Discussion:
Optical vs. Metric spacing
(too old to reply)
KPanthen
2007-03-30 14:18:34 UTC
Permalink
We are running Win XP, IDCS2 and have been using the default metric spacing in the control panel. I noticed that the spacing changes in Optical but also noticed that the text wrap is much more spacy.

What is everyones take on using Optical vs. Metric.

KEN PANTHEN, Albany, NY
M***@adobeforums.com
2007-03-30 14:48:00 UTC
Permalink
I tend to use Optical because Metrics are at the whim of the font's makers. We have some LT ITC Garamond in which the foundry in its widsom decided that a space after a period needs to be kerned out of existance. Gives a whole new meaning to run on sentences. <groan>

The problem I have with using Optical is that using the keyboard shortcut to Reset Kerning and Tracking doesn't work with Optically tracked text. Instead it changes the text's tracking to Metric, so rather than reset the tracking, it negates any relationship with the tracking you had. Quite the opposite that anyone would want but I'm not holding my breath for a fix. For all I know, Adobe considers it a feature. Nonetheless, I put in a Bug Report about it.
D***@adobeforums.com
2007-03-30 14:53:59 UTC
Permalink
I too prefer optical. It also works between characters of different sizes and even of different fonts. It also provides some degree of auto-tracking, tightening pairs as the font size increases.

It's interesting that the vast majority of the visitors to the Typography forum strongly prefer metrics. Perhaps they have more empathy for the blood, sweat and tears that went into making the metrics kern pairs and use them out of respect for the artist.

Dave
t***@adobeforums.com
2007-03-31 05:10:33 UTC
Permalink
We use metrical. In most cases the optical kerning isn't that much better. It may reduce the problems of some pairs, but other pairs go worse. So both are imperfect for me.
It is hard enough to bring InDesign to not change the wrap, when reopening documents again and again, so selecting optical kerning forces InDesign to use a further algorithm that has to be calculated again and again - that's a bit precarious for me.
We use optical mostly for page numbers or numbers anyway.

I strongly recommend Adobe to build in a document based kerning feature for editing the metrical pairs as Quark has (since version 3 or so). And in this tool I would love to see the ability to kern glyphs against the left or right box margin, so that I can determine how the optical margin alignment will take place, if the feature is used. (BTW: optical margin alignment should at least be a feature of a paragraph and not of a story...)

Tobias
P***@adobeforums.com
2007-03-31 11:22:30 UTC
Permalink
So there have to be some kind of check box to use metrical, but on those
changes use optical or something like this.




But this is already possible, Just select the particular glyphs you want optically kerned and select optical from the kerning field. The rest will maintain the metric kerning.

If you want some sort of automation for this, create a character style whose only attribute is optical kerning. You can then assign a keyboard shortcut, and use it in a nested style, find/change, and scripts.

Peter
P***@adobeforums.com
2007-03-31 11:44:51 UTC
Permalink
After thinking a little more, it might make sense to add optical kerning to another style in some cases, rather than keeping it as an isolated attribute, since you can't apply more than one style to a character.

Peter
t***@adobeforums.com
2007-04-01 04:58:24 UTC
Permalink
Peter,

yes, I can do I right now, but it's not a feature. And I'd like to have such a feature (or how about a plug-in?).
And about your second thought: This is another feature I'd like to see: the possibility to use multiple character styles on the same character(s).
(I know: I can already do this via nested styles. But that's all a bit of a workaround.)

Tobias
D***@adobeforums.com
2007-04-01 23:40:48 UTC
Permalink
I also use metrics in preference for optical, though it's not due to empathy for the blood, sweat and tears that went into making the metrics or out of respect for the artist. It's just that, given the choice between kerning pairs that have been generated by a kerning algorithm in ID or those generated by a kerning algorithm in a font editor (or, separately, as in Kernus) and/or tweaked by the font designer who drew the glyphs, I choose the latter. I might consider using optical kerning where I'm mixing fonts, but I prefer to manually kern in these situations. And, if I dsagree with a kerning pair in a font, I prefer to manually kern. As for tracking, I prefer to set my tracking myself.

