Open davidmjones opened 7 years ago
Thanks for the feedback. This will be reviewed and considered for inclusion in the next release.
Original comment by: stipub
I'd like to confirm that I'm also seeing this bug with MS Word 2016. In MS Word, the positioning of the subscript only looks bad when the letter immediately preceding the subscripted character is one of {V, Y, W}. I only checked the range from U+0041 to U+007A, though.
I do not see the same behavior with LuaLaTeX (MiKTeX,version 1.0.4) and unicode-math.
Original comment by: gilberhb
Logged to investigate in 2.0.1. This will likely involve updates to the MATH table cut-in kerning.
I'd like to request moving this to 2.1.0
I've investigated and it looks like it is most likely related to cut-in kerning, but the present MS MATH table editing tool is rather crude, and I'm not sufficiently familiar with it to tackle the work yet. I don't want to delay delivery of the other 2.0.1 fixes and extensions, which are now complete.
We ended up not doing review of MATH table cut in kerning for v2.10, but it is a priority item for 2.20.
Capital N, S or U as the baseline letter also produces somewhat unnatural spacing for subscripts with ascenders, but they're not as offensive as the spacing produced with capital V, W or Y at the baseline.
Also, the problem seems to appear if a letter with an ascender appears anywhere in the subscript, as shown in the example below: (Screen caps produced with STIX Two Math 2.12 b168a and MS Word 16, a.k.a. Office 365.)
I've also tested the same combinations with other math fonts, such as Asana Math and the TeX Gyre math fonts, and the problem doesn't seem to be appearing with those fonts, so it seems like the problem could be mitigated in the math table?
Not sure if this information helps you with debugging/fixing the problem.
Thanks. This is probably a perfect storm of changes we made to some spacing, the italics correction properties in the MATH table, and the cut-in kerning in the MATH table.
Hi,
I've been using Stix 2 with MS Word. I've noticed that ascenders seem to disable kerning. For instance, Y_a looks fine, but the gap in Y_b is too wide. Although I understand that clashes may occur, improvement seems possible!
Reported by: bittarello
Original Ticket: stixfonts/tracking/83