3

I noticed that when we use \underbrace and \overbrace, it creates a bracket that is assembled by several parts, please see the picture below demonstrating what I mean: enter image description here

If we zoom in, we can see: enter image description here

Andenter image description here

Which causes a small inconsistency, and I am puzzled because this is not a glitch, it's how the package made it to be, I just wonder if there's any reason why they make it this way?

Here is a minimal working example:

\documentclass[12pt]{article}
\usepackage{amsmath}
\begin{document}
$\underbrace{\left(A_{N} \cdot \int_{\left(m_{-i}, v\right)} \frac{v}
{\operatorname{rank}\left(m_{i} ; m\right)} \rho\left(d m_{-i}, d v | 
m_{i}\right)-X_{N}, t_{i}(1)\right)}_{ \leq A_{N} \cdot 
t_{i}^{\prime}\left(m_{i}\right)-X_{N}=t_{i}\left(m_{i}\right) \text { if } m_{i}<1}$
\end{document}
8
  • 2
    Please, make a complete example, in particular it's needed to know the font package you're using.
    – egreg
    May 4, 2019 at 21:07
  • @egreg Thank you for reminding me, I've added a minimal code, the only package I use is amsmath
    – The R
    May 4, 2019 at 21:16
  • 1
    This is the picture (click) I get, which is quite different from yours. The font is different.
    – egreg
    May 4, 2019 at 21:25
  • @egreg Yes, the picture is from a journal article and the code is reproduced, but this doesn't affect my question because the gaps are even more obvious in your picture.
    – The R
    May 4, 2019 at 21:28
  • 1
    What you see in the picture is not “the real thing”, because of rasterization at small resolution; I was just pointing out that the font is not the same. Here (click) is the central part at high resolution. The journal is using a different font and that makes the difference. The PDF viewer and your screen can have a role.
    – egreg
    May 4, 2019 at 21:34

1 Answer 1

2

My proposal is only an aesthetic alternative using, for example, \undercbrace from lite mtpro2 package. I have reconsidered some changes of your original code.

enter image description here

\documentclass[12pt]{article}
\usepackage{newtxtext,mathtools}
\usepackage[lite]{mtpro2}
\begin{document}
\[t_i(m_i)\geq \min\undercbrace{\left(A_N\cdot \int_{m_{-i},v} \frac{v}{\operatorname{rank}(m_{-i};m)}\rho(dm_{-i},dv\mid m_i)-X_Nm t_{i}(1)\right) }_{\leq A_{N} \cdot t'_{i}(m_{i})-X_{N}=t_{i}(m_{i}) \text { if } m_{i}<1}\]

\end{document}
2
  • Thanks for your recommendation, I can see in mtpro2 the brackets are "individually designed" according to length so I suppose in amsmath package they didn't do so, so it retains the gap.
    – The R
    May 4, 2019 at 21:34
  • @TheR vertical and horizontal braces are the same, each font has a fixed number of glyphs then an extendable version made up of central and edge parts with repeatable straight extenders. Some fonts have a larger range of designed sizes before using the extenders but the basic principle is the same in any case. May 4, 2019 at 22:45

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .