Comparison to PHOENIX in the CaT region

Hi,

I've synthesized a bunch of spectra in the 3000A to 10000A wavelength range using Korg on a logg/teff/feh/alpha grid that I can compare with PHOENIX. One of the features I observe is major difference around the CaT (in the wings particularly).
here's a T=4500K, logg=5, [Fe/H]=0 [a/Fe]=0 case
I understand that you are not responsible for PHOENIX, but I heard a view from an expert )) that we see here is more likely korg's deficiency rather than PHOENIX due to some kind of strong line treatment. Do you have any advice/opinion on that ? Or maybe I'm not using the right options ?

Thanks

PS I just used the standard Korg synthesize call with MARCS models + all the lines from VALD + individual alpha element abundances specified according to [Fe/H], [a/Fe].