Re: [ft-devel] Auto Embolding of Lohit Devanagari
You must not use ftview to test whether glyphs are correctly joined! This programs inserts an additional pixel horizontally between each glyph. For simple stuff, use ftstring. Unfortunately, it doesn't support embolding... Patches welcomed :-) Werner ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On 12 March 2012 13:37, Werner LEMBERG w...@gnu.org wrote: You must not use ftview to test whether glyphs are correctly joined! This programs inserts an additional pixel horizontally between each glyph. For simple stuff, use ftstring. Unfortunately, it doesn't support embolding... Patches welcomed :-) Thanks, i was not aware. In that case can i get patch for freetype with it i can do better testing. - Pravin Satpute ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
With patch it is better now, see the attached image. You must not use ftview to test whether glyphs are correctly joined! This programs inserts an additional pixel horizontally between each glyph. For simple stuff, use ftstring. For complex shaping, you need a demo program from, say, Pango. I hope this comment does not reflect on my patch that is intended to remove excessive space between emboldened glyphs in ftview. No, it reflects Pravin's `better'. I just wanted to make sure that people know that a one pixel space between adjacent glyphs is expected, since ftview inserts this deliberately -- this demo program often has to display zero-width, empty characters, and it is quite useful to see a minimal movement of surrounding glyphs if you increase or decrease the glyph index. Werner ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
I tried this issue (unflat horizontal stroke of Lohit Devanagari when it is autoembolden) on Mac OS X 10.6 on amd64 platform, the issue is reproduced. The result is attached. I was afraid that some overflow occurs in ILP32 platform and does not occur in LP64 platform, but it would not be so simple. I will check the situation in other GNU/Linux in tomorrow. Regards, mpsuzuki suzuki toshiya wrote: Dear Alexei, Thank you for trying to reproduce this issue. Today I'm out of my office and I don't bring 64-bit machine, but I have some (Debian GNU/Linux and Mac OS X on amd64), I will try within a few days. Alexei Podtelezhnikov wrote: On Fri, Feb 24, 2012 at 9:54 PM, suzuki toshiya mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. b) unflat horizontal stroke at the left side. If you look closer the unflatness results from incorrect outline tags at the outline point. I was NOT able to reproduce this on x86-64 with very lightly patched freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel inline: lohit_hi-macosx106.gif___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On 27 February 2012 22:52, Alexei Podtelezhnikov apodt...@gmail.com wrote: On Mon, Feb 27, 2012 at 10:41 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 19:36, Alexei Podtelezhnikov apodt...@gmail.com wrote: On Mon, Feb 27, 2012 at 2:12 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 07:58, Alexei Podtelezhnikov apodt...@gmail.com wrote: mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? Still the major problem is broken upper horizontal BAR, in local language we call it Shirorekha. It should be continuous as it is in Lohit normal. In this particular font the glyph width is larger than it advance, which is unusual. FT_Outline_Embolden does not adjust glyph metrics. So I wonder how you handle the advance values or which higher level text layout library does that for you. Earlier even me too thought it might be problem. But even if we make right and left kerning ZERO result does not change. Text कखगघचाक u0915u0916 I have changes width of only above characters in font (see u'\u0915\u0916\u0917\u0918\u091a\u093e\u0915' characters) Updated TTF file http://pravins.fedorapeople.org/Lohit-Devanagari.ttf Updated Sfd file http://pravins.fedorapeople.org/Lohit-Devanagari.sfd Screenshot http://pravins.fedorapeople.org/Screenshot-lohit-test-3.png Does hinting makes any difference here with Bold? Kerning and advance are different beasts. Your emails lack critical details. Please give more details. We need a snippet of your code to understand why advance values increased larger than they should have been. I am pretty sure it is about the glyph advance. As you can see the whole string is wider than the original one, probably a bit too much. Yes, width is definitely looks more than expected. did you able to reproduce it? I took above screenshot in gedit gnome-shell-3.2.2.1-1.fc16.x86_64 pango-1.29.4-1.fc16.x86_64 freetype-2.4.8-2.fc17.x86_64 gedit-3.2.6-1.fc16.x86_64 yes I think pango is also involve in this. Can i test Freetype auto-embolden with any freetype DEMO? Thanks, Pravin Satpute ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
Please show Lohit Devanagari with ftview, at appropriate size, then, hit space bar once. You may find the auto-embolden shapes. If you hit space bar once more, you may find the auto-oblique shapes. Regards, mpsuzuki pravin@gmail.com wrote: yes I think pango is also involve in this. Can i test Freetype auto-embolden with any freetype DEMO? Thanks, Pravin Satpute ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
