f



Q.: PDF rendering failures, BB PlayBook. Why?

Having finally reached the point of assembling a full proof PDF of a
massive tome we're compiling, seeing it print flawlessly for hard galleys,
and then sending it to our tablets for proofing at odd moments, I was
floored to see how badly the fonts rendered in a B'Berry PlayBook.

Seeing each page was a new, unnerving, helzapoppin experience for me.

Take page 18 as an example. Here it is in a standard Windows PDF viewer:

<http://misc.tlvp.net/PDF-ER/p18inPDFxChangeViewer.png>.

But in the Playbook's version of Adobe Reader, it looks like so:

<http://misc.tlvp.net/PDF-ER/p18inAdobeOnBB-PB.png>.

It looks better on the Playbook in Qpdf, but it's still not entirely right:

<http://misc.tlvp.net/PDF-ER/p18inQpdfOnBB-PB.png>.

Yet on a simple hundred-clam Android 4.4.4 tablet, it all looks just the
way it looks in Windows, or in hardcopy form, w/o any problems at all.

What's all that signify? Do I infer that neither Adobe nor Qpdf did a great
job at providing a Playbook-oriented PDF viewer? That the PDF file itself
is deficient in some way? (Fonts were all embedded, anyway.) Other?

The PDF whose page 18 serves for these three comparison .pngs is also there
in our little PDF E.R.: <http://misc.tlvp.net/PDF-ER/PDFwFlakyP18.pdf>. The
page of interest in that file is of course also p. 18 ("of course", no?).

How'd we make the PDF? We let Word 2007 make it in Word's usual way, from a
..docx file with all graphics linked and embedded, and all fonts embedded.

Relevant remarks, comments, suggestions, tips, advice, hints, kibbitzing,
etc., all gratefully welcomed, as I need to forestall any similar problems
when submitting the PDF file to a PoD DTP service bureau for production.

Many thanks. Cheers, -- tlvp
-- 
Avant de repondre, jeter la poubelle, SVP.
0
tlvp
3/14/2015 1:01:01 AM
comp.text.pdf 5600 articles. 0 followers. ramon (1518) is leader. Post Follow

9 Replies
871 Views

Similar Articles

[PageSpeed] 3

tlvp wrote:
> We let Word 2007 make it in Word's usual way,

Although this is never what people in your situation want to hear, this
problem is typical of MS-Word and nearly unheard of with any other
program. That said it seems to me that the very newest versions of Open
Office are getting in on the act, though it might be that for all the
broken documents I saw, an MS.docx was somewhere in the work chain.

Anyway, both pdfinfo:
name                            type              emb sub uni object ID
------------------------------- ----------------- --- --- --- ---------
Times New Roman                 TrueType          no  no  no     762  0
Times New Roman                 CID TrueType      yes no  yes    763  0
Times New Roman,Italic          TrueType          no  no  no     764  0
Times New Roman,Bold            TrueType          no  no  no     767  0
Times New Roman,Bold            CID TrueType      yes no  yes    768  0
Times New Roman,Italic          CID TrueType      yes no  yes    769  0
ABCDEE+Rockwell                 TrueType          yes yes no     770  0
ABCDEE+Courier New              TrueType          yes yes no     773  0
ABCDEE+Century                  TrueType          yes yes no     774  0
Times New Roman,BoldItalic      TrueType          no  no  no     782  0
ABCDEE+Arial Narrow             CID TrueType      yes yes yes    893  0
ABCDEE+Arial Narrow             TrueType          yes yes no     894  0
ABCDEE+Arial Narrow,Italic      TrueType          yes yes no     897  0
Times New Roman,BoldItalic      CID TrueType      yes no  yes    922  0
ABCDEE+Cambria Math             TrueType          yes yes no    1181  0
ABCDEE+Arial Narrow,Italic      CID TrueType      yes yes yes   1273  0

and Adobe Acrobat tell me that not all fonts are embedded.