I'm with Tobias in wishing ID had a kerning editor and a way to adjust the optical margins (and even suppress them for a character). I would like to see optical margins be a paragraph and a character level attribute.
t***@adobeforums.com
2007-04-02 04:59:51 UTC
Permalink
Dominic,

how do you kern »bad« pairs of a font right now? I'm curious ...

Tobias
D***@adobeforums.com
2007-04-02 05:20:28 UTC
Permalink
Manually, then search and replace for all such instances.
t***@adobeforums.com
2007-04-02 13:15:08 UTC
Permalink
There were times when one could use fontographer or fontlab to do the kerning as one liked. But nowadays all EULA's prohibit such an action.

Since version 2.02 I made regular feature request for this. CS3 doesn't have it either, so I'll go and post ...

sigh
Tobias
KPanthen
2007-04-02 13:22:49 UTC
Permalink
Thanks to all. It's a big consideration to switch over because the text wraps differently, something our designers will cring over.

Thanks to all

KEN PANTHEN, Albany, NY
D***@adobeforums.com
2007-04-02 22:00:52 UTC
Permalink
Post by t***@adobeforums.com
But nowadays all EULA's prohibit such an action.
Well, Adobe's EULA may be silent on the specific case of editing glyphs or metrics (I'm sure it didn't use to be so), but the font FAQ makes it clear that you are allowed to make such changes to Adobe fonts. See also the recent thread in the typography forum where Thomas Phinney makes it clear that the FAQ does apply:

<http://www.adobeforums.com/cgi-bin/webx?128@@.3bc38410>

I haven't checked other foundries' EULAs recently, so I don't know if they all prohibit such editing, but I never was very keen on going that route anyway, for several reasons:

* Some font editors are lossy.

* Some apps didn't seem to like such edited fonts (I had issues with Ventura).

* You have to keep the specific font files with the publication for evermore, and any future users risk text reflow if they open the file without that font. And, in my experience, it's very easy for fonts to get separated from the files, no matter how well-intentioned the original user may be.

I far prefer to keep changes to the kerning within the app file itsefl, either through a kerning editor or through manual search and replace operations. You're not the only one who's repeatedly requested a kerning editor - I was amazed that ID version 1 didn't have it considering how typographically superior it was advertised as being. Adobe are aware of the issue and say they are working on it but that it affects more than just ID.
töff
2007-04-02 22:22:18 UTC
Permalink
Love OPTICAL except for the connected fonts (script faces), where you
kinda have to use metrics, and even then things are often a nightmare.
t***@adobeforums.com
2007-04-03 04:34:54 UTC
Permalink
FontBureau's, Linotype's, FontShop's EULA's prohibit explicitly. And if I buy my Adobe font via FontShop or Linotype, I'm trapped.

Your list about »not to modify fonts with font editing programs« are true. It's definitely not the best way to handle this issue, but a workaround. If I had to manually kern 1500 pages of book layout with find/change, it breaks the financial calculation ...

It definitely has to be an InDesign function. I can modify spacing, size, transform a font manyfold, but cannot change the kerning pairs? Ridiculous.

Tobias
D***@adobeforums.com
2007-04-03 05:28:43 UTC
Permalink
Post by t***@adobeforums.com
And if I buy my Adobe font via FontShop or Linotype, I'm trapped.
Why are you trapped? What does the vendor have to do with Adobe's EULA?
Post by t***@adobeforums.com
If I had to manually kern 1500 pages of book layout with find/change, it breaks the financial calculation.
I've had to do it for about 1200 pages and it's no big deal, though I would prefer a kerning editor. I usually don't have to adjust a huge number of kerns (usually only one or two) and I can search and replace across all files at once. Takes only a few seconds. If I had to adjust a lot of kerns, I'd probably think twice about using that font in the first place.
Post by t***@adobeforums.com
I can modify spacing, size, transform a font manyfold, but cannot change the kerning pairs?
Well, to be fair, you can change kerning, but just like the spacing, sizing, and transfomations, those changes are done on a case-by-case basis, not automatically applied to every such instance.
M***@adobeforums.com
2007-04-03 15:04:28 UTC
Permalink
Don't quite get the "to be fair" rationale. Spacing et al can be done automatically by paragraph style. Kerning at best is a Character style.

