confusion with NWRITE=3

[askhetan]

Dear vasp users. I did a frequency calculation for a big ceria system with the following INCAR:

#======Accuracy&Speed======
ENCUT=500
#SYMPREC=1E-08
ISMEAR=-5 # -5 for single point energy calculation, DOS and vibrational, 1 for relaxation
#SIGMA=0.05 # comment out for ISMEAR = -4 or -5
EDIFF=1.E-07 # for DOS, vibrational, comment out for relaxation
ALGO=FAST
LREAL=AUTO
ADDGRID=.TRUE.
IDIPOL=3 # dipole correcion in the z direction
LDIPOL=.TRUE. # use after preconverging
#NPAR=2 # comment out for DOS and vibrational calculations, choose carefully for NEB, IDM
KPAR=4 # comment out for DOS,NEB and IDM calculations, give LAPACK errors

#======Relaxation==========
ISIF=2
IBRION=7 # 2 for relaxation, 7 for vibrational, -1 for DOS
#NSW=400 # comment out for static or DOS or CHARGE, 1 for vibrational
ISPIN=2
NELM=1000
#NBANDS=480 # use only if VASP suggests

#======Verbosity===========
LWAVE=.FALSE.
LCHARGE=.FALSE. # .TRUE. for CHGCAR
#LVTOT=.TRUE. # for absolute local potential
#LAECHG=.TRUE. # for Bader Charge Analysis

#======Vibrational=========
LEPSILON = .TRUE. #for the vibrational analysis
NFREE=2 # other value can be 4 for higher accuracy, use only for vibrational
NWRITE=3 #for dividing vibrational mode eigenvalues by sqrt(mass), only for vibrational

#======DOS=================
#LORBIT=11 # use when DOS is needed
#EMIN=-20 #lower limit of DOS graph
#EMAX=20 # upper limit of DOS graph
#NEDOS=4000 # division of points in DOS graph
#ICHARG=11 #non-self consitent calculation for DOS

#======DFT+U===============
LASPH=.TRUE.
LDAU=.TRUE.
LDAUTYPE= 3
LDAUL= -1 -1 -1 -1 -1 -1 3 #for H,O,H,O,Pt,O,Ce
LDAUU= 0 0 0 0 0 0 5 #for H,O,H,O,Pt,O,Ce
LDAUJ= 0 0 0 0 0 0 1 #for H,O,H,O,Pt,O,Ce
LMAXMIX=6
NSIM=4
LPLANE=.TRUE.

and after that i did, grep THz OUTCAR to get the following result:

Eigenvectors and eigenvalues of the dynamical matrix
----------------------------------------------------


1 f = 104.961540 THz 659.492808 2PiTHz 3501.140015 cm-1 434.086194 meV
2 f = 94.661861 THz 594.778015 2PiTHz 3157.579707 cm-1 391.490129 meV
3 f = 47.980975 THz 301.473357 2PiTHz 1600.473003 cm-1 198.433434 meV
4 f = 23.776863 THz 149.394436 2PiTHz 793.110756 cm-1 98.333237 meV
5 f = 21.536533 THz 135.318025 2PiTHz 718.381377 cm-1 89.067971 meV
6 f = 19.325407 THz 121.425116 2PiTHz 644.626183 cm-1 79.923489 meV
7 f = 17.448276 THz 109.630754 2PiTHz 582.011837 cm-1 72.160297 meV
8 f = 6.462858 THz 40.607333 2PiTHz 215.577722 cm-1 26.728241 meV
9 f = 4.900376 THz 30.789972 2PiTHz 163.458952 cm-1 20.266334 meV
10 f = 4.476339 THz 28.125670 2PiTHz 149.314604 cm-1 18.512658 meV
11 f = 3.803273 THz 23.896672 2PiTHz 126.863544 cm-1 15.729080 meV
12 f = 2.526878 THz 15.876842 2PiTHz 84.287573 cm-1 10.450331 meV
13 f = 2.188588 THz 13.751304 2PiTHz 73.003436 cm-1 9.051276 meV
14 f/i= 6.192176 THz 38.906587 2PiTHz 206.548742 cm-1 25.608789 meV
15 f/i= 6.457086 THz 40.571066 2PiTHz 215.385186 cm-1 26.704369 meV
16 f/i= 7.101363 THz 44.619177 2PiTHz 236.875949 cm-1 29.368885 meV
17 f/i= 8.794710 THz 55.258791 2PiTHz 293.359930 cm-1 36.372009 meV
18 f/i= 13.193513 THz 82.897287 2PiTHz 440.088208 cm-1 54.564003 meV

--------------------------------------------------------------------------------

Eigenvectors after division by SQRT(mass)

Eigenvectors and eigenvalues of the dynamical matrix
----------------------------------------------------

1 f = 104.961540 THz 659.492808 2PiTHz 3501.140015 cm-1 434.086194 meV
2 f = 94.661861 THz 594.778015 2PiTHz 3157.579707 cm-1 391.490129 meV
3 f = 47.980975 THz 301.473357 2PiTHz 1600.473003 cm-1 198.433434 meV
4 f = 23.776863 THz 149.394436 2PiTHz 793.110756 cm-1 98.333237 meV
5 f = 21.536533 THz 135.318025 2PiTHz 718.381377 cm-1 89.067971 meV
6 f = 19.325407 THz 121.425116 2PiTHz 644.626183 cm-1 79.923489 meV
7 f = 17.448276 THz 109.630754 2PiTHz 582.011837 cm-1 72.160297 meV
8 f = 6.462858 THz 40.607333 2PiTHz 215.577722 cm-1 26.728241 meV
9 f = 4.900376 THz 30.789972 2PiTHz 163.458952 cm-1 20.266334 meV
10 f = 4.476339 THz 28.125670 2PiTHz 149.314604 cm-1 18.512658 meV
11 f = 3.803273 THz 23.896672 2PiTHz 126.863544 cm-1 15.729080 meV
12 f = 2.526878 THz 15.876842 2PiTHz 84.287573 cm-1 10.450331 meV
13 f = 2.188588 THz 13.751304 2PiTHz 73.003436 cm-1 9.051276 meV
14 f/i= 6.192176 THz 38.906587 2PiTHz 206.548742 cm-1 25.608789 meV
15 f/i= 6.457086 THz 40.571066 2PiTHz 215.385186 cm-1 26.704369 meV
16 f/i= 7.101363 THz 44.619177 2PiTHz 236.875949 cm-1 29.368885 meV
17 f/i= 8.794710 THz 55.258791 2PiTHz 293.359930 cm-1 36.372009 meV
18 f/i= 13.193513 THz 82.897287 2PiTHz 440.088208 cm-1 54.564003 meV

I learnt from the vasp website that NWRITE = 3 gives the modes divided by sqrt of mass. But what I see is that the frequency values are printed twice. And just before being printed the second time, they have the line that the eigenvectors are divided by mass.

Are these two values supposed to be different? or are they supposed to be the same (i.e. both values are written after dividing by sqrt(mass)) ?

[admin]

you show the eigenvalues of the modes, which of course have to be the same. The eigenvectors are printed in the lines FOLLOWING the frequencies.