Margin in IEEETran conference mode

[dyhu...@gmail.com]

Hello,

I am trying to submit a paper to IEEE using IEEEtran. The requirement for the margin is top:0.75inch, bottom: 1 inch, side: 0.625 inch. I believe this is the default margin of IEEEtran conference mode, so at first I did not set margin. But when I upload my paper, the system gives a warning saying top and bottom margin is not compliant with the requirements.

Later I add
\newcommand{\CLASSINPUTtoptextmargin}{0.75in}
\newcommand{\CLASSINPUTbottomtextmargin}{1in}
to the beginning of my file, but it still does not work.

Here is my configuration at the beginning of the file, it would be great if anyone can help me check that. Thanks.

\newcommand{\CLASSINPUTtoptextmargin}{0.75in}
\newcommand{\CLASSINPUTbottomtextmargin}{1in}
\documentclass[conference]{IEEEtran}

\makeatletter
\def\ps@headings{%
\def\@oddhead{\mbox{}\scriptsize\rightmark \hfil \thepage}%
\def\@evenhead{\scriptsize\thepage \hfil \leftmark\mbox{}}%
\def\@oddfoot{}%
\def\@evenfoot{}}
\makeatother
\pagestyle{headings}

\usepackage[pdftex]{graphicx}
\usepackage[cmex10]{amsmath}
\usepackage{algorithmic}
\usepackage{algorithm}
\usepackage{amssymb}
\usepackage{cite}

\newtheorem{definition}{\textbf{Definition}}\newtheorem{lemma}{\textbf{Lemma}}\newtheorem{theorem}{\textbf{Theorem}}\newtheorem{corollary}{Corollary}\newtheorem{remark}{Remark}
\DeclareMathOperator*{\argmin}{arg\,min} %
\DeclareMathOperator*{\argmax}{arg\,max} %

\hyphenation{op-tical net-works semi-conduc-tor}

\begin{document}

[Uwe Siart]

dyhu...@gmail.com writes:

> But when I upload my paper, the system gives a warning saying top and
> bottom margin is not compliant with the requirements.

The IEEE PDF checker really sucks. I had this problem before and I think
this is because the PDF checker behaves *very* rigid with no tolerance
at all. LaTeX may put the last line of the page right at the top margin.
This means that the glyphs' descenders lie outside the text area (glyphs
may also have a depth). And this little thing is what the checker
declines.

Increase the bottom margin a little (say by 1mm) and it should be fine.

--
Uwe

[Uwe Siart]

Uwe Siart <use...@siart.de> writes:

> LaTeX may put the last line of the page right at the top margin.

»right at the bottom margin« of course.

--
Uwe

[Michael Shell]

On Fri, 17 Aug 2012 15:25:46 -0700 (PDT)
dyhu...@gmail.com wrote:

> But when I upload my paper, the system gives a warning saying top and
> bottom margin is not compliant with the requirements.

On Sat, 18 Aug 2012 08:17:28 +0200
Uwe Siart <use...@siart.de> wrote:

: Increase the bottom margin a little (say by 1mm) and it should be fine.



If either of you happens to know the minimum amount the bottom margin
needs to be increased to clear IEEE's test, please do let me know.
(e.g., will as little as 1pt or 2pt do the trick?)

Also, (another possible solution) could you try setting:

\topskip 10pt

(use the same point size as your main text size) just after you
\documentclass line and tell me if that small adjustment allows it
to pass IEEE's inspection.

I'll then either tweak IEEEtran to give a little extra margin of safety
or ask that IEEE does so with their margin checker so people won't have
to put up with this in the future.




Cheers,

Mike Shell

[Uwe Siart]

Michael Shell <ne...@michaelshell.org> writes:

> If either of you happens to know the minimum amount the bottom margin
> needs to be increased to clear IEEE's test, please do let me know.
> (e.g., will as little as 1pt or 2pt do the trick?)

