To convince yourself of this, consider the output of

cd ~/rsrch/data/z_piezo_calib         (should be rsrch/data/calibcant)
for X in A B C D; do
  cc_analyze.py -b `cc_bump_analyze.py -ct 20080919_cant${X}_tweak.bump`
                -t `cc_T_analyze.py    -ct 20080919_cant${X}_tweak.Ts  `
                -v `cc_vib_analyze.py  -ct 20080919_cant${X}_tweak.vib -p`
                --verbose --plot-inputs
done

The Q for cantilever D is so low it's hard to see resonance at all,
but it looks like at least 25 kHz.  For cantilever A, it's the ?second
resonance? by then.  Not the first anyway.

Hmm...  The tweakfiles are actually working fine, we just need to keep
scripts "20080919_cant_analyze.sh" with the target files and freq ranges.
Much better solution, but don't forget about it ;p.