| | <<<Back 1 day (to 2016/02/16) | 20160217 |
| Asuran | its kinda intresting how much better results i get when using an other printer driver, i now use as kens told me the windows one, which seems produce nice sharp fonts compared to older one | 12:29.19 |
| | chrisl, do you got an preflight tool or something else to inspect pdfs? i could need someone who tells me if my fonts are embedded | 13:37.01 |
| chrisl | Asuran: there's a PS tool that ships with Ghostscript - pdf_info.ps. That'll tell you. | 13:39.12 |
| Asuran | chrisl, how to execute it? | 13:39.29 |
| chrisl | Asuran: it's in the comments in the file | 13:39.59 |
| Asuran | hmm dont got such file | 13:40.50 |
| | seems the windows version doesn't have it? | 13:40.58 |
| chrisl | Possibly we don't ship toolbin on Windows.... not sure | 13:41.17 |
| Asuran | lol why? | 13:41.26 |
| chrisl | I dunno - not my decision | 13:41.40 |
| | Asuran: maybe Reader will tell you | 13:41.53 |
| Asuran | no | 13:41.58 |
| | adobe wants you to buy acrobat i guess | 13:42.05 |
| | i mean sure 500$ | 13:42.09 |
| | no problem :D | 13:42.14 |
| | or 17$ a month | 13:42.23 |
| chrisl | Isn't there are "Properties" entry in the File menu? | 13:42.42 |
| Asuran | ah there is | 13:43.11 |
| | thanks | 13:43.13 |
| | cool seems to work | 13:43.46 |
| | even when in commandline no message appers loading fonts from system folder or like that | 13:44.00 |
| chrisl | Do you expect it to? | 13:44.48 |
| Asuran | yes and no, i use now different virtual printers | 13:45.06 |
| | and the microsoft one produces nice output, but the message that sccaning system fonts folder for loading doesn't appear there | 13:45.30 |
| | and the file was little bit 0,7kb smaller so i thought maybe theres a problem | 13:45.50 |
| chrisl | That size of difference could be any number of things that wouldn't affect the perceived quality | 13:50.15 |
| Asuran | chrisl, "Querying operating system for font files..." this only appears on other virtual driver ps output | 14:17.49 |
| chrisl | Asuran: you already said that above | 14:19.12 |
| Asuran | well i stick then to the ms printer, its said that only ms printer produces nice sharp looking font result :-/ | 14:24.31 |
| chrisl | Well, it is, presumably, embedding all the fonts | 14:25.37 |
| | Have to go out...... | 14:26.35 |
| Asuran | i cant believe that but idk, yea ^^ | 14:27.09 |
| | thanks anyways | 14:27.25 |
| norbertj | Hi chrisl, do you know if it is possible (with the windows solutions) to have ghostpcl with UFST ? | 15:14.59 |
| Robin_Watts | norbertj: That is not one of the configurations enabled in the vanilla windows solution, no. | 15:28.33 |
| | norbertj: It shouldn't be too hard to add though. | 15:28.57 |
| norbertj | Robin_Wattsl: I'm just looking at the NMake line in the ghostpcl property pages. I think I have add something there. | 15:29.02 |
| Robin_Watts | Do you happen to know what the nmake line needs to change to ? :) | 15:29.23 |
| norbertj | Robin_Watts: problem is that we have customerproblem only reproducible on our controller, not with standalone. I wanted to check whether UFST (charwidth space) is different from Freetype/URW charwidth space. | 15:30.06 |
| Robin_Watts | norbertj: OK, so the easiest thing to do is to: | 15:30.38 |
| | 1) Open the Visual Studio solution in VS. | 15:30.48 |
| | 2) Copy the configuration you want (say Debug) as "Debug-ufst" | 15:31.10 |
| | 3) Edit the nmake line for Debug-ufst so that instead of making "debug" it makes "ufst-debug" | 15:31.32 |
| | and that should be it. | 15:31.37 |
| | oh, and instead of making "debugbsc" it should make "ufst-debugbsc" | 15:32.04 |
| norbertj | Robin_Watts: thanks, that's even easier than I had in mind. | 15:32.11 |
| Robin_Watts | norbertj: I can't see what that wouldn't work, but then this is gs/makefiles/vs.... | 15:33.04 |
| norbertj | Robin_Watts: thanks, going home now ;) | 15:37.10 |
| Asuran | sorry to bother again, but the microsoft printer tend to make: %Creating CID Keyed Font: F0 => Cinzel | 17:10.39 |
| | % and the other i use seems just embed the font | 17:10.39 |
| | so im unsure if i should let it be or look if gs can prevent this conversion? | 17:10.58 |
| kens | I have no clue what you mean | 17:11.07 |
| Asuran | i read postscript fonts are more quality (press) but | 17:11.14 |
| | microsoft printer for ps | 17:11.22 |
| kens | Again, I do not understand what you are talkign about | 17:11.56 |
| Asuran | the cid font looks much sharper while the ttf looks bit blurrier | 17:12.07 |
| | one virtual printer (microsoft) convert the fonts to cid | 17:12.24 |
| kens | Once more you are comparing apples with orangesw. | 17:12.32 |
| Asuran | other dont and the ms ones converted to CID looks more sharp but thiner | 17:12.39 |
| | maybe | 17:12.57 |
| kens | The blurriness is, as I have said before, due to sub-pixel differences in placement causing slightly fifferent results in the anti-aliasing. THe fonts you are talkign about are (almost certainly, and without looking at the source) the same | 17:13.15 |
| | A CIDFont may contain TrueType outlines | 17:13.35 |
| Asuran | the "problem" is the ms printer converts the truetype fonts to CID fonts | 17:13.56 |
| | which looks bit better but thiner. while the other ones tend to just embed the font without conversion which looks stronger but blurrier | 17:14.30 |
| kens | That is not a problem, and is, as I just said, almost certainly irrelevant. THe font outliens are almost undoubtedly TrueType in both cases, and the same in both cases. | 17:14.41 |
| | And since the printer driver is creating a CIDFont, no there is nothign Ghostscript can do to 'prevent ths conversion' since it is happening before we see it | 17:15.12 |
| Asuran | which should i use then? i would tend to let it convert it because the result is better but what is the best to do? i mean if letting embed the real ttf file would be better in general i would do it perhaps | 17:15.38 |
| | hmm okay | 17:15.55 |
| chrisl | There's no conversion happening | 17:16.04 |
| Asuran | but what does %Creating CID Keyed Font: F0 => Cinzel | 17:16.23 |
| | % then means? | 17:16.23 |
| chrisl | That's covered in the PDF reference manual | 17:16.45 |
| kens | It does not matter whether the font is a CIDFOnt or not, in both cases the outlines are very likely TrueType, and indeed the same outlines. WHat results in a difference in your screen display is a tiny shift in the position of the glyphs on the page. Ths leads to tiny sub-pixel differences in the position, and that results in slightly different results when anti-aliasing is applied. | 17:16.52 |
| Asuran | so what should i do now kens, im no expert | 17:17.17 |
| | but i assume you guys are, atleast you seem to know not less about it | 17:17.37 |
| chrisl | You have a solution that apparently produces output you like, stick with it | 17:18.09 |
| kens | Asuran we have reached the limit of support we are preared to give free users. I believe we have answered your questions fully over the last several days. If you have new questiosn we may be prepared to answer them but for now we must get on and attend to ourt paid work. Good luck with your use of Ghostscript. | 17:18.18 |
| | Forward 1 day (to 2016/02/18)>>> | |
IRC Logs