For me it's difficult to figure out the exact minimum amount. I'm not a
font expert and I don't know how to identify the largest depth of a
glyph in a font. At least it depends on the font size. Next question is,
which fonts can ever appear in inline material? Times, Computer Modern,
and Computer Modern Math come to my mind. But is that all?

My choice was something like 30% of the font size (3pt for a 10pt
document) and that did the trick for me. But I can't tell if this will
also cover the worst case (inline fraction, or what ever is thinkable).

> I'll then either tweak IEEEtran to give a little extra margin of
> safety or ask that IEEE does so with their margin checker so people
> won't have to put up with this in the future.

My suggestion would be not to tweak IEEEtran but to make the checker
more intelligent. E.g. to give it some tolerance at the bottom margin
that allows glyph descenders to lie outside the text area. While not
being a PDF expert I can imagine that it is also possible to detect
whether the faulty material is graphical material or belongs to font
glyphs. After all the optical margin is at the bottom of the line, not
at the bottom of the descenders. So LaTeX does it perfectly right IMO.

AFAIK IEEEtran.cls won't violate the left, right, and top margin. So the
problem appears only at the bottom. But maybe one day IEEE may also want
to advance to character protrusion (package microtype). Then the checker
should also behave fuzzy about left and right margin.

BTW: The IEEE checker also causes trouble with PSTricks. The \psdots
macro produces Type 3 objects which the checker criticizes as bitmap
fonts. Wrong. The checker should be aware that Type 3 does not
necessarily mean "bitmap font". Type 3 just refers to a group of
PostScript objects with restricted instruction set.

--
Uwe

[dyhu...@gmail.com]

在 2012年8月18日星期六UTC-5下午8时18分41秒，Michael Shell写道：

It still does not work. I increase both top margin and bottom margin for 0.1 inch, but it still does not work.

Without changing anything, the checker reports my top margin is only 0.375inch. But I find if I only delete the following code (which is used to add page number), it will pass the check. Or else, if I increase the top margin to 1.125inch, the checker reports
"margins are 0.667 (L) x 0.666 (R) x 0.75 (T) x 1.014 (B)". But this is obviously not a good way.

Please help me. Thanks.

[dyhu...@gmail.com]

在 2012年8月21日星期二UTC-5下午4时38分27秒，dyhu...@gmail.com写道：

Yes, and if I move the page number to the bottom, then top margin pass the checker, but bottom margin does not. It seems that after adding page number, the corresponding margin is reduced, which results in the failure.

[Michael Shell]

On Tue, 21 Aug 2012 14:38:27 -0700 (PDT)
dyhu...@gmail.com wrote:


> But I find if I only delete the following code (which is used to add page
> number), it will pass the check.

That does suggest that that is the problem. Per IEEE standards, conference
papers are not supposed to have page numbers and this is why IEEEtran
disables them by default under conference mode.

My question now is which conference wants those page numbers/headers?
Because it seems like a case of IEEE's right hand (the PDF checker)
doesn't know what its left hand (a certain conference's requirements)
is doing.


Uwe's problem (given that even a small adjustment will fix it) is probably
not the same. However, that one could be fixed as a result of other
adjustments. If I knew that even as little as 2-3pt (via setting \topskip)
would do it, then I could be sure.


Cheers,

Mike

[Uwe Siart]

Michael Shell <ne...@michaelshell.org> writes:

> Uwe's problem (given that even a small adjustment will fix it) is
> probably not the same.

Could well be that the OP's problem has different reasons. I'm not sure
as well.

> If I knew that even as little as 2-3pt (via setting \topskip) would do
> it, then I could be sure.

When I ran into conflict with the IEEE checker I could track it down to
the glyph descenders in the bottom line. Using a package like
'showframe' I could clearly see that those were the one and only objects
outside the IEEEtran textbody area. So I increased the bottom margin by
1mm (I didn't touch \topskip, I used 'geometry' to get it done, AFAIR)
and all went smooth.

That's all I can report about it. There may be more opportunities for
LaTeX to shift material into the margins. So better don't touch
IEEEtran.cls unless you can be sure it will fix most of the cases.

--
Uwe