Superscript and Subscript

[Xcel]

I cannot find any previous listings for this topic so if it has been raised before I apologise.

There seems to be a major problem with superscript and subscript. Only the numbers 1, 2 and 3 and one letter n superscript and only 4 letters and no numbers subscript. The chart i'm including was created with Times New Roman Open Type.

I have used the same text and font in a Pages document and all correctly format.

 

[A_B_C]

Yes, there are still a few issues with superscript and subscript glyphs. Times New Roman, as installed with macOS, does not have the full set of superscript and subscript glyphs, but almost only the ones that are included in the standard encoding. That means, there are no superscript glyphs for the figures 4, 5, 6, …, 9, 0 and no superscript glyphs for the letters b, c, d, f, … in the font file. And almost the same goes for the subscript equivalents.

If a superscript or a subscript glyph is not present in a font, text processing applications will use a scaled down and repositioned version of the standard glyph that is present in the font. This is a fallback solution that currently does not work in Affinity Publisher, presumably due to the slightly smarter way of addressing superscript and subscript (in general, glyph substitution) handling compared to Affinity Designer. But I am sure this will be sorted out.

For the developers it might be interesting to note that you can create a document with superscript and subscript glyphs in Affinity Designer, open it in Publisher, and the substitution fallback will work there as in Designer (blue text in my screenshot). However, when you recreate the same text frame in Publisher, it will use the superscript and subscript glyphs that are present in the font, but fail to apply the fallback solution (red text in my screenshot).

 

[Alfred]

10 minutes ago, A_B_C said:

For the developers it might be interesting to note that you can create a document with superscript and subscript glyphs in Affinity Designer, open it in Publisher, and the substitution fallback will work there as in Designer (blue text in my screenshot). However, when you recreate the same text frame in Publisher, it will use the superscript and subscript glyphs that are present in the font, but fail to apply the fallback solution (red text in my screenshot).

The size and weight of the superscript ‘123’ looks better in the red version. The blue version is too large and too light.

[A_B_C]

Yes, Alfred, that’s because Publisher uses the glyphs that had been drawn by the font designer. So Publisher is a little smarter than Designer at this point. But Publisher doesn’t create (fake) superscripts and subscripts for the missing ones.

[Dave Harris]

2 minutes ago, A_B_C said:

Yes, Alfred, that’s because Publisher uses the glyphs that had been drawn by the font designer. So Publisher is a little smarter than Designer at this point. But Publisher doesn’t create (fake) superscripts and subscripts for the missing ones.

We don't do that automatically because they tend not to match. Instead we give you the choice. There is a control at bottom-right of the Character panel Position and Transform section that will produce fake superscripts and subscripts by scaling and positioning the original glyphs. Using this (and only this) will give superscripts that work for all characters, and will look uniform compared to each other. The control in the Typography section will produce superscripts by switching glyphs (either by OpenType or by our own fallback if the font doesn't do OpenType). This only works if the glyphs are available, but produces nicer results if they are.

Here's a typographic superscript 2, a normal 2, and a fake superscript 2 in Arial. Notice that the two superscripts have difference heights and stem thicknesses. That difference would mean that having a mix of real and fake superscripts in the same text would look uneven, which is why we don't use fake superscripts as a fallback for typographic ones. Also notice that the stem thickness for the typographic superscript is "better", more appropriate to the weight on the full-sized 2. That is why we have to offer typographic superscripts, because they look better when available. And we have to offer fake superscripts to cover for when the typographic superscripts aren't supported for the characters you want.

The difference with Designer is because the user interface 1.6 wasn't very clear about which superscript you were getting. As I recall, the control in the Typography section of the Character panel actually gave fake superscripts rather than typographic ones. The apps should be consistent when 1.7 comes out.

[Xcel]

Thanks Dave and A_B_C. 

I have many fonts from different sources and have been checking for the presence of the super and sub glyphs. I have found that many of the Adobe .otf do have the full set but the procedure described by Dave Harris clearly provides better results for those that don't. This, however, may be a time consuming exercise in a technical document with many superscript references but I guess that it could be saved as a Style.

Thanks again.

[A_B_C]

Thank you, Dave. Somehow I missed this setting on the Character panel.

There is still a slight inconvenience … or bug, if you will … when your font does not have dedicated subscript (or superscript) glyphs and you engage both the subscript (or superscript respectively) options on the Character panel and the Typography panel, you won’t get superscript (or subscript) glyphs.

I believe it would make sense if the Character panel setting overruled the Typography panel setting when there are no dedicated superscript or subscript glyphs in the font. 

[Xcel]

Thanks A_B_C for your very complete response. I had a feeling that it may be font related. I'm not sure of the origin of my Times New Roman as I have fonts from many different sources and some are .ttf some .otf and I even have some Apple .dfont. I also have some fonts I designed myself for special purposes and these would not have the super and sub glyphs.

I, too, missed the Superscript pull down in the Positioning and transform dialogue and thought that Dave Harris was suggesting that the font size and baseline should be adjusted to achieve the superscript.

I have all of the fonts that came with Adobe Creative Suite and I have opened some of these in Fontographer and a number do have all of the glyphs. One example is Adobe Caslon Pro which has all the numerical superscripts and many more alpha superscripts.

Using the unicodes from the Adobe Caslon Pro I have generated the complete set of superscripts for Times. I may have to tweak the spacing but it's a start.

[Alfred]

19 hours ago, Dave Harris said:

Here's a typographic superscript 2, a normal 2, and a fake superscript 2 in Arial. Notice that the two superscripts have difference heights and stem thicknesses. That difference would mean that having a mix of real and fake superscripts in the same text would look uneven, which is why we don't use fake superscripts as a fallback for typographic ones. Also notice that the stem thickness for the typographic superscript is "better", more appropriate to the weight on the full-sized 2. That is why we have to offer typographic superscripts, because they look better when available. And we have to offer fake superscripts to cover for when the typographic superscripts aren't supported for the characters you want.

It seems to me that there’s a strong case for reducing the height and increasing the stem thickness of the fake versions.

[A_B_C]

(I doubt it will be possible to write an extrapolation algorithm that will be able to increase the stem widths of arbritrary fonts in a visually pleasing way. )

[Oval]

Affinity könnte ja zumindest optional eine passendere Schriftstärke nutzen/automatisch auf Bold schalten, falls der Nutzer gar keine Erfahrung hat. 

[A_B_C]

(“AI” is the future. )

[Oval]

8 hours ago, A_B_C said:

(“AI” is the future. )

(Hopefully AD is the future.  Even without AI. )