixed.

I’ve ranted at length about this “feature” before on this forum, so I won’t repeat myself here. But just have your designer check for this.

@Theunis

That was my first idea but I'm not sure if it's really accurate. Also, in the case mentionned above, I didn't have the original to compare.

On windows, if you double-click on a font, calibri.ttf by example, it opens a simple preview of that font and it gives you different info like 'File Size', 'Typeface Name' and 'Version' (oh, and it also activates it by the way!). I have tested some random fonts and compared the results with the info given by the MS Font Properties Extension 2.30. Some fonts (not too sure but I think it was a Calibri font installed with the PowerPoint2007 preview software on Windows XP) had a different version number in the two apps, like Version 5.00 on the preview app and 1.0x on the other, aargh!

So I'm not sure if I would trust the version number. What if someone opens the font with an editor, modify some kerning pairs and saves it? I guess it does not change the version number unless you specify it but I have no experience with those tools.

@Kent

Excellent tip! I have only little experience with InDesign so i forgot about that feature (never tried it). I'll check that monday morning before asking to re-install everything.

@ Fontology

If the font can be purchased via myfonts, you can compare Indesign and the myfonts font preview with regard to the kerning of the pair. Myfonts (on the the old site is a check box, that activates the kerning) displays the kerning at least partly.

And you can post the name of the font style and the kerning value here. Indesign displays the value. There is probably someone here, that has access to the font.

The version number is indeed not evident.