X-Git-Url: https://git.lukelau.me/?p=opengl.git;a=blobdiff_plain;f=assignment-2%2Freport.latex;fp=assignment-2%2Freport.latex;h=6730aa9bfb4dd75762ef3e2336fd44ddb4d90b23;hp=0000000000000000000000000000000000000000;hb=150ba29f1ce70777762cb7fc44e00e52339bb66e;hpb=5fa27af6222eab1f43e6ca01f1c37cbfa1d86fcd;ds=sidebyside

diff --git a/assignment-2/report.latex b/assignment-2/report.latex
new file mode 100644
index 0000000..6730aa9
--- /dev/null
+++ b/assignment-2/report.latex
@@ -0,0 +1,67 @@
+\documentclass{article}
+\usepackage{fullpage}
+\usepackage{graphicx}
+\usepackage{minted}
+\begin{document}
+\title{Assignment 2}
+\author{Luke Lau 16336810}
+\maketitle
+
+\setminted[c++]{stripall,linenos,breaklines,tabsize=4}
+
+\section{3D}
+For 3D, I first had to add an additional number of vertices to create a square pyramid model:
+
+\inputminted[firstline=181,lastline=198,gobble=1]{c++}{../main.cpp}
+
+The vertex shader then got 3 new matrices: the model, view and projection matrices.
+
+\inputminted{glsl}{../vertex.glsl}
+
+These matrices are computed and bound during the display function:
+
+\inputminted[firstline=36,lastline=45,gobble=2]{c++}{../main.cpp}
+\inputminted[firstline=62,lastline=62,gobble=2]{c++}{../main.cpp}
+
+\texttt{GL\_DEPTH\_TEST} also needed to be enabled so that each side of pyramid was drawn in the correct order (lower Z values first, higher Z values last).
+
+\section{Transformations and Animations}
+
+Various scale, rotation and translation transformations can be performed by manipulating the model matrix.
+To animate it, the amount that the model is transformed by can be multiplied with the elapsed time at the time of drawing.
+\inputminted[firstline=47,lastline=59,gobble=2]{c++}{../main.cpp}
+It was important to translate the model into its position in the world first, before performing any of the rotation or scaling, otherwise the rotation/scaling ended up affecting its position.
+
+\section{Multiple Pyramids}
+Originally I planned on duplicating the vertices stored in the vertex buffer, but as it turns out is sufficient enough to redraw the same 12 vertices and just translate them beforehand.
+
+\begin{figure}[!htb]
+\minipage{0.32\textwidth}
+  \includegraphics[width=\linewidth]{plain}
+  \caption{The pyramids with no transformations}
+\endminipage\hfill
+\minipage{0.32\textwidth}
+  \includegraphics[width=\linewidth]{rotateAndScale}
+  \caption{Scaling and rotation applied}
+\endminipage\hfill
+\minipage{0.32\textwidth}
+  \includegraphics[width=\linewidth]{translate}
+  \caption{Sinusoidal translations, individual to each pyramid}
+\endminipage
+\end{figure}
+
+\section{Camera and movement}
+
+I handle keyboard input by storing a map of which keys are currently pressed down, and then updating the position on a timer.
+\inputminted[firstline=210,lastline=246]{c++}{../main.cpp}
+
+The yaw used above is calculated alongside the pitch based on mouse movement.
+\inputminted[firstline=248,lastline=271]{c++}{../main.cpp}
+
+By working on the pitch and yaw of the camera, we can calculate the direction the camera should be facing. This is stored in the \texttt{camFront} vector.
+This is combined with \texttt{camUp}, which defines the roll of the camera, and is used to calculate the view matrix with \mintinline{c++}{glm::lookAt}.
+Since we don't want the camera to be angled on its side, \texttt{camUp} is left as just the $y$ normal vector.
+
+
+
+\end{document}