[Gzz-commits] manuscripts/AniFont anifont.tex

[Janne V. Kujala]

CVSROOT:        /cvsroot/gzz
Module name:    manuscripts
Branch:         
Changes by:     Janne V. Kujala <address@hidden>        03/10/02 09:53:27

Modified files:
        AniFont        : anifont.tex 

Log message:
        some points

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/manuscripts/AniFont/anifont.tex.diff?tr1=1.6&tr2=1.7&r1=text&r2=text

Patches:
Index: manuscripts/AniFont/anifont.tex
diff -u manuscripts/AniFont/anifont.tex:1.6 manuscripts/AniFont/anifont.tex:1.7
--- manuscripts/AniFont/anifont.tex:1.6 Tue Sep 30 07:23:04 2003
+++ manuscripts/AniFont/anifont.tex     Thu Oct  2 09:53:27 2003
@@ -87,6 +87,13 @@
 
 - nonlinearities - to our knowledge none yet
 
+- assumptions about the contents of the mipmap levels
+
+    - aniso filter might compute trilinear samples using only one level, etc.
+    
+    - for analysis, the contribution of one texel could be represented
+      as the contribution of four texels on a higher mimap level as per
+      the assumption of generating the mipmap levels in the usual way
 
 \section{Surprise: stretch-squish can yield better images in orthogonal 
transformations}
 
@@ -147,6 +154,9 @@
       NV GF FX supports that, nothing else. Others can only bias the lod so 
have to use supersampling 1 or 2x2
 
 
+- benefit / cost ratio analysis: how much slower than trilinear and
+  how must faster than real supersampling -- does aniso filter provide
+  ``optimal'' quality / cost ratio?
 
 - effect was first noticed when fitting a Letter PDF page into 2048x2048 
texture:
   earlier used only part of resolution, anisotropic version surprisingly 
improved
@@ -162,6 +172,10 @@
   can still be useful - a trick we haven't seen used elsewhere
 
 - compared various filters for text
+
+- we use aniso filter not because of optimality for the task but just because 
+  it is widely implemented and optimized and thus it can be used to obtain 
+  a significant improvement for a moderate performance hit.
 
 Seveeral aspects we did not touch