Axel
0
Axel
3/15/2015 5:15:23 AM
On Sun, 15 Mar 2015 06:15:23 +0100, Axel Berger wrote:

> tlvp wrote:
>> We let Word 2007 make it in Word's usual way,
> 
> Although this is never what people in your situation want to hear, 

Actually, I'm familiar with, and have learned to repair, other manglings
that Word carries out in conversing from .docx; this will just be another.

> ... this
> problem is typical of MS-Word ... 
> 
> ... [a little snippage] ...
> 
> Anyway, both pdfinfo:
> name                            type              emb sub uni object ID
> ------------------------------- ----------------- --- --- --- ---------
> Times New Roman                 TrueType          no  no  no     762  0
> Times New Roman                 CID TrueType      yes no  yes    763  0
> Times New Roman,Italic          TrueType          no  no  no     764  0
> Times New Roman,Bold            TrueType          no  no  no     767  0
> Times New Roman,Bold            CID TrueType      yes no  yes    768  0
> Times New Roman,Italic          CID TrueType      yes no  yes    769  0
> ABCDEE+Rockwell                 TrueType          yes yes no     770  0
> ABCDEE+Courier New              TrueType          yes yes no     773  0
> ABCDEE+Century                  TrueType          yes yes no     774  0
> Times New Roman,BoldItalic      TrueType          no  no  no     782  0
> ABCDEE+Arial Narrow             CID TrueType      yes yes yes    893  0
> ABCDEE+Arial Narrow             TrueType          yes yes no     894  0
> ABCDEE+Arial Narrow,Italic      TrueType          yes yes no     897  0
> Times New Roman,BoldItalic      CID TrueType      yes no  yes    922  0
> ABCDEE+Cambria Math             TrueType          yes yes no    1181  0
> ABCDEE+Arial Narrow,Italic      CID TrueType      yes yes yes   1273  0
> 
> and Adobe Acrobat tell me that not all fonts are embedded.
> 
> Axel

Hoo, boy! Thanks, Axel. Who'd ha' thunk ... tell Word explicitly to embed
all fonts, and it only embeds the fonts it wants to, and sometimes even
only subsets those. How to fix that?

What's the difference, btw, between "TrueType" and "CID TrueType" in the
font type column? And why should characters be encoded sometimes following
the Unicode standard (uni = yes) and sometimes not (uni = no)? (And: any
way to get Word to use Unicode only in the PDF?)

Thanks for your attentions. Cheers, -- tlvp
-- 
Avant de repondre, jeter la poubelle, SVP.
0
tlvp
3/15/2015 7:03:56 PM
On Sun, 15 Mar 2015 15:03:56 -0400, tlvp wrote:

> On Sun, 15 Mar 2015 06:15:23 +0100, Axel Berger wrote:
> 
>> tlvp wrote:
>>> We let Word 2007 make it in Word's usual way,
>> ...
>> ... this
>> problem is typical of MS-Word ... 
>> ...
>> ... Adobe Acrobat tell me that not all fonts are embedded.
> 
> Hoo, boy! ... tell Word explicitly to embed
> all fonts, and it only embeds the fonts it wants to, and sometimes even
> only subsets those. How to fix that?

Heh-heh ... turns out setting Word's font-embedding options to embed all
fonts (including those sometimes excluded, and never merely subsetting)
isn't a program-wide setting but operates on a document-by-document basis
.... so that it hadn't yet been set for the document in question.

We'll be sure to set it before the next docx-to- PDF-to-print proofing run,
and hope for the best. Meanwhile this has been a real eye-opener, so again:

> Thanks for your attentions. 

Cheers, -- tlvp
-- 
Avant de repondre, jeter la poubelle, SVP.
0
tlvp
3/15/2015 8:09:46 PM
tlvp wrote:
> and sometimes even only subsets those.

That's the normal thing and you'll find few exceptions. Only those
letters are embedded, that are actually used in the current document.
This is meant to make it harder to extract a copyrighted and possibly
expensive font from a PDf comepletely.

> What's the difference, btw, between "TrueType" and "CID TrueType" in the
> font type column?

I haven't the slightest idea. As far as I can remember, I've never seen
CID before. What's funny here is, that the *same* font seems to be used
twice in two different versions. Is that text a collaboration with
contributions from several people on several machines?

