# Adjusting the size of round parenthesis in an equation

I am not sure what is the common rule in adjusting parenthesis in equations. Let me illustrate my question by the following example

\documentclass{article}
\usepackage[T1]{fontenc}
\usepackage{fouriernc}
\usepackage{mathtools}

\begin{document}

\begin{equation*}
f_{\Lambda}(\rho,1)\hspace{2mm}\text{or}\hspace{2mm}f_{\Lambda}\left(\rho,1\right)
\end{equation*}
in case if
\begin{equation*}
f(x)\mapsto f(\rho)
\end{equation*}
which one is preferred ?
\begin{equation*}
f\left(x\right)\mapsto f\left(\rho\right)\hspace{2mm}\text{or}\hspace{2mm} f\bigl(x\bigr)\mapsto f\left(\rho\right)
\end{equation*}

\end{document}


I would say the second option looks better, but in particular, I have adjusted the parenthesis manually by using \bigl( and \bigr(, while the right hand side is adjusted by \left( and \right(. My question is : will the parenthesis adjusted in this way always have the same length?

Thank you

• Neither \bigl-\bigr nor \left-\right. – egreg Sep 22 '14 at 10:10
• Plus, don't use \hspace{3mm} to add space around \text{or}, in this case, use \quad – daleif Sep 22 '14 at 10:17
• @egreg How to produce brackets of the same length in the above setting? – user124471 Sep 22 '14 at 10:22
• Without any \left, \right, \bigl and \bigr. – egreg Sep 22 '14 at 10:26

The final size is a question of opinion, so I'll just give mine, which is very simple:

for these cases, just use unadorned parentheses

Let me say why this is my opinion.

1. When you write $f\left(x\right)$, a thin space is added between “f” and “(”; you can see it in the last display in the picture, that you should compare with the middle one.

2. With $f\left(\rho\right)$, besides the added thin space, the descender in the letter makes TeX choose a bigger parenthesis, that leaves too much blank space at the top and, most important, does not guarantee equal size of the parentheses (left part of the last display).

3. There's no reason for the parentheses to “fully cover” the material the parentheses delimit. The classical example is

$6\left(\sum_{k=1}^{\infty}\frac{1}{k^{2}}\right)=\pi^{2}$


versus

$6\biggl(\,\sum_{k=1}^{\infty}\frac{1}{k^{2}}\biggr)=\pi^{2}$


but here font depending adjustments can be necessary. The fouriernc version turns out to be

(still I'm preferring the bottom one).

So, in the end,

$f(x)\mapsto f(\rho)$


is the better choice.

• Very good answer and useful advices. Thank you! – user124471 Sep 22 '14 at 11:41