blob: 4f29bea6fd997fcfb1e44b35a3d987e8af68b673 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
|
\chapter{Color mapping}
In chapter 2 we saw a figure (figure 2) which showed a fluid in motion. The fluid had a very bright grey color. How is this color determined at every vertex?
\section{Description}
The technique that maps a value to a specific color is called color mapping. We already explained that the simulation is divided into a two dimensional array of
different values. A colormap calculates the color, given a certain colormap function, for every value at a vertex. Example: \\
In figure 2 we saw the smoke using a grey scaled colormap. If we know that the values at the vertices ranges from 0 to 1, we can use the value to determine each
color aspect, red, green and blue. This means, for each vertex:
$$ red = green = blue = value $$
To be able to reason about the colored images, we added a legend at the top of the screen. The leftmost colors indicate low values and the rightmost colors indicate
high values. With such a colormap legend, it's easier to understand the produced images and say something about the value of the fluid. \\
You are also able to set the number of used colors. At default this value is set to 256 colors, but you can easily set that amount to 16 using a slider. You will
see big bands of colors appear. This way, the line between certain values becomes more visible. \\
\section{Implementation}
We implemented a number of colormaps.
\section{Difficulties}
The last mechanisms to implement for this assignment were scaling and clamping. With clamping you let the user set a minimum and maximum for the values. Actual data
higher than the maximum or lower to the minimum are clamped to the maximum or minimum respectively. \\
The clamping mechanism was a bit more subtle. At every frame, the minimum and maximum values are stored and the entire dataset at the current time moment is mapped
to the visible colormap. This is not so hard to do, but we had not foreseen that values could also be negative. It wasn't until we implemented the divergence that
we found this problem. \\
|
