I tried to use EBGaramond otf fonts in writer, but when I export to pdf, some letters have error in shapes, I tried the same font with Scribus and Microsoft word but the exported pdf look perfect without errors.
Steps to Reproduce:
1- Open writer
2- Write some text with ebgramond font, and type some capital letters like N, M, T, F, Z
3- export to pdf
4- open pdf with some viewer and zoom letters to maximum
some letters have wrong shapes and are corrupted
All letters should look fine
User Profile Reset: Yes
Created attachment 152576 [details]
OTF Wrong drawings in exported pdf
1. Attach your EBGaramond font here
2. Attach sample document in ODT format
3. Add info from Help->About dialog
Created attachment 152597 [details]
source odt + pdf output + version + fonts
Here's the font and the odt file plus the output pdf with screenshot of writer version.
Created attachment 152598 [details]
And here's a sample of scribus pdf output using the same font, another remark is that writer deforms the output shape of font letters, they are not like scripus output.
[Automated Action] NeedInfo-To-Unconfirmed
I confirm it with
Version: 22.214.171.124 (x64)
CPU-Threads: 4; BS: Windows 10.0; UI-Render: Standard; VCL: win;
Gebietsschema: de-DE (de_DE); UI-Sprache: de-DE
This is a dupe of bug 99114 or bug 105979 where in https://bugs.documentfoundation.org/show_bug.cgi?id=105979#c12 Khaled notes "our subsetter does not support producing CFF fonts" and we get difference between PDF generated by gs printing (embedding CFF font) and our PDF export (embedding PS Type1 font).
Per 105979 this is inherited from OOo where Herbert Dürr (who implmented the CFF -> PS Type1 subsetting) comments in https://bz.apache.org/ooo/show_bug.cgi?id=43029#c88
Caolán's https://gerrit.libreoffice.org/#/c/47607/ for bug 114704 looks to not have fully resolved issue of bug 105979 but maybe on the right track
Guess the question might be if now, 10 years on, we can fully move to CFF and drop the Type1 subsetting altogether?
*** Bug 99114 has been marked as a duplicate of this bug. ***
*** Bug 105979 has been marked as a duplicate of this bug. ***