SoundTraps, like all high frequency recorders, are highly unlikely to have a flat frequency response. Although there is some calibration information at 250 Hz from the manufacturer this will not reflect the calibration at high frequencies. In addition, the variation in calibration values between devices may be greater at higher frequencies.
Has anyone calibrated multiple SoundTraps across their operation frequency spectrum i.e. up to 160 kHz (the cutoff frequency for the anti-aliasing filter)?
Yes! I have painstakingly done this, and while there is a ~ flat freq. response up to ~80 kHz, there is a distinct hump from ~100-150 kHz (which is of concern if you study NBHF Kogia or porpoies, as I do). See my results below, on the calibration of several ST300HFs. Details on calibration procedure below.
Calibration procedure: Sent pulses in 10 kHz steps to all 15 STs (10-200 kHz for 12 HF STs). This was done in the 3m deep cedar calibration tank. Three different hydrophones projected short pulses at different frequency ranges (HS27: 10-50 kHz. HS70: 60-100 kHz. B&K 8105: 110-200 kHz) from the Agilent wave generator (20 Vp2p, 5 cycles, pinging once every 50 ms, with 2m between transmitting hydrophone and a given ST). Note that I used a fixed # of cycles in each of my tone projections, meaning that the number of samples for a 5-cycle tone is greater for lower frequency tones than for higher frequency tones. All SoundTraps (2 V range) were recording in high gain mode, and were compared to a reference hydrophone (Reson 4034, resonant at 300 kHz, with +50 dB of gain on a Reson VP2000 amplifier with a high pass filter of 1 kHz and a low pass filter of 250 kHz, also recording on a 2 V range).
All devices were at the same depth of 1.4 m, in a 3 m deep pool. Hydrophones were taped to rods, with the tip of the sensor just below (7 cm below) the bottom of the rod, to ensure that they were hanging straight. Rods were attached to wooden blocks, which were clamped onto a ladder which lay atop the calibration tank, across the centre. The three projectors were aligned on one side of the tank, and the 2 receivers (reference and given SoundTrap) were positioned 2 m from the transmitters, on the other side of the tank.
I then computed the clip level of each ST, at each frequency, averaged this for 100 pulses, and compared this to the average of 100 pulses on the reference hydrophone. The frequency response of the reference hydrophone was taken into account. Results show a full scale response that is pretty flat at lower frequencies, but clip levels are ~+5dB around porpoise frequencies.
Disclaimer: some of these devices were brand new, and some had been to depths of >1000 m several times. I do not have a good handle on how much an individual devices' calibration curve changes with time, but it is always best practice to calibrate an individual instrument before and after an experiment/field trip.
This Wiggins & Morris (2019) calibration report may be helpful for checking ST500 tank calibrations at lower frequencies, but their max sample rate was only 96kHz.
