Hi everyone,
 
After some test and trial I have arrived to the conclusion that two bezier curves (you know, these
S-shaped made by almost all CAD programs) are not enough for *all* the analemmas of *any* flat
sundial... In fact, I guess the best choice is taking as nodes our well-nown zodiacal points and... that's it!
Anyhow, this approximation is very good for sundiallig and I believe there is no need to calculate any more
points than these twelve; if we take as well the semi-zodiacal points, I have never come across a worst-case deviation
higher than 1 min from the corresponding 365-points analemmae. The savings of space and computing time are 1 to 15 !
 
Anselmo Pérez Serrada 

Reply via email to