Gerald,

You're mistaken about the optical size thing. The OpenType version of Minion Pro comes as a set of four different optical size variants, including one labeled "display." This covers all the characters in the font, not just caps as in the original "titling" release. Indeed, the main type page on Adobe.com right now has a prominent billboard for an article on optical size in Adobe type: see http://www.adobe.com/type/topics/opticalsize.html

That being said, in this case the "titling" companion would be used for 12-13 point text. The display and titling faces were designed for sizes >24 point. One could possibly use the subhead face, at a slight stretch.

However, regardless of what optical size of type is in use, that alone does not create visual contrast given that they are being used at reasonably appropriate sizes. So I'd also think about using a heavier weight, or perhaps italics, to create that contrast.

One could of course also change to another type family. Some sort of humanist sans or classic grotesque would go well with Minion.

I'm sad about the loss of MM technology myself. But....

> The sacrifice of MM technology (due to many things but not least pure old greed on the part of Adobe's owners)

What nonsense.

Adobe's "owners" are and were stockholders, who do not participate directly in corporate decision-making. The people primarily making this decision were at a low enough level that none of them had particularly large amounts of stock. The amounts of money involved in making this decision either way would not have been large enough to have a noticeable impact on Adobe's stock price (though many such decisions do add up).

I suppose you can put it down to "greed" if by that you mean that the people deciding to phase out MM technology came to the conclusion that the costs of keeping it were very large (true), and that the folks who buy fonts didn't care much (true for the overwhelming majority of font customers).

Although I personally wish I could do MM OpenType fonts, I know that I'm in a minority. Although many people miss MM technology sort of "in principle" it seems very few folks outside of Adobe ever released MM fonts.

T

I think I will retire from this discussion. Almost all of what you (Hrant) assert about why MM has died is simply untrue. You weren't there, and I am pretty sure none of your information comes from people who were; it's just your creative supposition about the reasons.

FWIW, I don't disagree very much about the broad issues of publicly held companies. They need to satisfy their shareholders by making a profit. Luckily, Adobe's primary means of making profits is to try to make software that our customers want. But there's no question that with publicly held companies the profit motive is often orthogonal and sometimes opposed to quality.

Later,

T

While I have purchased and used many of the MM fonts from Adobe since they were first introduced by Mr. Warnock at Seybold, I, in retrospect, realize that I almost never used them to create instances of my own and simply used their numeric equivalent of font variations for bold, light, etc., much like Adrian Frutiger used his numbering system in Univers (55, 75, 57, etc.).
Their was something relavent and fasinating about the concept of MM but I never seemed to get it to fulfill its potential so I just used MM fonts like other font families.
There was also something funny about the optical corrections (or rather lack-there-of) when going from weight to weight. I don't know if this was due to the nature of MM or just that the font designers were just too new to it to make it work better. An example is Myriad as opposed to Frutiger, Whatever was needed to corect the forms from weight to weight was done perfectly in Frutiger but there is something seemingly left undone in Myriad. This may be just my 60-year-old eyes (although I was much younger at the time).
I am sorry we have chased off Mr. Phinney from this thread since he may be the only one among us who can help clarify this. I for one promise to keep the discussion centered on typography and not on the failings or merits of Capitalism. I would hope that Thomas might either reconsider his departure from this topic or that he might send me his thoughts on the matter directly.

Chris L

"And then there's MS: some people say that when MS and Adobe started colluding in the new OT format, MS insisted that MM had to go. Their typical users are even denser than Adobe's, after all."

Hrant

I had heard this as well from an Adobe insider so I believe it to be true. The early propaganda on OT did include multiple axis capability as well as a number of other quite desirable typographic bells and whistles. Most of which are no longer on the prospectus.

I would think that Microsoft's main concern was to make it work and make it acceptable and get it the hell out there on the market. So, it happened, and up until about a year ago, when I stopped dropping in on the OT list, there was still argument about standardization of the tables. I never really liked TrueType, and I like it less with its new name.

