Fred Nader's Comments

MS probably avoided Linotype because they simply don't have the technical savvy for something of this caliber. They would have had to fragment and outsource it, and they'd still be lacking somebody who could tie it all seamlessly into the OS.

hhp

Having made core fonts for Microsoft and others, I disagree with Hrant's comments: there is nothing about Segoe that Linotype and their partners could not have developed.

I'll second that (having hinted a couple of MS core fonts for Agfa Monotype, years ago before I joined Adobe).

T

So is Segoe like the chap says, a
rip of Frutiger?

In order to know whether Segoe is a rip-off of Frutiger, one would need to know the mind of the person who made it. A more reasonable question to ask is: Does Segoe look too much like Frutiger? (regardless of intent).

I think it does, and I have had occasion to see a lot more Segoe than most other people outside of Microsoft and Agfa Monotype. I don't know whether Segoe is an intentional rip-off -- and to a large extent I think the question is irrelevant --: the similarity between the two faces is too great not to cause scandal, and MS should be concerned about this.

In order to know whether Segoe is a rip-off of Frutiger, one would need to know the mind of the person who made it.

Cough cough, what's to know?
�����������������
"Does Segoe look too much like Frutiger? "
You bet it does.

�����������������
Certainly�I agree with John.
there is nothing about Segoe that Linotype and their partners could not have developed.

�����������������

Some inside info:
- The font requires ClearType hinting, and I think that's proprietary to MS.
- Apparently Segoe's italic is more "divergent", shall we say.
- Agfa/Monotype, Linotype and MS are in contact to try to resolve the issue.

And what about Myriad? People who complain about Segoe should complain about Adobe's "clone" as well.

In terms of "sufficient originality", personally I haven't seen Segoe close enough to decide - certainly that lo-res image on Nader's page doesn't cut it.

hhp

ClearType hinting is not proprietary to Microsoft, and is easier than "traditional" superhinting. Microsoft and others have repeatedly done workshops that covered ClearType hinting.

What is somewhat tricky is doing a font that offers *switchable* hinting for ClearType or regular TT rendering. I don't know if Segoe does that.

T

> ClearType hinting is ... easier than "traditional" superhinting.

What about ClearType superhinting? The triple horizontal resolution surely makes that harder (hard enough that I don't think anybody actually does it... at the moment). And significantly here, it might introduce the color filtering/balancing between the three channels, which I'm pretty sure is proprietary.

hhp

Since I first saw Segoe, I've looked a lot at it, and at Frutiger and Myriad, and also at Stone Sans. The latter three constitute what I consider a sub-genre of humanist sans -- the Frutiger Line --, although Stone Sans also owes something to Meier's Syntax. It is a narrow sub-genre, clearly defined by Frutiger, but within it Myriad and Stone Sans manage to present novelty and originality: I don't consider Myriad a clone of Frutiger. Whatever the intent of the designer of Segoe, and whatever his design process/method, the simple fact is that there is neither novelty nor originality in the result: it just looks like Frutiger.

- Agfa/Monotype, Linotype and MS are in contact to try to resolve the issue.

And I think we should let them get on with it. They all know the issues, Microsoft most of all, and hardly need to hear our opinions.

Stone Sans? I think it's way different, and in fact pretty distinctive - otherwise Frothy would never have been unmasked.

hhp

Hrant, the whole point of ClearType is to reduce the need for extensive hinting in the x-direction. The ClearType renderer actually ignores x-direction delta hints. Ross and I have done extensive work with ClearType, and there is no gain in trying to control x-direction stems or proportions: the ClearType resolution gain allows greater fidelity to the glyph form, so optimal screen readability can be addressed in the design process rather than in the hinting. Indeed, if a font does contain x-direction stem hints -- e.g. for the benefit of 300 dpi laser printers -- it is often beneficial to turn them off for the CT renderer, which produces better results without them.

Note that these comments, while generally true of ClearType in its currently shipping incarnation, are even more pertinent to recent improvements that will ship with future MS software.

Stone Sans? I think it's way different, and in fact pretty distinctive...

Put Frutiger, Myriad and Stone Sans side by side and you will see relationships between them. Myriad is closer to Frutiger than Stone Sans, to be sure, but they do form a kind of sub-genre: there is a familial relationship. If Myriad is the direct descendant of Frutiger, Stone Sans is a kind of grandchild. In any case, my main point is that Myriad has both novelty and originality, whereas Segoe does not.

And significantly here, it might introduce the color filtering/balancing between the three channels, which I'm pretty sure is proprietary.

When MS first shipped a product with the CT renderer -- the first version of the MS Reader -- they went to great lengths to try to hint the Reader fonts -- ironically, including Frutiger Linotype -- so that stem edges would fall on the same subpixel boundaries, but this was because the first version of the renderer was not as good as it has since become. It is now possible to achieve excellent, consistent, high contrast bitmaps with greater fidelity to the outline width and weight. Again, this means that screen readability can be optimised in glyph design.

Could someone please post Nader's sample here?

Don't know why, I can't access his website ('this page cannot be found').

