I am simulating O K-edge from LiNi0.5Mn1.5O4 with partial fluoride substitution. 

I am starting from a DFT simulated cif file with 30 different oxygen atoms, so I am simulating each Oxygen for a later average and comparing them with experimental results.

I am using these flags:

 *  O K edge energy = ##.## eV
EDGE    K
S02     1.0
 *         pot    xsph  fms   paths genfmt ff2chi
 CONTROL   1      1     1     1     1      1
 PRINT     1      0     0     0     0      0

 *         ixc  [ Vr  Vi ]
 EXCHANGE  0

 *         r_scf  [ l_scf   n_scf   ca ]
 SCF       6.50

 *         kmax   [ delta_k  delta_e ]
 XANES     4.0       0.05     0.3

 *         r_fms     l_fms
 FMS        6.50  0

 *         emin  emax   eimag
 LDOS      -30   20     0.1

 RPATH     0.1
 *EXAFS     20

POTENTIALS
*    IPOT  Z   Tag
      0     8   O
      1     8   O
      2    25   Mn
      3    28   Ni
      4     3   Li
      5     9   F

ATOMS
*    x         y         z       ipot  tag   distance  site_info
    0.00000   0.00000   0.00000    0   O     0.00000  * O_25

(...)

I just finished some simulations and I found the energy limits are shifted in xmu.dat for the different oxygen sites, just in absolute (omega) and relative (e) energy.
As I want to see the relative intensities in the pre-edge, I am worried about normalizing the energy before averaging all the spectra.
Is this ok? Is there a way to normalize the energies or am I making any mistakes? I m attaching a plot of three spectra.

Best regards.