2

F. Basan et al.

Some units lacked data on day two, limiting direct comparisons, but available
results suggest that under realistic open-water conditions the tested systems converge
 to broadly similar clip level characteristics across most of the operational
range. However, differences between the two frequency ranges (0.8-10 kHz
vs. 2-20 kHz) remain visible, underlining that calibration results are themselves
subject to deviations that should be further investigated to better quantify
uncertainties.

Comparison to Pistonphone Checks

We compared pontoon-derived and pistonphone-based clip levels (converted via fullscale
 voltage; Fig. 4). The comparison was restricted to the overlapping 2-10 kHz
band. Deviations mostly range between —1 and +4 dB, indicating good agreement.
This consistency suggests that the free-field calibrations results are broadly coherent
across systems, though each shows distinct frequency features.
For most recorders, deviations cluster around small positive offsets, indicating
slightly higher measured clip levels than those derived by pistonphone checks
(Le., lower effective sensitivities). A broad minimum in deviation appears around

[

f

h

/

Po

£

1

a
Le]

30
SZ
5

—

27-7
2000

200:

3J00
Frequency [Hz]

— AZ
-FOI ST300HF (pontoon))|
-AU ST600HF (pontoon)
—— BSH Sylence (pontoon}
—— FOI Sylence (pontoon)
= TUT Svience (nantnan)

10000

Fig. 4 Deviation between pontoon-derived clip levels and pistonphone-derived reference clip
‚evels (measured-reference) for the overlapping 2-10 kHz band. Reference values are based on
pistonphone checks listed in Table 3, measured at different frequencies (typically 125-250 Hz; FOI
Sylence at 1 kHz)