Eduardo, from Nader's site:

Thanks.
Well, I guess there's not much to say.
And BTW, I don't think Myriad is a Frutiger clone neither.
Between being 'not original' and being a 'clone' there's
some difference (see Nick's comment here).

I think this whole issue brings about new issues that need to be addressed, especially in the realm of copyright protection, which I know may be a point falling on deaf ears in the US. (I've seen some really crazy stuff in my day, but the treatment of type design in the US is one of the most astounding slaps in the face I've ever experienced--I know I'm preaching to the choir.)

I guess the biggest question I'd pose on this matter is: Does adding special treatment to an existing face (in this case, hinting, or specifically, "ClearTyping" [which is a phrase that I believe WILL be necessary in the near future, if it isn't already--so I'll take full credit for coining it when a definitive dictionary of type design emerges]) give it credibility as a new face? It obviously doesn't, or it shouldn't anyway. If that were the case then everyone would need to learn ClearTyping (which, frankly, I'm not prepared to do at this point); if you don't ClearType your face at the onset, "someone" else might do it in the future, which from what I can see in the comparison is what happened here.

I agree that to make a more accurate judgement I'd have to see some higher resolution; regardless, at this point I can see only a few minor differences... and those appear to be the result of hinting variation. Doesn't make it a different face, though.

The thing is, "technologies" such as ClearType affect the way the type is seen, and that's the most important thing. An outline font isn't necessarily the be-all and end-all of what the font really is. In fact sometimes it's merely an awkward vehicle: think of pixelfonts - especially grayscale ones, where the outlines are basically unusable as such.

hhp

In fact, come to think of it, even outline fonts aren't always what they seem to be. Consider the necessity for optical scaling: the same outline font looks very different at different sizes, and people make multiple size-versions of a font (or a MM optical axis) to counteract optical effects.

The true nature of a typeface isn't entirely expressed in the actual outlines.

hhp

John: your points about ClearType make sense to some extent, but I think not totally. You say ClearType makes superhinting less necessary. I agree. But it can't totally get rid of the need (not at current screen resolutions - not everybody is using your settings :-). You say that ClearType ignores x-hints. Sure. But I can think of two reasons for that, both temporary: hinting at triple the resolution with the color filtering trickery thrown in is much harder, and they didn't want to hold up the other MS products waiting for a first iteration of CT; and they wanted to leave room for improvement*, and could afford to since even a "low-tech" incarnation of subpixel rendering is still a big improvement.

* Leaving room for improvement has in fact become standard practice in the software industry, because you can charge for upgrades.

Basically what I'm getting at is that even with triple the resolution and anti-aliasing, there's some room left for improvement, and that would come from superhinting. But I admit it's possible that the gain is small enough not to warrant the effort.

hhp

I think there could be, in some instances, a very small gain in the contrast of some bowl shapes by hinting to specific subpixels; however, now that the colour filtering is doing such a good job there is a much karger benefit to be had in using subpixels for spacing and kerning, which so far has not been done. Once you start allowing the glyphs to fall on subpixel widths and to be kerned to subpixels, you lose the ability to predict on which subpixel boundary a stem will fall, but you gain a huge amount typographically.

> using subpixels for spacing and kerning , which so far has not been done.

But then you sacrifice the "one image per character" rule. Not to say I believe anything like that should be sacrosanct. But in the case of subpixel rendering, you'd be introducing different colors too - it might be too distracting to see two adjacent "el"s with different color edges.

And couldn't good spacing result from the font itself instead of the instructions - like you were saying about the forms themselves?

hhp

Well, Hrant, I can't argue with any of that, you're obviously right. Part of the difficulty in this discussion depends on which aspect (or, perhaps more accurately, which argument) we're going after. If we're talking about copyright protection then the points you brought up are major issues (not that they aren't otherwise). I would venture to say that type design is the most difficult copyright question to answer... it may be obvious to you and I, but it's clear that more than a few political types don't see it our way. Most of them can't see past the notion that we're not trying to claim the alphabet itself*, but our own interpretations of it; with that in mind, can you imagine the smoke that'd spout from their ears if they even heard the points we just discussed?

Of course other factors play here, too. What amount of work has to be performed on a (pre)existing object before that work itself is protected as a new object? Perhaps a lessen can be taken from the field of translation: Translations are only granted a "derivational copyright" because they are considered derivative works, and the translator must still gain permission from the original copyright owner to do anything to it. (This is all keeping in mind that, oftentimes, the translator will put in almost three times as much work as the original author.)

This and other issues need to be resolved before most policy-makers are ever going to take the matter seriously.


*I read an article a few years ago about a visual poet (was it Richard Kostelanetz?) who had given the copyright office so much trouble that he was actually assigned a personal copyright caseworker. This poet's work consisted of a lot of typographical work (the one I remember was three sizes of the lowercase a arranged on various parts of the page). The caseworker argued that the copyright office wasn't going to offer protection to a pre-existing system, and that any modification of the forms involved was "of a supernatural nature." Whatever that means.