ftstring is supposed to be an exemplar string renderer for other libraries because it handles lsb_delta and rsb_delta. This is short of kerning but at least it is better than ftview. Unfortunately, auto-emboldened fonts are not a part of ftstring's functionality. So other libraries could not copy the code from there, I guess. Repeating myself, auto-emboldened glyphs should get auto-increased advances according to the strength of the emboldening and not based on the width. I an not sure how to properly handle this. On Tue, Feb 28, 2012 at 10:43 AM, suzuki toshiya mpsuz...@hiroshima-u.ac.jp wrote: Please show Lohit Devanagari with ftview, at appropriate size, then, hit space bar once. You may find the auto-embolden shapes. If you hit space bar once more, you may find the auto-oblique shapes. Regards, mpsuzuki pravin@gmail.com wrote: yes I think pango is also involve in this. Can i test Freetype auto-embolden with any freetype DEMO? Thanks, Pravin Satpute -- Alexei A. Podtelezhnikov, PhD ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
Can i test Freetype auto-embolden with any freetype DEMO? Yes, e.g. with ftview. Press the space key to cycle through the rendering modes; one of them does artificial emboldening. Werner ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On Mon, Feb 27, 2012 at 2:12 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 07:58, Alexei Podtelezhnikov apodt...@gmail.com wrote: mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? Still the major problem is broken upper horizontal BAR, in local language we call it Shirorekha. It should be continuous as it is in Lohit normal. In this particular font the glyph width is larger than it advance, which is unusual. FT_Outline_Embolden does not adjust glyph metrics. So I wonder how you handle the advance values or which higher level text layout library does that for you. ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
Alexei Podtelezhnikov wrote: In this particular font the glyph width is larger than it advance, which is unusual. ... but it is typical of Devanagari fonts, and in general of several hanging scripts, where the upper bar is required to connect at both ends for most glyphs. I am not sure if it was done because of rounding errors while digitalizing the designs, or if this feature was put there to make sure the line is connected even if the outlines are somehow damaged by rounding; but in any case the Freetype library (and any upper-level libraries as well) should expect such a case if it is to handle Devanagari fonts. Antoine ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On Mon, Feb 27, 2012 at 10:41 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 19:36, Alexei Podtelezhnikov apodt...@gmail.com wrote: On Mon, Feb 27, 2012 at 2:12 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 07:58, Alexei Podtelezhnikov apodt...@gmail.com wrote: mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? Still the major problem is broken upper horizontal BAR, in local language we call it Shirorekha. It should be continuous as it is in Lohit normal. In this particular font the glyph width is larger than it advance, which is unusual. FT_Outline_Embolden does not adjust glyph metrics. So I wonder how you handle the advance values or which higher level text layout library does that for you. Earlier even me too thought it might be problem. But even if we make right and left kerning ZERO result does not change. Text कखगघचाक u0915u0916 I have changes width of only above characters in font (see u'\u0915\u0916\u0917\u0918\u091a\u093e\u0915' characters) Updated TTF file http://pravins.fedorapeople.org/Lohit-Devanagari.ttf Updated Sfd file http://pravins.fedorapeople.org/Lohit-Devanagari.sfd Screenshot http://pravins.fedorapeople.org/Screenshot-lohit-test-3.png Does hinting makes any difference here with Bold? Kerning and advance are different beasts. Your emails lack critical details. Please give more details. We need a snippet of your code to understand why advance values increased larger than they should have been. I am pretty sure it is about the glyph advance. As you can see the whole string is wider than the original one, probably a bit too much. ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On Mon, Feb 27, 2012 at 12:17 PM, Antoine Leca antoine-freet...@leca-marti.org wrote: Alexei Podtelezhnikov wrote: In this particular font the glyph width is larger than it advance, which is unusual. ... but it is typical of Devanagari fonts, and in general of several hanging scripts, where the upper bar is required to connect at both ends for most glyphs. I am not sure if it was done because of rounding errors while digitalizing the designs, or if this feature was put there to make sure the line is connected even if the outlines are somehow damaged by rounding; but in any case the Freetype library (and any upper-level libraries as well) should expect such a case if it is to handle Devanagari fonts. I'm afraid that something relies on the width to predict the advance instead of simply adding the emboldening strength to the original advance. Simple as that. I doubt that this is FreeType's responsibility. ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On 27 February 2012 19:36, Alexei Podtelezhnikov apodt...@gmail.com wrote: On Mon, Feb 27, 2012 at 2:12 AM, pravin@gmail.com pravin@gmail.com wrote: On 27 February 2012 07:58, Alexei Podtelezhnikov apodt...@gmail.com wrote: mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? Still the major problem is broken upper horizontal BAR, in local language we call it Shirorekha. It should be continuous as it is in Lohit normal. In this particular font the glyph width is larger than it advance, which is unusual. FT_Outline_Embolden does not adjust glyph metrics. So I wonder how you handle the advance values or which higher level text layout library does that for you. Earlier even me too thought it might be problem. But even if we make right and left kerning ZERO result does not change. Text कखगघचाक u0915u0916 I have changes width of only above characters in font (see u'\u0915\u0916\u0917\u0918\u091a\u093e\u0915' characters) Updated TTF file http://pravins.fedorapeople.org/Lohit-Devanagari.ttf Updated Sfd file http://pravins.fedorapeople.org/Lohit-Devanagari.sfd Screenshot http://pravins.fedorapeople.org/Screenshot-lohit-test-3.png Does hinting makes any difference here with Bold? Regards, Pravin Satpute ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
On Fri, Feb 24, 2012 at 9:54 PM, suzuki toshiya mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. b) unflat horizontal stroke at the left side. If you look closer the unflatness results from incorrect outline tags at the outline point. I was NOT able to reproduce this on x86-64 with very lightly patched freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
Dear Alexei, Thank you for trying to reproduce this issue. Today I'm out of my office and I don't bring 64-bit machine, but I have some (Debian GNU/Linux and Mac OS X on amd64), I will try within a few days. Alexei Podtelezhnikov wrote: On Fri, Feb 24, 2012 at 9:54 PM, suzuki toshiya mpsuz...@hiroshima-u.ac.jp wrote: I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph)4.8 It looks like advance value is incorrectly adjusted or calculated. I was able to reproduce this on Fedora 16 x86_64, 64-bit build of freetype-2.4.8. b) unflat horizontal stroke at the left side. If you look closer the unflatness results from incorrect outline tags at the outline point. I was NOT able to reproduce this on x86-64 with very lightly patched freetype-2.4.8. Pravin, can you check Fedora x86_64 spin? ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel
Re: [ft-devel] Auto Embolding of Lohit Devanagari
2012/2/25 suzuki toshiya mpsuz...@hiroshima-u.ac.jp Dear Alexei, Thank you again for comment. I think your comment (the lower-left corner of the horizontal stroke is interpreted as Bezier curve) is reasonable, although yet I've not checked how they are interpreted yet (sorry for my laziness). I think there are 2 issues in auto-embolden Lohit Devanagari; a) disconnection of horizontal strokes (which should be connected with previous/next glyph) b) unflat horizontal stroke at the left side. I think issue a) occurs in all glyphs, but issue b) is not for all (however, many glyphs show issue b)). In addition, issue b) is not only for Devanagari. Lohit shows same issue on some ASCII characters (like #, =), so this issue should not be expected as Devanagari only problem. I attached a pair of U+090A and U+090B, showing issue b) (U+090A) and showing no issue b) (U+090B). In attached pair if we check normal version of same pair horizontal upper bar is connected but after Bold it is broken in between. Do let me know if any changes required from font side i tried some though. Thanks, Pravin Satpute Also I uploaded 2 pictures of the outline parsed by fontforge, for U+090A and U+090B. I think there is a difference at the lower-left corner between these 2 glyphs, and it may cause issue b). http://home.hiroshima-u.ac.jp/mpsuzuki/Lohit-Devanagari_U+090A.gif http://home.hiroshima-u.ac.jp/mpsuzuki/Lohit-Devanagari_U+090B.gif Further investigation is needed, please wait (sorry again for my laziness). Regards, mpsuzuki Werner LEMBERG wrote: Lohit Devanagari does not have Bold variant and presently using auto bold algorithm. It does not work correctly in gnome see http://pravins.fedorapeople.org/Screenshot.png It seems that FreeType 2.4.5 shows same problem. I think it is clear what is happening: instead of interpreting a point as a corner between two straight segments, this point is considered a control point of a conic Bezier curve. Since every glyph suffers from it, could it be a first or a last segment? An off-by-one error somewhere? David B., can you have a look, please? The thread starts here: http://lists.gnu.org/archive/html/freetype-devel/2012-02/msg00090.html The font can be downloaded here: https://fedorahosted.org/lohit/ Werner ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel ___ Freetype-devel mailing list Freetype-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/freetype-devel