Gerald

Yes, I am rather short-tempered with Hrant. His past history of gratuitous insults is pretty long. Given that, his most recent insinuations about what it's "convenient" for me to believe (in this and another thread) seem to be starting down that same path. I am not terribly interested in getting into a "pissing match" which is what it might start to come to if I actually dealt with his nonsense. What he chooses to believe, even in the face of contradictory statements from people who were actually there, is up to him.

However, there's no reason to punish the discussion for my aggravation with Hrant. Perhaps we can try to talk about the design and technology issues, and even the reasons for MM's demise, and keep things away from the personal attacks--and away from presentations of complete speculation as if they were facts.

Chris's comments about his usage of MMs are right on the money. We did a survey of end users, and we discovered that very few of the people who had MM fonts ever made custom instances with them. Multiple masters were a technology that were very little used by end users. We've already talked about the reasons, which were nothing to do with capitalism per se but everything to do with insufficient evangelism both internally and externally.

That lack of MM evangelism came from a simple lack of understanding that such evangelism would be necessary. Yes, the reality is the scenario that Hrant dismisses as "unlikely."

Microsoft's role in the demise of MMs was relatively minor, although I think some of us misunderstood it briefly at the time. Adobe supports MMs via ATM. Microsoft doesn't support MMs at the system level. For MM support to continue indefinitely without Microsoft support, Adobe would have to keep on making ATM forever and update it for every OS version. That's assuming it would even be possible as MS continues to increase the security of the OS.

Originally OpenType did have MM capability built in (for PostScript flavored OpenType only). But it was clear that it would be a very hard sell. Like Apple's AAT, why would vendors who had ignored MMs for almost a decade suddenly change their mind because of this repackaging? Adobe would have had to evangelize both technologies simultaneously. This concern about diluting our focus, given limited resources, was a major part of the decision to discontinue MMs.

I'm not sure exactly which of several things Gerald is referring to about arguments regarding standardization of OpenType tables. It's certainly true that any font technology that offers decent support for complex scripts will itself be a complex technology (at least in that area), and have to be very clearly documented. Sometimes things are not well written and the spec must be made more clear. Also, people debate the addition of new layout feature tags occasionally, but that's a natural thing with an extensible font format. In both areas OpenType has advantages over both Type 1 and old-fashioned TrueType.

Hrant is not really getting the current situation with hinting and font rendering very well. The truth is, as the rendering technologies get more sophisticated, the hinting is largely just getting in the way. That's why traditional superhinted TrueType fonts don't work well with ClearType. I had once thought that TrueType's advantage in hinting would vanish when screen resolutions got high enough, maybe around 2015 or 2020. Instead, the rendering technologies and LCD screens are eliminating this advantage much quicker than that. Maybe cell phones and other portable and home devices will give traditional TrueType superhinting a niche for a while. But I wouldn't want to bet on it.

T

> In the way Adobe as a company defined success (ie gobs of cash)....

Your saying this over and over again doesn't make it true.

As for rendering technologies, typical resolutions continue to increase.

Hrant, have you turned on "CoolType" (here meaning color device-dependent rendering) in Acrobat 6 and used it on an LCD screen? If not, you don't have any basis to comment on current CoolType rendering.

It's not at all obvious that superhinting can benefit sub-pixel rendering, at the resolutions we've got today. It seems likely that Microsoft would have done it if it was so worthwhile, no? Instead, they do radically *less* hinting for their ClearType technology. I wish I could go into detail about Adobe's research in sub-pixel rendering, but of course I can't.

As for screen resolutions, they have been steadily increasing for the last few years, and there are numerous technologies in development that threaten to increase screen resolutions a lot more. It is inevitable that they will go up; it's just a matter of how quickly.

T

Certainly one aspect that hurt MM is instances named MyriadMM_640%234.
I prefer Myriad Bold Condensed (or did I just make Myriad Light Extended

Certainly multiple masters as an end user technology are dying. Adobe's stopped selling 'em.

However, as a font development tool, I expect multiple master style interpolation to stick around for a long time. Certainly it is heavily used at Adobe, in almost all our new font development.