> And why should characters be encoded sometimes following
> the Unicode standard (uni = yes) and sometimes not (uni = no)?

No idea, I'm far from a specialist.

Axel
0
Axel
3/16/2015 1:34:23 PM
In article <5506DBDF.7B56A580@B.Maus.De>, Axel_Berger@B.Maus.De says...

> > What's the difference, btw, between "TrueType" and "CID TrueType" in 
the
> > font type column?
> 
> I haven't the slightest idea. As far as I can remember, I've never seen
> CID before. What's funny here is, that the *same* font seems to be used
> twice in two different versions. Is that text a collaboration with
> contributions from several people on several machines?

TrueType means its a TrueType font, CID means its a CIDFont, in this 
case using TrueType outlines. Fonts are not the same as CIDFonts, but 
you can think of them as follows: a Font is limited to 256 glyphs, a 
CIDFont has no practical limit on the number of glyphs it can use. 

THat's why CIDFonts are often used, DroidSans for instance has pretty 
nealry complete coverage of Latin, Chines, Japanese, Korean and 
Vietnamese glyphs, as well as Greek, Hebrew and others.

There's just no way to address all the glyphs that font contains if you 
load it as a TrueType Font, you have ot load it as a CIDFont.

CIDFonts need not have TrueType outlines, they can also have type 1, CFF 
(type 2) and type 3 (user defined) outlines instead or, indeed, as well 
as TrueType outlines, potentially all mixed together and addressed as if 
they were one single font.

 
> > And why should characters be encoded sometimes following
> > the Unicode standard (uni = yes) and sometimes not (uni = no)?
> 
> No idea, I'm far from a specialist.

I doubt if uni=yes referes to them being encoded in Unicode (but without 
looking at the file I can't tell). I suspect this is more likely to mean 
that the fonts haev a ToUnicode CMap which allows the character codes to 
be converted to Unicode, for the purposes of search/copy/paste. 
ToUnicode CMaps are optional.



			Ken
0
ken
3/16/2015 2:17:01 PM
ken wrote in <news:MPG.2f70f63817bcb16b9898e5@usenet.plus.net>
> In article <5506DBDF.7B56A580@B.Maus.De>, Axel_Berger@B.Maus.De says...
>
>> > And why should characters be encoded sometimes following
>> > the Unicode standard (uni = yes) and sometimes not (uni = no)?
>> 
>> No idea, I'm far from a specialist.
>
> I doubt if uni=yes referes to them being encoded in Unicode (but without 
> looking at the file I can't tell)

It's still there: http://tlvp.net/~misc.tlvp/PDF-ER/PDFwFlakyP18.pdf

The fonts used have two encodings: either WinAnsiEncoding or Identity-H 
(then listed as "uni" in pdfinfo's output).

Regards,

Thomas
0
Thomas
3/16/2015 2:23:06 PM
ken wrote:
> I suspect this is more likely to mean that the fonts haev
> a ToUnicode CMap which allows the character codes to
> be converted to Unicode, for the purposes of search/copy/paste.
> ToUnicode CMaps are optional.

You're absolutely right. First it's pdffonts not pdfinfo - I call both
in succession in my batch used for the purpose. Second the help file
explains:

The following information is listed for each font:

name   the font name, exactly as given in  the  PDF  file
       (potentially including a subset prefix)

type   the font type -- see below for details

emb    "yes" if the font is embedded in the PDF file

sub    "yes" if the font is a subset

uni    "yes"  if  there  is an explicit "ToUnicode" map in the PDF
       file (the absence of a ToUnicode map doesn't necessarily 
       mean that the text can't be converted to Unicode)

object ID
       the font dictionary object ID (number and generation)

PDF files can contain the following types of fonts:

       Type 1
       Type 1C -- aka Compact Font Format (CFF)
       Type 1C (OT) -- OpenType with 8-bit CFF data
       Type 3
       TrueType
       TrueType (OT) -- OpenType with 8-bit TrueType data
       CID Type 0 -- 16-bit font with no specified type
       CID Type 0C -- 16-bit PostScript CFF font
       CID Type 0C (OT) -- OpenType with CID CFF data
       CID TrueType -- 16-bit TrueType font
       CID TrueType (OT) -- OpenType with CID TrueType data