My problem with using local formatting to mitigate unwanted metrics is that I have a long-standing habit of adjusting the tracking slightly one way or the other as a quik way to check for nicer line breaks and rags. If the copy gets edited, I like to select the entire paragraph and hit Reset Kerning and Tacking so that my adjustments don't multiply. (Hence my beef about the keyboard shortcut changing Optical text back to Metric.) Under this scenerio, I lose local kerning adjustments.

Using Character styles and/or S&R to affect kerning seems a bit tedious as well, unless one limits themselves to just a couple of major problems. You would have to maintain a list of character pairs that need to be altered for each font. (Hey, that's the job of the kerning editor that ID needs.) As I said, I have one font (I'll refrain from mentioning that it's Linotype's ITC Garamond) in which the Metrics are just horrendous. All punctuation+space is kerned -120, and some caps are kerned into a preceding space. A new sentence starting with a cap T, for instance, will almost tuck the preceding period under the left side of the T. This is not something that you want to have reset or miss in a S&R.

Bottom line: Without a kerning editor any font that has less than desirable metrics is better off set optically.
D***@adobeforums.com
2007-04-03 23:51:16 UTC
Permalink
Post by M***@adobeforums.com
Don't quite get the "to be fair" rationale.
I was just trying to make the point that none of the other things Tobias mentioned are stored in a font specific file, so I didn't see the fact that you could make those changes to text in ID to be any way incompatible with the lack of a kerning editor.
Post by M***@adobeforums.com
Without a kerning editor any font that has less than desirable metrics is better off set optically.
So in those cases do you set the whole text using optical kerning? If so, how do you know that ID's kerning algorithm will produce better results than the kerns in the font? If I had to use a font with numerous kern pairs that I disagreed with, I'd probably write a script to do all the searching and replacing, but I don't find it overly onerous. It's just one of the things I accept that I have to do to get the text the way I want it. As for losing local formatting, I avoid using it and generally stick to creating styles.
Post by M***@adobeforums.com
A new sentence starting with a cap T, for instance, will almost tuck the preceding period under the left side of the T.
It sounds to me like they've tried to make the spaces visually balanced in ITC Garamond. I haven't seen it in much published work, but some people hold that word spaces should be decreased considerably before certain letters, like T, A, W, and V (eg, see Dowding's Finer Points in the Spacing and Arrangement of Type). It's not a look that I'm fond of.
t***@adobeforums.com
2007-04-04 05:01:05 UTC
Permalink
Dominic,

in the beginning we could do find/change (manually or per script), but we often have customers who disagree with some pairs after correction. Then we have to go through the text step by step, because the line wrap could change and orphans may appear.

Your »to be fair«: spacing is inside the font. But that's a bit bean counting.

Tobias
D***@adobeforums.com
2007-04-04 05:16:50 UTC
Permalink
Agreed, S&R is not an ideal option. But surely a kerning editor would pose the same problems for you - if a client disagreed with a kerning pair (and most people I deal with wouldn't even know what a kerning pair is), then amending the pair in an editor or within the font itself would still require you to go through the text looking for problems.
Post by t***@adobeforums.com
spacing is inside the font.
Yes, but unless you're saying there's a way in ID to adjust the spacing values in a font so that those values are thereafter changed for all ocurrences of that font in that file, I don't see the relevance. Because all I meant by my response to your comment that you "can modify spacing, size, transform a font manyfold, but cannot change the kerning pairs" was that you can change kerns manually or by a (character) style and that in this regard it's the same as transforming a font or modifying its spacing and size. You do them all manually or by a style in ID but none of them can be done automatically. Now, if ID had a tracking editor but not a kerning editor, I could understand a statement like "I can adjust tracking but not kerning pairs? Ridiculous."
t***@adobeforums.com
2007-04-05 04:35:26 UTC
Permalink
But surely a kerning editor would pose the same problems for you Yes and
no. Yes, because there could be reflowing text. No, because time is money
and the nearer the press date the less time we have for applying and checking.






spacing is inside the font. O.k., I misunderstood your point before your
late posting. In that you're right.





