Vibrational Analysis Problems

[StKolev]

Dear CP2K users/developers

I have a problem analysing vibrational frequencies of a chosen atom. I
am performing vibrational analysis of the simple molecule: trimethyl
phosphate, and I want to analyse the valent vibration P=O. For this
reason my input is:

&VIBRATIONAL_ANALYSIS
  NPROC_REP 80
  INTENSITIES
    &MODE_SELECTIVE
     &INVOLVED_ATOMS
      INVOLVED_ATOMS 2
    &END
    &END

Where the involved atom "2" is the oxygen atom from P=O. Unfortunately
in the output I get lots of frequencies, including C-H vibrations in
3000 cm-1. I tried adding the phosphorous atom in the INVOLVED_ATOMS,
but the same problem occured. I also tried the same thing on other
molecules, with the same problem.

I also tried to analyse the vibrational frequencies in a chosen range.

&VIBRATIONAL_ANALYSIS
  NPROC_REP 80
  INTENSITIES
    &MODE_SELECTIVE
     &INVOLVED_ATOMS
      INVOLVED_ATOMS 1 2
      RANGE 1000 1400
    &END
    &END

But in the output I get again a lot of frequencies, far away from that
range. This is the output of trimethyl phosphate in gas phase:
 [FREQ]
       76.001365
      175.485916
      301.837727
      499.294563
      576.061745
      647.550391
      887.902065
     1077.117794
     1132.144099
     1185.118136
     1248.815876
     1432.806129
     1458.712616
     1473.337353
     1506.962366
     3023.183956
     3036.672705
     3049.558262
     3126.724066
     3131.690122
     3143.864104
     3151.974314
     3175.811496
     3198.346532
 My question is: what am I doing wrong? How to analyse the vibrations
of a chosen atom or set of atoms? And the other question: how to make
the output of the vibrational calculations contain only  frequencies
in a chosen range (because in the cases of large systems the output
file gets lots of MB and it is difficult or impossible to open with
MOLDEN) ?

Best Regards
Stefan Kolev