R help - Apr 2017 - difference metric info of same font on different device

Hi there,

I try to plot with custom fonts, which have good shape Latin and CJK 
characters. I set up all the fonts correctly. However, when I plot the 
same code on png() and postscript(), I get different result. The main 
problem is the space between characters is narrower in postscript() than 
that in png(), and some character also overlap in postscript().  You can 
see the differences from the attached png files.

Is there any way to get the same plot using postscript() and png()? 
Thanks in advance.

Best,
Jinsong

The code I used is here:

windowsFonts(song = windowsFont("SourceHanSerifSC-Regular"),
              hei  = windowsFont("SourceHanSansSC-Regular"),
              hwhei  = windowsFont("SourceHanSansHWSC-Regular"),
              fzsong  = windowsFont("FZShuSong-Z01"),
              fzhei = windowsFont("FZHei-B01"))

postscriptFonts(song = CIDFont("SourceHanSerifSC-Regular", 
"UniSourceHanSerifCN-UTF8-H", "UTF-8", ""),
                 hei  = CIDFont("SourceHanSansSC-Regular", 
"UniSourceHanSansCN-UTF8-H", "UTF-8", ""),
                 hwhei  = CIDFont("SourceHanSansHWSC-Regular", 
"UniSourceHanSansHWCN-UTF8-H", "UTF-8", ""),
                 fzsong  = CIDFont("FZShuSong-Z01",   
"GBK-EUC-H",
"GBK", ""),
                 fzhei = CIDFont("FZHei-B01", "GBK-EUC-H",
"GBK", ""))

fa <- c("sans", "serif", "song",
"hei", "hwhei", "fzsong", "fzhei")

postscript("font.eps", fonts = fa, onefile = FALSE, width = 4, height
=
4, horizontal = FALSE)

#png("font.png", width=4*300, height=4*300, res =300)

plot(0,xlab="",ylab="",type="n")
text(1, -0.75, expression(CO[2]-Hei), family = "hei")
text(1, -0.5, expression(CO[2]-HWHei), family = "hwhei")
text(1, -0.25, expression(CO[2]-FZHei), family = "fzhei")
text(1, 0.0, expression(CO[2]-Sans), family = "sans")
text(1, 0.25, expression(CO[2]-FZSong), family = "fzsong")
text(1, 0.5, expression(CO[2]-Song), family = "song")
text(1, 0.75, expression(CO[2]-Serif), family = "serif")

dev.off()
-------------- next part --------------
A non-text attachment was scrubbed...
Name: postscript.png
Type: image/png
Size: 6388 bytes
Desc: not available
URL:
<https://stat.ethz.ch/pipermail/r-help/attachments/20170407/af6f7b62/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: png.png
Type: image/png
Size: 21368 bytes
Desc: not available
URL:
<https://stat.ethz.ch/pipermail/r-help/attachments/20170407/af6f7b62/attachment-0001.png>

I think it is a fundamental characteristic of graphics drivers that output will
look different in the details... you are on a wild goose chase. Postscript in
particular has a huge advantage in font presentation over other graphics output
mechanisms.
-- 
Sent from my phone. Please excuse my brevity.