Tobias
D***@adobeforums.com
2007-04-05 04:51:29 UTC
Permalink
No, because time is money ...
Ah, but if you made individual character styles for each of your adjusted kern pairs, it would be as quick as a kerning editor. For example, a character style called "T." to adjust kerning between "T" and ".". Then it's just a matter of adjusting the value in the style, as in an editor.
M***@adobeforums.com
2007-04-05 13:33:06 UTC
Permalink
It's not a look that I'm fond of.




My example wasn't meant to be an indication of preference. It's just plain out wrong not to have a space between sentences.

And yes, because I must use optical tracking for that particular font, I am inclined to always use optical. I know ID's kerning algorith will produce better results because it won't do any of the specific things I have to avoid with the metrics. Besides, I like the attribute of optical kerning that Dave points out. Smaller type sizes are adjusted automatically.
J***@adobeforums.com
2007-04-05 18:32:28 UTC
Permalink
I like to use optical kerning because I'm not yet done with the JS I'm writing to replicate the functionality of Quark's kerning table editor. I typeset languages like Hmong and Somali on a regular basis, and optical kerning is the best & fastest way I've found to make words like "qaybgeshaa" and "xwm txheej" look decent.

The idea of making character styles makes sense if you're setting type in English (or another language to which type designers pay attention), and you know that the font you're using has poor metrics for particular pairs of characters. However, I'm working in languages for which common character pairs have not been given careful attention by a type designer. In particular, metrics around q, x, and v are generally poorly suited to my work, in almost every font I've ever seen. There are so many pairs that need to be tweaked manually that I'd rather not manage the massive stack of character styles needed to use that method.
M***@adobeforums.com
2007-04-05 19:59:53 UTC
Permalink
I typeset languages like Hmong and Somali on a regular basis




Don't both of those languages have their own alphabet? You must be setting Roman transcriptions then, eh. I get the idea that the original Hmong alphabet is falling into disuse, but surely the same can't be said for Somalian.
J***@adobeforums.com
2007-04-05 20:50:32 UTC
Permalink
Both Somali and Hmong have been written in Latin script for quite some time. I've seen the the Hmong Pahawh script, but it was shown to me by my Hmong translator as a curiousity, not as a writing system that people actually use. Likewise, there have been a few different scripts used to write Somali during the twentieth century, but at least for populations living in the US, Latin script is pretty standard. Omniglot reports that the 1969 coup had as "one of its stated aims the resolution of the debate over the country's writing system." (Omniglot sometimes has inaccuracies in its historical data, so I'm going to ask some of my translators, who actually lived through it, for confirmation.)

You could call contemporary Vietnamese a "Roman transcription," but most people who still know how to write the old Ch&#7919;-nôm script are scholars. If people grew up with Latin script in schools and such, is it still valid to think of it as a Roman transcription of some other script?

(Sorry for the derail, folks. I can't help myself.)
M***@adobeforums.com
2007-04-05 21:45:17 UTC
Permalink
(Sorry for the derail, folks. I can't help myself.)

Well I'll take part of the blame. Thanks for answering. I thought Somali would have been similar to Ethiopian, but now that I know different it explains a sign I saw recently on a grocery store: top line was Ethiopian and the bottom was what I thought was an English transcription of the Ethiopian. It was kinda strange though, and I'm guessing now that it was Somalian.

I have an interest in this stuff myself. I'm intrigued by foreign writing systems and took it upon myself to learn to read and write Lao. The Pahawh obviously has similar Indian roots with Lao and other SE Asian scripts. Haven't seen original Vietnamese, but I understand it was more like Chinese.
Loading...