I see Hrant is still putting down CoolType without actually answering my challenge as to whether he's actually looked at its current incarnation: Use Acrobat 6, turn on "CoolType (color DDR) which is off by default, and view on an LCD screen.

Cheers,

T

How could you criticize something you've never seen? Well, probably the same way you could assert as "fact" information that was pure supposition, for which you had no personal or even direct second-hand knowledge of.

But that aside, you did not say whether you were referring to CoolType the DLL, which includes grey-scale only rendering, or CoolType the color device-dependent rendering technology. I wanted to specify that the latter shows the current limits of what Adobe has done with on-screen text rendering. So I wanted to make sure we were talking about the same thing.

Several things to keep in mind:

1) Everyone should remember that this is only intended to be viewed on an LCD screen. Anyone who is viewing it on a CRT will not see the intended effect; there's a reason it's off by default. (Hrant *still* hasn't said whether he's viewing this on an LCD screen.)

2) Don't assume your personal experience of subpixel rendering is typical. As the research shows (c.f. Kevin's presentation at ATypI), different individuals have different responses, and they very widely. I can't recall the exact percentages, but I think it was on the order of 5-10% of viewers who are ultra-sensitive to color fringing in subpixel rendering, but again sensitivity was widely distributed.

As for this particular case, there are definitely some problems with the larger size being rendered there, such as the bar of the t as you note. I too hate the s as well. However, without examining the font in a font editor, I can't say whether those problems are due to an error in CoolType, odd letterforms, or bad hinting.

You don't say whether these are TrueType or PostScript outlines, by the way.

But in any case, without picking apart the actual font to see what the hinting was like, I can't say whether better hinting could improve the rendering. However, I can say that in most cases when starting with well-made fonts, better hinting within the limits of current hinting technology won't help all that much in either ClearType or CoolType. Adobe could improve on the current auto-hinting tech for PostScript outlines, but not immensely (and yes, we have looked into it pretty carefully).

If what you mean to say is that *not* using a sub-pixel rendering technology, simply using hinting alone, will produce better results on an LCD screen, the available research pretty consistently contradicts that assertion.

If you want to critique subpixel rendering technology, it's also best to show the alternative. Care to show the old-style system rendering of those same ppem sizes of FF Din?

Finally, in regards to your parting shot, Hrant, you attack lots of people without acknowledging it, so I'm not special in this regard. The funny part is you calling yourself "objective." I'm sure that a poll on this or any other forum of people who know you reasonably well would generate numerous positive adjectives to describe you, but I'd bet a very nice dinner that "objective" would make few if anybody's lists of your personality traits.

T

This animosity is misplaced, and not mutual.

--

Well OF COURSE I was looked at an LCD screen! Are you insinuating that I'm dumb, or that I'm trying to trick people? Should I have a hissy-fit about that?

But as far as that screen-grab collage I've done that doesn't even matter:
1) Most of the faults I cited are unrelated to the type of screen being used - they're obvious anywhere.
2) The sample's viewers (and not me) have to use an LCD screen to decide.

The font I showed is a legit PostScript DIN: there's no reason to believe there's a problem with the font instead of concluding the much more obvious reality: that CoolType sucks. And if you persist, I'll have to put up grabs of the crash-and-burn Electra rendering...

And of course I'm not saying subpixel rendering is a bad idea, what I'm saying is that it doesn't eliminate the need for human input, at least not for the forseeable future. Again: you can't get the money to finance the writing of an algorithm that can match a human in such tight conditions.

Your thoughts on the limits on subpixel rendering are completely blind to the possibilities. I could make subpixel bitmaps by hand that would blow away this mediocre escapism. BTW, I'm not the only person who thinks this should have been called CrapType. Even some people who dislike me as a person and disagree with much of what I think have a serious problem with the backwards (or at best sideways) step that this constitutes.

Some people are so obsessed with automation that it becomes the objective instead of the means.

hhp

Who is baiting whom? And who is avoiding the necessary conclusion (that CoolType is not good enough)? This truth is right there for everybody to see. "I too hate the s as well"?! What kind of thing to say is that? Don't make it sound like this is about personal preference. That "s" simply SUCKS. Just admit it. And it's not DIN's fault. ClearType does a better job. And it's still not good enough.