Axel
0
Axel
3/16/2015 3:55:44 PM
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03/16/2015 01:34 PM, Axel Berger wrote:
> tlvp wrote:
>> and sometimes even only subsets those.
> 
> That's the normal thing and you'll find few exceptions. Only those 
> letters are embedded, that are actually used in the current
> document. This is meant to make it harder to extract a copyrighted
> and possibly expensive font from a PDf comepletely.

I think that is indeed the normal way, certainly dvips and pdflatex do
that.

This (among 32,768 other reasons) was why we gave up using Word .docx
files for anything except the actual text and the style names. Fixing
up style names (where the user has omitted or buggered around with
them) is simple with a tall (eg A4) screen in Word, and the result can
be transformed quite easily out of OOXML into something useable to
create the PDF. But we're dealing with documents which are mostly
text: some graphics, but no math or complex tables, for which we'd
need a different workflow.

///Peter

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQEcBAEBAgAGBQJVEeP9AAoJEHt9ZfbX6inQiRcH/2F3QLGXhoDDhxXNJW4LHUFX
msy/cLcaen9V/G/+iELv9s7xGtLfoGRSI/6EUSK2GD5Iis4kzfr8AceyGmQLFbn3
mieo+hUBWmOzLkjYKnd90AHPS6WpVKrtOAhXFupAIR3T1jLW1u1slKCRgpjV+yA0
mtNGwJpIIjke7IwkFTZXpxa2XPHI24hCfnWrZuikrDtxMhxLRWw38HQoxr2sabKQ
kehQFPsOuEsyeeFqJ2Pn3MZXJAbD8igNgNr+ux11powgRpUAmo3r1J2lKbDQcAFa
G2CIv6/IrfRFkDWQcFdFUFoTxBbBr3W/IcxVbJD8DXAk7xxmXIiksXkc1G5bNqs=
=CByY
-----END PGP SIGNATURE-----
0
Peter
3/24/2015 10:23:57 PM
On Mon, 16 Mar 2015 14:34:23 +0100, Axel Berger wrote:

> What's funny here is, that the *same* font seems to be used
> twice in two different versions. Is that text a collaboration with
> contributions from several people on several machines?

Nope. One person, one machine, one version of Word (from MS Office).

Even a PDF Word produced after the document has resaved with all fonts
embedded (including common system fonts), and *not* merely subsetted, the
Adobe Reader in the BB PB uses fanciful substitutions for T.N.R. more often
than not, yet the PB's installation of Qpdf gets everything looking right.

A puzzler. Thanks for thinking about it. Cheers, -- tlvp
-- 
Avant de repondre, jeter la poubelle, SVP.
0
tlvp
3/25/2015 8:38:34 AM
Reply:

Similar Artilces:

PDF PDF PDF
For anyone struggling to figure out how to create a PDF in SWX it's pretty simple but you may have to have the Bluebeam version of swx FIRST go to Tools>Options and check "Save as PDF" Then you can save them right from the save as dialog. Maybe I'm the only dumbass that could't figure that out! ;0) But it was in help under "PDF" An easier way is to download a program from www.pdf995.com that "prints" you files to a PDF foramat. This program works with SW and any other program you use to print with. "3d" <jmiller at marvelindustri...

pdf \ text (get rid of text in pdf)
Is there a way to remove all text from PDF? Will extract images work for you? If so, PDF-Tools by Tracker Software will do it. http://www.docu-track.com/ -- Don Vancouver, USA "MarosV" <maros.vranec@gmail.com> wrote in message news:ebb897e1-c8e3-4b3a-9274-dfd9d2c845c3@c4g2000hsg.googlegroups.com... > Is there a way to remove all text from PDF? ...

PDF TO PDF/A
Hello, is it possible to convert a PDF file to PDFA file ? i tried the -dPDFA option and i vefy my pdf file with PDF longlife. I have an error. Thanks for help. ...

