Bat detectors, limitations, and field tips

We explored the science of echolocation, and we met the calls of Jersey’s bat species. Now for the next piece of the puzzle – the technology that makes it all audible.

How does a bat detector actually convert ultrasound into something we can hear? What can acoustic data tell us, and what can’t it? And how do you get the most out of a detector in the field?

How your detectors converts ultrasound

There are several fundamentally different ways a bat detector can make ultrasound audible. Each has different strengths and trade-offs, and understanding which type you’re using helps you interpret what you’re hearing.

Heterodyne detectors

A heterodyne detector works by mixing the incoming ultrasonic signal with a fixed internal frequency set by your tuning dial. What you hear is the beat frequency — the difference between the bat’s call frequency and the detector’s tuned frequency. So if a bat calls at 47 kHz and your detector is tuned to 45 kHz, you hear a tone at 2 kHz. It’s the same principle that produces the “beating” sound when two slightly detuned guitar strings play together.

The beauty of heterodyne is that it works in real time and it exaggerates frequency changes, making Doppler shifts and subtle frequency differences much more obvious to the human ear. The limitation is that you can only listen to a narrow window of about 5 kHz at once, so you might miss species calling outside your tuned range. This is the type most commonly used on JBG bat walks — it’s intuitive, immediate, and hugely rewarding.

Frequency division detectors

Frequency division (FD) detectors convert the incoming signal into a square wave and divide the frequency by a fixed ratio, usually ten. A bat calling at 45 kHz becomes an audible signal at 4.5 kHz. The advantage is that you hear the whole frequency range simultaneously — no tuning required. The trade-off is lower sound quality (converting to a square wave discards a lot of detail) and the fact that calls stay at their original speed, so fast FM sweeps still sound like brief clicks. Some dual-output detectors combine FD and heterodyne, giving you the best of both.

Full-spectrum detectors

Full-spectrum detectors digitise the entire ultrasonic signal at a very high sampling rate — typically 300–500 kHz — preserving all the frequency and amplitude detail. The recording can then be analysed visually as a spectrogram on a computer, or slowed down for listening. Modern full-spectrum detectors like the AudioMoth, Batlogger, and Wildlife Acoustics SM range can record continuously to memory cards for hours or days.

It’s worth noting that full-spectrum detectors evolved from earlier “time expansion” detectors, which recorded short snippets of ultrasound and played them back at reduced speed. Modern full-spectrum units record continuously in real time — no gaps, no missed bats. This is a significant practical advance, and it’s the technology behind the static monitoring that JBG and researchers use across Jersey: small weatherproof units left in the field to record every bat pass at a site.

Zero-crossing analysis (ZCA) detectors

ZCA detectors, such as the Anabat system, record only the points where the signal crosses zero amplitude, producing a highly compressed data stream. This allows very long recording times on minimal storage but discards amplitude information entirely. ZCA recordings can be analysed as time-frequency plots and are still widely used, particularly for long-term automated monitoring.

What detectors can’t tell you

Acoustic monitoring is incredibly powerful, but it has real limitations that are important to understand — especially if you’re involved in survey work or interpreting data.

Detection bias

Because high-frequency sound attenuates rapidly, a detector does not sample the airspace evenly. Loud, low-frequency callers like noctules can be picked up from 50–100+ metres. Quiet, high-frequency callers like long-eared bats or lesser horseshoes might only be detected within 5–10 metres. This means raw bat pass counts are strongly biased towards loud species, and can’t be directly compared across species without correction for detection probability. Humidity, temperature, and wind also affect sound transmission, so detection range varies night to night.

Acoustic cryptic species

As we saw in Part 2, the Myotis genus is notoriously difficult to separate on calls alone. Automated classification software (such as Kaleidoscope, BatClassify, or the Acoustic Pipeline) can assign species-level identifications, but these always carry uncertainty, particularly for Myotis. Manual verification by an experienced analyst is essential for reliable results.

A bat pass is not a bat

A “bat pass” — one sequence of recorded calls — tells you that a bat flew past the microphone. It doesn’t tell you whether it’s the same individual passing back and forth twenty times or twenty different bats. Acoustic data gives you an index of activity, not a count of individuals. This distinction matters hugely for monitoring work.

Call plasticity

The same species can produce quite different calls depending on habitat (open versus cluttered), behaviour (commuting versus foraging), and social context. A common pipistrelle in open air may end its call at 42 kHz; the same species in a confined space might call at 48 kHz. This is why rigid frequency thresholds for identification can be misleading — experienced analysts look at the overall call shape, rhythm, and context, not just the peak frequency number.

And it’s not just bats…

One of the most fascinating discoveries to come out of passive acoustic monitoring is that bat detectors don’t only record bats. Ultrasonic microphones also pick up other animals that vocalise at high frequencies, including insects like bush-crickets, small mammals, amphibians, and birds!

Shrews produce ultrasonic calls that show up clearly in bat detector recordings, and this is something a local PhD student has been studying: whether the vast archives of bat acoustic data collected across the Channel Islands might also hold valuable information about shrew distribution and ecology. It turns out that years of bat monitoring may have been quietly recording more than we realised. But that’s a story for another post …

Getting more from your detector in the field

Whether you’re a complete beginner or a seasoned bat-walker, here are some tips that will help you hear more and understand more of what you’re hearing.

Tune actively

If you’re using a heterodyne detector, don’t just park it on 45 kHz. Start there for common pipistrelles, but sweep regularly between 20 kHz and 55 kHz to check for noctules, serotines, soprano pipistrelles, and Kuhl’s. If you’re near a known horseshoe site, try 80 kHz and 110 kHz. Each frequency window opens a different part of the bat community.

Listen to the rhythm

A fast, accelerating rhythm ending in a buzz means the bat is foraging actively. A steady metronomic pulse means commuting. An irregular, variable pattern in autumn might be a mating display. The temporal pattern of calls often tells you what the bat is doing more clearly than the frequency alone.

Think about habitat

Near water: listen for Daubenton’s bats skimming the surface. Along woodland edges: long-eared bats (though faint). Open grassland and high above treeline: serotines. Around streetlights: Kuhl’s pipistrelle. Habitat tells you what to expect, and what to listen for.

Time your visit

Pipistrelles emerge early, often within minutes of sunset. Horseshoe bats emerge later. Noctules often fly high and early, sometimes even before sunset. If you’re hearing bats before it’s fully dark, they’re most likely pipistrelles or noctules.

Don’t worry about getting every ID

Even the best bat workers can’t identify every call in the field. Some calls are ambiguous, some detections are too brief, and some species simply can’t be separated acoustically. Enjoy the experience of hearing what was previously invisible. The identification comes with time.

The best way to experience all of this first-hand is to join us on a JBG bat walk. 

Become a member today to learn more, or get in touch if you’d like to get involved with acoustic monitoring!