U 
F. Basan et al. 
considered essential for recovery, and several groups reported positive experiences 
with additional measures such as trawl buoys, contact labels, or photographic 
documentation before deployment. Example deployment designs are presented in 
Fig. 7. Biofouling was identified as a recurring issue, particularly during summer 
months, with mitigation strategies including deeper placement or more frequent 
servicing. While details vary between institutions, these exchanges underlined that 
rg design can influence data quality and that continuous sharing of practical 
solutions across the community is an important part of harmonization. 
Conclusion and 
Outlan 
The HELCOM workshop demonstrated in practice that pistonphone and free-field 
calibrations complement each other. Pistonphones remain a reliable and affordable 
tool for quick functionality checks and for documenting long-term stability, but they 
cannot reveal the frequency-dependent features that matter when interpreting one- 
third-octave band levels. The broadband calibrations carried out in both the tank and 
at the pontoon facility of FOI confirmed that none of the tested autonomous 
recorders showed a perfectly flat response. Instead, system-specific roll-offs, dips, 
and mounting effects became visible. 
At the same time, the pontoon results also showed that under realistic open-water 
conditions, curves converge and appear flatter than in the tank, suggesting that some 
of the deviations seen in controlled facilities may be linked to setup artifacts rather 
{han inherent device behavior. The comparison to nominal sensitivities further 
underlined that free-field results and manufacturer specifications can show substan- 
tial differences. All systems show lower free-field sensitivities than nominal—with 
frequency-dependent undulations. The resulting uncertainty curves confirm that the 
observed tank-pontoon differences are systematic and fall within a reproducible, 
frequency-dependent range. However, the outcomes also underline that the analysis 
presented here is not yet complete: not all of the factors that shape the sensitivity of 
autonomous recorders are fully understood, and further work is needed to disentan- 
gle hardware, deployment, and calibration effects. 
Beyond the calibration results themselves, the workshop provided an important 
forum for exchanging experiences on rig design and practical deployment issues. 
Even small details—such as the use of holders, cable routing, choice of anchor 
weights and rope connections—can affect sensitivity curves and in the long run the 
comparability of long-term monitoring data. Sharing these experiences openly is 
therefore as valuable as the calibration exercises themselves. 
Our tank results further illustrated this point: for the BSH Sylence recorder, using 
a 3D-printed clamp to fix the hydrophone to the rope (to mimic the float-mounted 
setup used in real deployments) systematically shifted the clip level upwards com- 
pared to the free-hanging configuration. This demonstrates that even seemingly 
minor choices in mounting hardware can change the effective device response by 
several decibels.