On April 7, 2017 1:05:45 AM PDT, Jinsong Zhao <jszhao at yeah.net>
wrote:
>Hi there,
>
>I try to plot with custom fonts, which have good shape Latin and CJK 
>characters. I set up all the fonts correctly. However, when I plot the 
>same code on png() and postscript(), I get different result. The main 
>problem is the space between characters is narrower in postscript()
>than 
>that in png(), and some character also overlap in postscript().  You
>can 
>see the differences from the attached png files.
>
>Is there any way to get the same plot using postscript() and png()? 
>Thanks in advance.
>
>Best,
>Jinsong
>
>The code I used is here:
>
>windowsFonts(song = windowsFont("SourceHanSerifSC-Regular"),
>              hei  = windowsFont("SourceHanSansSC-Regular"),
>              hwhei  = windowsFont("SourceHanSansHWSC-Regular"),
>              fzsong  = windowsFont("FZShuSong-Z01"),
>              fzhei = windowsFont("FZHei-B01"))
>
>postscriptFonts(song = CIDFont("SourceHanSerifSC-Regular", 
>"UniSourceHanSerifCN-UTF8-H", "UTF-8", ""),
>                 hei  = CIDFont("SourceHanSansSC-Regular", 
>"UniSourceHanSansCN-UTF8-H", "UTF-8", ""),
>                 hwhei  = CIDFont("SourceHanSansHWSC-Regular", 
>"UniSourceHanSansHWCN-UTF8-H", "UTF-8", ""),
>                 fzsong  = CIDFont("FZShuSong-Z01",   
"GBK-EUC-H",
>"GBK", ""),
>                 fzhei = CIDFont("FZHei-B01",
"GBK-EUC-H", "GBK", ""))
>
>fa <- c("sans", "serif", "song",
"hei", "hwhei", "fzsong", "fzhei")
>
>postscript("font.eps", fonts = fa, onefile = FALSE, width = 4,
height >
>4, horizontal = FALSE)
>
>#png("font.png", width=4*300, height=4*300, res =300)
>
>plot(0,xlab="",ylab="",type="n")
>text(1, -0.75, expression(CO[2]-Hei), family = "hei")
>text(1, -0.5, expression(CO[2]-HWHei), family = "hwhei")
>text(1, -0.25, expression(CO[2]-FZHei), family = "fzhei")
>text(1, 0.0, expression(CO[2]-Sans), family = "sans")
>text(1, 0.25, expression(CO[2]-FZSong), family = "fzsong")
>text(1, 0.5, expression(CO[2]-Song), family = "song")
>text(1, 0.75, expression(CO[2]-Serif), family = "serif")
>
>dev.off()

On 2017/4/7 23:13, Jeff Newmiller wrote:
> I think it is a fundamental characteristic of graphics drivers that output
will look different in the details... you are on a wild goose chase. Postscript
in particular has a huge advantage in font presentation over other graphics
output mechanisms.
>
Well, the problem stems from the MetricInfo of CID-keyed fonts, which 
are intended only for use for the glyphs of East Asian languages, which 
are all monospaced and are all treated as filling the same bounding box. 
(from the help page of CIDFont)

However, is it possible to use the same MetricInfo of CID-keyed fonts as 
that for png() or windows()?

Best,
Jinsong

On 2017/4/7 23:13, Jeff Newmiller wrote:
> I think it is a fundamental characteristic of graphics drivers that output
will look different in the details... you are on a wild goose chase. Postscript
in particular has a huge advantage in font presentation over other graphics
output mechanisms.
>
I agreed with your opinion on Postscript.

However, as shown in the attached plots in previous post, the glyph 
metric info for any CID-keyed font is based on assumption in R. In fact, 
it's not possible to get the metric info of a CID-keyed font without 
accessing the actual font which may be in truetype or opentype/CFF 
format. And, I don't think R could find the actual font.

Best,
Jinsong

Hi

I think you are hitting the limit of what R's PostScript device can do 
with CID fonts (particularly with Latin characters).

Have you tried the cairo_ps() device ?

Paul

On 7/04/2017 8:05 p.m., Jinsong Zhao wrote:
> Hi there,
>
> I try to plot with custom fonts, which have good shape Latin and CJK
> characters. I set up all the fonts correctly. However, when I plot the
> same code on png() and postscript(), I get different result. The main
> problem is the space between characters is narrower in postscript() than
> that in png(), and some character also overlap in postscript().  You can
> see the differences from the attached png files.
>
> Is there any way to get the same plot using postscript() and png()?
> Thanks in advance.
>
> Best,
> Jinsong
>
> The code I used is here:
>
> windowsFonts(song = windowsFont("SourceHanSerifSC-Regular"),
>              hei  = windowsFont("SourceHanSansSC-Regular"),
>              hwhei  = windowsFont("SourceHanSansHWSC-Regular"),
>              fzsong  = windowsFont("FZShuSong-Z01"),
>              fzhei = windowsFont("FZHei-B01"))
>
> postscriptFonts(song = CIDFont("SourceHanSerifSC-Regular",
> "UniSourceHanSerifCN-UTF8-H", "UTF-8", ""),
>                 hei  = CIDFont("SourceHanSansSC-Regular",
> "UniSourceHanSansCN-UTF8-H", "UTF-8", ""),
>                 hwhei  = CIDFont("SourceHanSansHWSC-Regular",
> "UniSourceHanSansHWCN-UTF8-H", "UTF-8", ""),
>                 fzsong  = CIDFont("FZShuSong-Z01",   
"GBK-EUC-H",
> "GBK", ""),
>                 fzhei = CIDFont("FZHei-B01",
"GBK-EUC-H", "GBK", ""))
>
> fa <- c("sans", "serif", "song",
"hei", "hwhei", "fzsong", "fzhei")
>
> postscript("font.eps", fonts = fa, onefile = FALSE, width = 4,
height > 4, horizontal = FALSE)
>
> #png("font.png", width=4*300, height=4*300, res =300)
>
> plot(0,xlab="",ylab="",type="n")
> text(1, -0.75, expression(CO[2]-Hei), family = "hei")
> text(1, -0.5, expression(CO[2]-HWHei), family = "hwhei")
> text(1, -0.25, expression(CO[2]-FZHei), family = "fzhei")
> text(1, 0.0, expression(CO[2]-Sans), family = "sans")
> text(1, 0.25, expression(CO[2]-FZSong), family = "fzsong")
> text(1, 0.5, expression(CO[2]-Song), family = "song")
> text(1, 0.75, expression(CO[2]-Serif), family = "serif")
>
> dev.off()
>
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
-- 
Dr Paul Murrell
Department of Statistics
The University of Auckland
Private Bag 92019
Auckland
New Zealand
64 9 3737599 x85392
paul at stat.auckland.ac.nz
http://www.stat.auckland.ac.nz/~paul/

