# Tighten spacing in multiple-element subscripts that include primes and uppercase-Greek letters

My MWE compiled with pdfLaTeX:

\documentclass[12pt]{article}
\usepackage{amsmath}

\begin{document}

$$\frac{ \Gamma_{v'J'\Lambda'} \Gamma_{v'J'\Lambda' \rightarrow\Lambda''} } { (E-E_{v'J'\Lambda'})^2 + \left( \Gamma_{v'J'\Lambda' \rightarrow \Lambda''}/2 \right)^2 }$$

\end{document}


Result:

This is just a fraction of the equation and the subscripts obviously need to get a bit tighter; the primes seem to get the same space as a full character. Any advice is appreciated. Cheers /J

• i'm not sure the extra space is due entirely to the primes. i'm not able to test just now, so am making a suggestion. try this fraction without the primes; i think you will find that there is space between the v, j, and \Lambda. if true, then the recommended method for closing things up is to "backspace" after the primes with \!. the same wouldn't hurt at the beginning of the subscripts following \Gamma. Jul 27 '15 at 16:08

The spacing issues you raise may be related more to the Greek letters in use (Gamma, Lambda) than to the occurrence of the "prime" characters.

For extra-fine control of math spacing, consider using \mkern with positive or negative amounts of mu instead of \, (positive thinspace, equivalent to \mkern3mu when in math mode) or \! (negative thinspace, equivalent to \mkern-3mu).

The following suggestions came about via experimentation. Feel free to adjust the spacing amounts to suit your aesthetics. You could

• Shrink the space between \Lambda and a following ' by inserting \mkern-1.5mu

• Shrink the space between J' and \Lambda by inserting \mkern-2mu

• Shrink the space between v' and J by inserting \mkern-1.5mu

• Shrink the space between \Gamma and the subscript material by inserting \!

• Increase the space between the two "Gamma" terms in the numerator by inserting \,

• Last but not least, make sure the two 2 exponent terms in the denominator are at the same height relative to the adjacent parentheses, say, by switching to \displaystyle in the denominator and not using \left and \right for the parentheses that surround the second summand.

Needless to say, the resulting code won't exactly be easy to read. For sure, don't engage in such fine-tuning until the document is otherwise pretty much finished. In particular, do make sure the material is factually correct first...

\documentclass[12pt]{article}
\usepackage{amsmath}
\usepackage[textwidth=8cm]{geometry} % just for this example
\begin{document}

before:
$\frac{ \Gamma_{v'J'\Lambda'} \Gamma_{v'J'\Lambda' \rightarrow\Lambda''} } { (E-E_{v'J'\Lambda'})^2 + \left( \Gamma_{v'J'\Lambda' \rightarrow \Lambda''}/2 \right)^2 }$

\bigskip
after:
$\frac{ \Gamma\!_{v'\mkern-1.5muJ'\mkern-2mu\Lambda\mkern-1.5mu'} \, \Gamma\!_{v'\mkern-1.5muJ'\mkern-2mu\Lambda\mkern-1.5mu' \rightarrow \Lambda\mkern-1.5mu''} } { \displaystyle (E-E_{v'\mkern-1.5muJ'\mkern-2mu\Lambda\mkern-1.5mu'})^2 + (\Gamma\!_{v'\mkern-1.5muJ'\mkern-2mu\Lambda\mkern-1.5mu' \rightarrow\Lambda\mkern-1.5mu''}/2 )^2 }$

\end{document}

• You are absolutely right about the second power! Thanks a lot for the solution. I added my own answer below that made it more compact. Jul 28 '15 at 13:25

I'm gonnna add my own solution based on that by Mico. I dug into the \newcommand syntax and created some aliases in the preamble that makes my equations look like I want them and makes the equation code more compact. This is a working example:

\documentclass[12pt]{article}
\usepackage{amsmath}

% My commands!
% kerning:
\newcommand{\mk}[1]{\mkern-#1mu}

% primes:
\newcommand{\myp}{'\mk2}
\newcommand{\mypp}{'\mk{2.5}'\mk2}

% Lambda:
\newcommand{\Lp}{{\Lambda\mk2\myp}}
\newcommand{\Lpp}{{\Lambda\mk2\mypp}}

% J:
\newcommand{\Jp}{{J\mk1\myp}}
\newcommand{\Jpp}{{J\mk1\mypp}}

% v:
\newcommand{\Vp}{{v\myp\mk1}}
\newcommand{\Vpp}{{v\mypp}}

% arrowy:
\newcommand{\rar}{\rightarrow}
\newcommand{\LL}{{\Lp \rar \Lpp}}
\newcommand{\eff}{\text{eff}}

%%% Breit Wigner:
\newcommand{\Gtun}{\Gamma_{\mk3\Vp\Jp\Lp}^\text{tun}}

\begin{document}

$$\sigma_\LL (E) = \frac{\pi\hbar^2}{2\mu E} P_\Lp \sum_{\Vp\Jp} (2J'+1) \frac{ \Gtun \Grad } {(E-E_{\Vp\Jp\Lp})^2 + (\Grad/2 )^2 }$$

\end{document}


And this is what it looks like:

• +1 for the use of macros to systematize the kerning corrections. I'd still be tempted to apply \displaystyle to the denominator of the final fraction, so that the second powers are raised ever so slightly relative to the preceding ) symbols. However, that's clearly just a matter of personal taste.
– Mico
Jul 28 '15 at 13:52
• That looks even better, thanks! But if using \left( and \right) this is done automatically. In that case though, the two denominator expressions have parentheses of different hight. Is there a good way to make LaTeX adjust both pairs to the same height (presumably to the bigger of them)? Jul 28 '15 at 14:15
• Actually, \left.\right)^2 will result in the 2 glyph being raised by quite a bit more than if )^2 occurs in displaystyle math mode. A "compromise" suggestion: Define the macro \ddfrac as follows: \newcommand\ddfrac[2]{\dfrac{\displaystyle#1}{\displaystyle#2}} and use it instead of \frac; that way, both numerator and denominator material will be typeset in displaystyle math. You might like the result.
– Mico
Jul 28 '15 at 14:26
• Regarding your other point: How big might the discrepancy in the sizes of the parentheses be? If the discrepancy is quite pronounced, more harm than good will probably come from making all parentheses the same size.
– Mico
Jul 28 '15 at 14:27
• Regarding the different height: In this case they just become annoyingly almost similar... I think I'm going for your suggestion \displaystyle (...)^2 in this case. But if you know a good way to make them equal in height with \left(...\right)-syntax please tell me and I can test that too. Jul 28 '15 at 14:39