Also, TT superhinting is not rendered superfluous at all - it's simply that nobody has used it to improve subpixel rendering yet. And it's not nice of you to keep spinning the PS-versus-TT war to your favor: the latter format allows more control, hence better results. TT is a superior format. Period. The war is over. Don't be like those Japanese soldiers hiding in trees three decades after the horse is dead.

But ignoring all the diversions and attacks, let me ask this:
Exactly how does CoolType use the hinting information (PS and/or TT) in a font?

hhp

Duplicate post.

This is all rather moot anyway, since ClearType deliberately ignores x-direction delta hints, which means you can't really do 'superhinting' for CT. Ironically, the reason for this is that existing superhinted fonts originally intended for b/w rendering, e.g. the MS core and web fonts, render better in ClearType if the x-direction deltas are ignored.

I'm still not convinced that there is any significant gain to be had in trying to superhint in the x-direction for ClearType. That is to say, I'm not sure that the desired results would be appreciably different from what you get without such hinting, presuming of course that you are starting with a typeface that is suitable to the medium. Some typefaces are unlikely ever to be very readable on screen, regardless of how good the renderer gets, how high monitor resolutions get, or how much work is put into superhinting them. Some typefaces are also crap on paper. What I like about ClearType, and what I would think would appeal to you Hrant, is that it throws readability back to type designers; screen readability used to be almost completely a technical issue, now it is a design issue.

No, FontLab does not support device exceptions in hinting.

Does this mean you think hinting is not type design?

No, that is not what I said. We were talking about text on screen, presumably at typical reading size, under a specific rendering technology. Regardless of whether one views hinting as design or not -- and I think the degree of design involved in hinting depends on the typeface and on the particular ppem size --, type on screen used, necessarily to rely on hinting as a largely separate process from the design of outlines. One could optimise the outline design process to make hinting easier, e.g. by wrapping outlines to pixels at specific sizes, but making readable type meant distorting the outline to achieve something that may or may not actually look like the typeface in question. My point is that ClearType, by increasing effective resolution while drastically improving contrast over greyscale rendering, reduces the need for distortion of outlines to present clear, readable forms on screen at a wide range of sizes. If less distortion is necessary, screen readability can be addressed in outline design in ways that were previously impossible.

Well, there's still quite a lot of minute fiddling you can do in ClearType if you want. Clarifying my comment above re. x-direction hints (thank you, Ross):

x-direction DELTAP's (those applied to points) are processed in CT (at least in the current version of the rasterizer). DELTAC's, which are those applied to cvt values, are ignored along with the other horizontal cvts.

I'm still not at all sure that the potential benefit is worth the effort, i.e. that what you would want to achieve by delta'ing points in CT as much as one used to for b/w is significantly different from what you get without so much extra work.

What's 'significant'? A statistically significant measurable impact on reading speed and comprehension.

Righto. Attached is a .pdf with a sample
of a spread. Scala Sans is the greyed type,
Minion Pro for the rest. I am looking for
something that would replace the words
"Godwin" and "Valete." Or confirmation
that what I have is appropriate.

cheers Stephen.

Minion.pdf (45.3 k)

Very nice to see that spread Mr. Kris Sowersby.

Oh! Thnx dude.
Just a curiosity,
have you tried in Swift,
I find it gorgeous.

Kris, I agree it looks fine. If you are doing large titles, above 36 pt, a possible alternative is Poetica's small caps, which are a bit lighter and more varied, but from the same hand and similar in design to Minion.

Yes

Jokes aside, I'm going to change this topic title to better represent
the discussion. I hope kris won't mind. Funny that he started this
thread with "Just a quickie."

Nice pages, kris. I don't think you need to change that type at all.
Minion SC is working just fine.

It's not the first time I make fun of my capillary deficiency...

Actually I'll have to watch out the joke doesn't get stale...