On 2017/4/10 10:06, Paul Murrell wrote:
> Hi
>
> I think you are hitting the limit of what R's PostScript device can do
> with CID fonts (particularly with Latin characters).
>
> Have you tried the cairo_ps() device ?
>
> Paul
Thank you very much. It works very well. It's more simple than 
postscript() with CIDFont setting.

Best,
Jinsong
>
> On 7/04/2017 8:05 p.m., Jinsong Zhao wrote:
>> Hi there,
>>
>> I try to plot with custom fonts, which have good shape Latin and CJK
>> characters. I set up all the fonts correctly. However, when I plot the
>> same code on png() and postscript(), I get different result. The main
>> problem is the space between characters is narrower in postscript()
than
>> that in png(), and some character also overlap in postscript().  You
can
>> see the differences from the attached png files.
>>
>> Is there any way to get the same plot using postscript() and png()?
>> Thanks in advance.
>>
>> Best,
>> Jinsong
>>
>> The code I used is here:
>>
>> windowsFonts(song = windowsFont("SourceHanSerifSC-Regular"),
>>              hei  = windowsFont("SourceHanSansSC-Regular"),
>>              hwhei  =
windowsFont("SourceHanSansHWSC-Regular"),
>>              fzsong  = windowsFont("FZShuSong-Z01"),
>>              fzhei = windowsFont("FZHei-B01"))
>>
>> postscriptFonts(song = CIDFont("SourceHanSerifSC-Regular",
>> "UniSourceHanSerifCN-UTF8-H", "UTF-8",
""),
>>                 hei  = CIDFont("SourceHanSansSC-Regular",
>> "UniSourceHanSansCN-UTF8-H", "UTF-8",
""),
>>                 hwhei  = CIDFont("SourceHanSansHWSC-Regular",
>> "UniSourceHanSansHWCN-UTF8-H", "UTF-8",
""),
>>                 fzsong  = CIDFont("FZShuSong-Z01",   
"GBK-EUC-H",
>> "GBK", ""),
>>                 fzhei = CIDFont("FZHei-B01",
"GBK-EUC-H", "GBK", ""))
>>
>> fa <- c("sans", "serif", "song",
"hei", "hwhei", "fzsong", "fzhei")
>>
>> postscript("font.eps", fonts = fa, onefile = FALSE, width =
4, height >> 4, horizontal = FALSE)
>>
>> #png("font.png", width=4*300, height=4*300, res =300)
>>
>> plot(0,xlab="",ylab="",type="n")
>> text(1, -0.75, expression(CO[2]-Hei), family = "hei")
>> text(1, -0.5, expression(CO[2]-HWHei), family = "hwhei")
>> text(1, -0.25, expression(CO[2]-FZHei), family = "fzhei")
>> text(1, 0.0, expression(CO[2]-Sans), family = "sans")
>> text(1, 0.25, expression(CO[2]-FZSong), family = "fzsong")
>> text(1, 0.5, expression(CO[2]-Song), family = "song")
>> text(1, 0.75, expression(CO[2]-Serif), family = "serif")
>>
>> dev.off()
>>