[Kstars-devel] Quaternions & Co (was: Plans for threading in KStars)

[Daniel-Marian Baboiu]

On Sat, 2011-11-19 at 18:43 +0300, Aleksey Khudyakov wrote:

> Refraction correction could be implemented for vectors similarly to what
> we do now. We'll need two lookup tables. One for new Z and one for
> factor for X and Y which are needed to keep vector unit.

Refraction could be implemented as function of Z only; renormalization
of vectors might not be necessary. Near horizon, where refraction is at
its worst (~0.5 degree), correction in Z is 0.008726, and the norm of
the vector changes to 1.000038. A fraction of pixel when plotted on
screen, except maybe at the highest zoom (but then all points in the
displayed field haev the same offset). 

> As for relativistic correction for the Sun. I'm not sure whether we need
> them at all.

I don't think these relativistic corrections are needed. These are
needed only for the highest precision possible. Deflection of light by
the Sun falls in the same category - it is observable in visible light
only within a few degrees from the Sun, and it is an effect deemed
unobservable in 1800, and very difficult in 1919.

Position of Sun using the Meeus truncation of VSOP87 is within 0.75" in
longitude and 0.29" in latitude from full series (standard deviations of
0.15" and 0.075" respectively).

DMB