Open be5invis opened 7 years ago
If it helps, here’s test strings for the International and the the Uralic phonetic alphabet.
@brawer your strings seem to have disappeared.
@brawer @pauldhunt I find this: https://github.com/googlei18n/noto-source/commit/b85e6f0fb52d4e411cd2e76fa2af4c09eb26416a
I created supplementary IPA glyphs that appear in the Full IPA Chart and additional Latin glyphs. But my glyphs are based on Source Han Serif whose proportional glyphs have been derived from Source Serif Pro, rather than based on the original Source Serif Pro.
I wonder if even such glyphs can help IPA support in Source Serif Pro, e.g., as materials. (Incidentally, I created three weights, Light, Regular, and Bold.)
@nozoka That sounds great, thank you!
Currently, Source Serif itself does not support many IPA characters, but IPA support is definitely something to think about. I’d be curious to see which glyphs you have created, in order to assess when and if they can be included in the project.
The best way to contribute glyphs to Source Serif would be forking the repository, and adding the outlines to all seven masters – ExtraLight-Text-Black + Italics + the high-contrast master (found in the contrast
branch) – considering the proportions found within those files. Source Han Serif’s Latin glyphs are scaled derivatives of the original Source Serif, so the closer submissions are to the original design, the better.
Each supp-.cff in the zip file includes my additional glyphs to the corresponding weight of Source Han Serif, and each all-.cff includes both my additional glyphs and the original ones in Source Han Serif. Since Source Serif Pro includes more Latin glyphs than Source Han Serif, not a few glyphs in my addition are not new to Source Serif Pro.
Recently I struggle to create additional glyphs for the variable font version of SSPs. I think I've understood the mechanism of it, but unfortunately, it is difficult to create serif glyphs being consistent through multiple weights, mainly because of lack of available tools.
Thank you @nozoka! I can definitely use those glyphs in a future IPA update of Source Serif.
The drafting quality in your files is very high, I am impressed!
( BTW: to assess your CFF files, I converted them to UFO like this:
for f in $(find . -name "*.cff") ; do tx -ufo -o ${f%.*}.ufo $f; done
)
It’s true – some of the Latin glyphs found in Source Serif have been omitted from Source Han Serif – mainly to save space. The same is true for Source Sans; I know that @pauldhunt drew IPA characters for the family, but they had to be omitted in the Han release (@kenlunde may know more). Therefore, it would be useful to plan IPA as a Source Serif addition primarily.
I don’t know if you are aware of this – it is not necessary to draw all the glyphs in all the weights – most weights can be interpolated from a handful of masters. Source Serif has three masters for each of the slopes, so six in total. I edit those masters with Robofont, but there are other font editors available (also free ones, such as FontForge). One other benefit of working with UFO masters is that drawing can be done with overlaps, which are only removed when generating OTFs (and in the case of a Variable Font – are not removed at all).
Please let me know how you’d like to proceed on this – I can use your submission as-is for a future IPA extension (I’d need to massage it a bit before inserting the glyphs into the master UFOs), or you could fork the repository and implement your additions directly in the source files. You would be credited in any case.
I am very glad that my glyphs can help IPA support in Source Serif Pro! Thank you!
Because I am not familiar with GitHub, I would like to entrust these glyphs to you this time, rather than doing fork. My fullname is Nozomu Katō. If/when ō cannot be used in text, please replace ō with oo, double o.
When I was creating these glyphs, I referenced the widths of horizontal and vertical stems, miny and maxy values of ZzfOopqyj glyphs for ascent-mean-descent lines, the height and ratio of the serif component, etc.. But there might be other factors that should have been taken into account. If any inconsistency is found in my glyphs, please do not hesitate to modify (massage?) it. The tools I am using to create glyphs are FontForge and Calculator included in Windows. So, I was and am not able to do any complex and/or high-level operations.
I've been learning variable fonts and UFO since last week. For future refererence, please correct me if I am wrong:
I did not notice the -ufo mode since "tx -u" does not display it. Thank you for teaching me. I look into what this can be used for, this weekend.
IPA also require many mark attachments, SIL has some good examples in Charis SIL and Doulos SIL if you want to see which diacritics need for IPA.
@frankrolf
About AL5, do you intend to add support for the entire character set in one update? Because there seems to be a few tiers between AL4 and AL5. Basic IPA and African Latin support is obviously the first step, but there are a number of characters in AL5 that don’t seem to be too useful as they’re only used for obsolete orthographies or non-IPA phonetic transcriptions.
What’s obvious to one user might not be so obvious to the other :-)
I am thinking of AL5 as one goal to achieve; and a logical progression from the currently supported AL4. Also, the AL5 label just helps me group things.
FYI – currently I don’t have a specific timeline for working on AL5 support.
What’s obvious to one user might not be so obvious to the other :-)
I meant the next step in the process of supporting AL5. Obviously there are other goals besides AL5, like polytonic Greek and maybe even Hebrew and Arabic.
Here are some examples from Charis SIL, a serif typeface with IPA support:
If you want to see more letters, you can download it from here: https://software.sil.org/charis/download/
Same here (as in #21) – please provide background info, and let me know how you plan to help.