ANN: Fly Text to PDF
Hi All: Fly Text to PDF 1.3 is powerful tool which can convert your text files into PDF. This tool is powerful converter tool running on Microsoft Windows Operating System. You can use this tool to convert your text report, text documents and other text files into PDF quickly and easily. You also can set the PDF properties in each text files by using special tags, or set the default properties for every output PDF files. Please visit our website for more information: http://www.medafan.com/pdf-tools For the output sample, please click on: http://www.medafan.com/pdf-tools/license.pdf Key fea...

Convert PDF image to PDF text
Hi, I have a PDF file with image text documents. hence I can't search the text. How can I convert it to searchable PDF file? Thanx You will have to scan it using an OCR (Optical Character Recognition) application and then create a PDF file. -- Don Vancouver, USA <WhiteLen@gmail.com> wrote in message news:1193737890.691267.165820@z9g2000hsf.googlegroups.com... > Hi, > > I have a PDF file with image text documents. hence I can't search the > text. How can I convert it to searchable PDF file? > > Thanx > Any recommendation for OCR application? ...

Postscript to PDF with pdf-tools, pdf-writer, or other
Does anyone have a way to convert Postscript to PDF using pdf-tools, pdf-writer, or other method? Any suggestions? On 14/04/2008, Sean Nakasone <seannakasone@yahoo.com> wrote: > Does anyone have a way to convert Postscript to PDF using pdf-tools, > pdf-writer, or other method? Any suggestions? ps2pdf is standard on any install of Ghostscript, at least on most flavours of Linux/BSD Farrel ...

convert PDF images to PDF text (OCR)
Hi, I have a book which was entirely scanned into a PDF, there are approximately 300 pages. The problem is that each page is an image of the paper page, not selectable text. How can I convert this PDF made of images to a text-selectable PDF? Or directly convert it to Word? I suppose I'll need some form of OCR. Thanks in advance, Raphael Yes, since you have an image, you need to OCR it into data. You will not have much luck with any formating info, but most OCR programs today to a pretty good job if the image is fairly clean . Larry T. larrynospam@nospamjbmsystems.com wrote in message...

scanned image pdf to searchable text pdf
We have a lot of pdf files that are just scanned images of documents. How easy is it to change these into pdf's that we can search for specific words. I believe Adobe Acrobat Capture will do this when the document is originally scanned, but can it use pdf's that have already been created. If it can, can this process be automated to convert 100s of pdf files? Are there other alternatives to Adobe Acrobat Capture as it is not cheap for a large number of documents? I am new to this, so please keep it simple. Thanks for any help. Adrian HI, As you indicated, your pdf's are ...

PDF::Writer, create pdf and insert in other pdf file.
Hi folks(wharever it mean my english is so bad...) the following code create a PDF: #!/usr/bin/ruby require 'pdf/writer' pdf = PDF::Writer.new pdf.text "Hello Matz", :font_size => 22 pdf.save_as "novo.pdf" ######################################## now I wanna save this pdf as a last page of a existing pdf. any tips? -- Posted via http://www.ruby-forum.com/. On Wed, Mar 26, 2008 at 1:23 PM, Ricardo Pog <ricardo85x@gmail.com> wrote: > Hi folks(wharever it mean my english is so bad...) > > > the following code create a PDF: > >...

PDF TO PDF MULTIPAGES
Hello, (gs 8.54 -Windows XP) i try this to convert a pdf file to one pdf file per page: gs.exe -sDEVICE=pdfwrite -o c:\fic-%d.pdf -q c:\colle.pdf It doesn't work. Is works with jpeg files: gs.exe -sDEVICE=jpeg -o c:\fic-%d.jpg -q c:\colle.pdf Thanks for Help. Christophe BADET Christophe BADET wrote: > Hello, > > (gs 8.54 -Windows XP) > i try this to convert a pdf file to one pdf file per page: > > gs.exe -sDEVICE=pdfwrite -o c:\fic-%d.pdf -q c:\colle.pdf > It doesn't work. > Is works with jpeg files: > gs.exe -sDEVICE=jpeg -o c:\fic-%d.jpg -q c:\colle....

