/[gzz]/manuscripts/AniFont/SCRATCH
ViewVC logotype

Diff of /manuscripts/AniFont/SCRATCH

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.1 by tjl, Fri Oct 17 09:15:22 2003 UTC revision 1.2 by tjl, Wed Oct 22 09:38:05 2003 UTC
# Line 17  samples the mipmap levels in XXX aniso X Line 17  samples the mipmap levels in XXX aniso X
17  }  }
18  \end{figure*}  \end{figure*}
19    
20    
21    The most important area for reading is naturally the center of the fisheye,
22    where the transformation is nearly orthonormal - the
23    edges are mostly used for getting a sense of the context, not for reading.
24    
25    After some investigation, we discovered that we had found a special case
26    of a general principle: if a texture image is only transformed through
27    rotation and isotropic scaling, a better filtering result is always obtained
28    by applying the stretch-squish operation.
29    
30    \begin{figure}
31    \begin{tabular*}{\columnwidth}{rc}
32    a) & \includegraphics[width=5cm]{footprint.1}  \\
33    b) & \includegraphics[width=4cm]{footprint.2}  
34    \end{tabular*}
35    \caption{
36    \label{figfootprint}
37    Pixel footprint in screen space (PFSS) diagram.
38    Texture samples' contribution to a pixel's value.
39    a) An explanation of PFSS diagrams: the diagrams
40    show the contribution of each texel to the pixel
41    as a color (black = no contribution, white = large contribution).
42    b) An example PFSS of an EWA texture filterer without
43    mipmapping (mockup, just diagrammatic).
44    In screen space, the
45    filter is circular and has soft edges, while in texture space it would be
46    elliptical.
47    }
48    \end{figure}
49    
50    Figure~\ref{figfootprint} shows a legend of PFSS diagrams and
51    a diagram for the EWA filtering method.
52    Appendix A shows how
53    PFSS diagrams can be generated easily to show the actual behaviour of a hardware accelerator.
54    
55    In Section~\ref{seccomp}, we compare the performance of different filtering
56    methods, including trilinear, stretch-squish aniso and supersampling, on a test image.
57    
58    
59    - For text, setting of the problem: orthogonal transformations are most important,
60      TrueType shows maybe not the right model but ...
61    
62    
63    - case we consider: sharp edges, orthogonal (or nearly so) transformations, e.g. text
64    
65    - simple solution for improving the situation in one direction:
66      stretch the texture in one direction, squish back by texture coordinates. activate
67      the aniso filter. Aniso filters planned so that they don't flicker, either.
68    

Legend:
Removed from v.1.1  
changed lines
  Added in v.1.2

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26