<font size=2 face="Arial">Hi Gerd,</font>
<br>
<br><font size=2 face="Arial">we have ever 2 problems: how to calculate
the next value and how we have to encoding this value.</font>
<br>
<br><font size=2 face="Arial">#1: If we have not a plateau, then we have
a "Standardwert". For the calculation of a "Standardwert"
we have to 3 differend cases depend on the value hdiff(n, m-1) = h(n, m)
- h(n-1, m) (horizontal difference, see at the end of "Datenberechnung").
max is highest possible value in this 64x64-subtile.</font>
<br>
<br><font size=2 face="Arial">#2: We have to decide, is it "Längencodierung",
"Hybride Codierung" or if they are not possible, "Binärcodierung
für große Zahlen". But we have no choice, we have to use that kind,
that the algorithms say (see "Umschaltung der Codierart"). That
do in TileEncoder.cs the class CodingType and the derived classes.</font>
<br>
<br><font size=2 face="Arial">I'm sorry, but TileEncoder.cs is not written
for speed or good understanding. It is only for exploring DEM and, in Input2,
for showing, what's going on. There is much place for optimizing. ;)</font>
<br>
<br>
<br><font size=2 face="Arial">Frank</font>