pdf to dvi to pdf?
I need help with what seems to be a simple problem. I have a collection of pdf manuscript to submitt to a on-demand publisher. They want them to be: (1)6in x 9in with 1/2in margins allound (2) and embedded fonts. The problem is there are: (1) 8.5in x 11in with 1in margins (2) and the fonts aren't embedded. Perl pdf mod is one possibility, what would be involved with doing this with LaTex? I wanting to stay with open source solutions(Linux). TIA, David editor@deepmeanings.com (David Jackon) schrieb: > I need help with what seems to be a simple problem. > I have a collection of ...

PDF::API2
Hello All, I am new to PDF files so I don't really know if what I want to do is possible and how to use the PDF::API2 modules. I need to extract information from columns in a table ( I assume that PDF does not know anything about tables). What I was thinking of doing was finding the horizontal location of the header (I know what it should be), then extract all text that starts at that location. I have played around with the PDF::API2 module and read the 'Using PDF::API2 - The code' help page, however it doesn't show me how to extract information from an exi...

Copy from a PDF to PDF
Is it possible to copy a piece of text from a page (not the whole page) from a PDF file and paste it into another PDF file in a desired position? Thanks in article YT2Fb.22892$vi2.19893@twister.auna.com, jesusp at jesusp@arrakis.es wrote on 12/20/03 1:09 PM: > Is it possible to copy a piece of text from a page (not the whole page) from > a PDF file and paste it into another PDF file in a desired position? - Yep - copy and paste. You may have to crop it first to get just what you want. And it may not paste into position - you need to move it around. - I suppose it depends on which ap...

4x PDF Help
Hi Everyone -- we've launched a new question and answer site for all PDF questions. It's based on the highly popular technology platform used by StackOverflow.com, and provides a great environment for getting your questions answered and answering other people's questions. Check it out --> http://help.4xpdf.com/ Cheers, - Rowan. ...

Web resources about - Q.: PDF rendering failures, BB PlayBook. Why? - comp.text.pdf

Color rendering index - Wikipedia, the free encyclopedia
The color rendering index ( CRI ), sometimes called color rendition index , is a quantitative measure of the ability of a light source to reveal ...

Rendering Facebook Connect in Other Languages
Rendering Facebook Connect in Other Languages

apple cupertino rendering 3 by Shay Levanon and Amir Levanon - Flickr - Photo Sharing!
issuu.com/amirlevanon/docs/cupertinoapple2030?mode=window...

iPhone 5 Renderings Based on Leaked Parts - YouTube
This is a CAD rendering of the upcoming new iPhone 5, based upon available sources of supposed leaked parts. More information and many more renderings ...

Pop singer whose smooth rendering of country music made her a star
Patti Page offered a soothing counterpart to the revolutionary new sound of rock'n'roll.

Boston 2024 Olympic Designs & Renderings Released [PHOTOs] - BostInno
Boston's possible bid for the 2024 Summer Olympics cleared a major hurdle on Thursday night, when the United States Olympic Committee selected ...

Google Chrome/OS to use new rendering engine ‘Blink’ based on predecessor Webkit
In a surprise announcement made at the Chromium Blog today, Google announced that Chrome OS, Chrome, and Opera will use a new rendering engine ...

This is where Apple will give keynotes in a few years and a hundred other Campus 2 renderings
[ Cupertino.org ] Related Articles Apple CFO Peter Oppenheimer sends brochure to Cupertino neighbors inviting feedback on new ‘Campus 2′ Apple ...

New renderings released of proposed NFL stadium in Inglewood
New renderings have been released of St. Louis Rams&#39; owner Stan Kroenke&#39;s proposed &#36;1.86-billion venue at the former Hollywood Park ...

Congress is great … at ducking issues, rendering itself irrelevant
According to the Associated Press, Senate Majority Leader Mitch McConnell "grudgingly" conceded this week that he and his fellow Republicans ...

Resources last updated: 1/25/2016 9:37:05 PM