I am trying XANES simulation in FEFF9. My system of calculation is SrZnO2 and I am simulating Sr K edge. 

I noticed that on changing SO2 card from 0.0 to 1.0, there is change in position as well as profile of absorption edge. Also, by selecting or deselecting EXCHANGE card, some changes are also there. 

So, which approach is to be followed to obtain trust-able spectra?

Please help.