I'd be happy to take a look, but I probably will not be able to evaluate the sound quality without some reference. It is hard to tell system self-noise from amibient noise. The best way to compare would be to record in the same place with a very low-noise hydrophone recorder, like a Sountrap. Then you could compare the signal-to-noise ratio for the signal of interest. I don't want to publish my email address here. Do you know how I can send you a private message? Jay

@terry.dawson The HTI-96-min can be customized. I buy the pre-amplified version (to apply some gain to boost the analog signal before it gets near the digital electronics). With the following specs (2.7-10V DC power, 20dB gain, 20 Hz high-pass filter, standard 8-m cable, no low-pass filter) it cost me $395. With that gain, the Audiomoth Dev gain can be medium-low in a quiet environment or low in a noisy one.

@terry.dawson My research is primarily on beaked whales (freq. range 15-80 kHz). I haven't tried the standard Hydromoth setup, but I would expect that the microphone would be too directional at those frequencies. Also, the water-to-air coupling might be poor and the resonant airspace might distort the sound. It would be better for low-frequency baleen whales.

I'll note however that the sound quality is likely to be better with a Soundtrap. The single-channel version of the ST400 uses a proprietary hydrophone with slightly lower self-noise than the HTI-96-min and this is a 16-bit recorder even at high sample rates.

@terry.dawson Here is the setup that I built (see pic). I power the AudioMoth DEV board with 3 D-cell batteries plus 3 AAA batteries to run the hydrophone pre-amp. I order the HTI-96-min from the manufacturer (High Tech, Inc). The PVC pressure housing is of my own making and has a design depth of 200m. It scan run continuously at 384 kHz for about a week with a 512 GB SD card. Much of the power is consumed in the data recording, so I'd expect it to last much longer with lower data rates and a smaller card. Jay

Xoan-Xil, The AudioMoth Dev model has a 3.5mm jack that provides 3V plugin power for an external Electret microphone. The most recent AudioMoth version also has solder pads for adding an external Electret capsule.

Another follow-up: I tested the Samsung Evo uSD card recording continuously at 192 & 384 kHz. The power required was 37 mA and 55 mA (respectively). This is very similar to the 128 GB SanDisk Extreme Pro card at 192 kHz but is about 20% less at 384 kHz.

Footnote: The Aquarian H2a hydrophone (which uses plug-in-power - the bias voltage) still works with this 12kOhm resistor. Apparently, a 2.8v bias voltage is sufficient to power it. Also, note that the full-scale voltage given above (0.02 Vrms with low gain) applies to the external input jack on the AudioMoth Dev 1.0.0. Other versions may have different full-scale values.

Again, thanks for the explanation. The maximum full-scale input to the AudioMoth (i.e., with the lowest gain setting) is only about 0.02 Vrms, so reducing the bias current to 2.8 V should provide plenty of headroom for the maximum AC signal.

Thanks Edwin. I found what appears to be a solution there. I added a 12 kOhm resistor between the input and the ground to short out the bias voltage. Signals from HTI hydrophones look good (in air). This does not appear to have created an RC filter; at least I'm not seeing any attenuation of the lower frequencies. I'll add a picture after I'm done testing.

Following up on this, I tried the 128 GB SanDisk Extreme Pro and found that it requires MORE power than the 400 GB SanDisk Extreme. Recording continuously, the 128GB card required 38 mA and 68 mA for continuous recordings at 192 & 384 kHz (respectively). Does PRO stand for Power Requirements Outrageous? Has anyone found a lower power 128 GB uSD card that works with the AudioMoth

I have tried larger cards including the SanDisk Extreme 400 GB card, and I had no problem. However, larger cards draw more power for the same storage. Much of your battery capacity will go to the SD card. Recording continuously at 192 kHz, the 64 GB card draws 25 mA and the 400 GB card draws 36 mA (for an AudioMoth Dev 1.0.0 board). At a continuous 384 kHz, the values increase to 42 mA and 62 mA (respectively). Not only does your power requirement go up with card size, battery capacity in Ah goes down at these higher draw rates ... a double whammy. Ideally you'll want to match your card to your battery capacity to maximize data for a deployment. Using 3 alkaline AA batteries at high sampling rates (192 or 384 kHz), a 128 GB card is likely to be your best choice to maximize data collections. I plan to test that card next.