For UK ecological consultants

A season of bat recordings, sorted by species — in hours, not weeks.

BioSonic identifies every UK bat species from your static-detector and transect recordings, then hands you a clean, confidence-ranked shortlist for manual verification. Built for the way UK survey teams actually work — activity surveys, EIA and planning conditions.

Works with full-spectrum WAV from AudioMoth, Wildlife Acoustics (Song Meter, Echo Meter), Elekon Batlogger, Titley Anabat and Pettersson.

The bottleneck

Static survey data buries your team.

A single landscape-scale project can leave your team with hundreds of detector-nights of passive data — and millions of triggered recordings — before a report is even started. Auto-classifiers help, but they throw false positives on the calls that matter most, and Myotis and Plecotus still need an experienced eye.1

Analysts lose days scrubbing noise and re-checking common-pipistrelle bulk before the writing can begin.

100K files
one analyst's season of static data — reviewed in 2 days, not 14
— EcoFauna, BioSonic customer

How it works

A night's recordings, sorted in three steps.

Upload your recordings

Drag in full-spectrum WAV files from any detector. No installs, no local processing limits — it runs in the browser.

AI sorts the night by species

Image recognition on spectrograms classifies every pass, flags feeding buzzes and social calls, and surfaces low-confidence calls so noise and empty files never reach your desk.

Verify and export

Review a ranked, confidence-scored shortlist — keeping manual verification where it counts, in line with BCT good-practice guidance2 — then export client-ready results for your Environmental Statement.

Built for UK survey work

Made for activity surveys, EIA and planning.

Every UK species

All 17 breeding UK species (18 recorded), including the hard Myotis and Plecotus groups where auto-classifiers struggle most.1,4

Verification-first, not black-box

A confidence score on every call keeps you aligned with the 4th-edition Good Practice Guidelines (Collins, ed., 2023), which set the expectation that automated identifications are manually verified.2,3

Benchmarked accuracy

98.7% F1 in an independent third-party benchmark (Wilder Sensing — 6,126 calls, 15 species), ranking first of all classifiers tested, with 13.4× fewer false positives than Kaleidoscope Pro.5 See the benchmarks ›

EIA and planning ready

Outputs structured for activity-survey reporting and the ecology chapters of an Environmental Statement.

Collaborative review

Your whole team works the same project — shared organisation account, no per-seat fees, no licence dongles or project files to pass around. Every change is tracked for a clean audit trail.

Your data, in the UK and EU

Recordings and results stored in UK and EU data centres, encrypted in transit and at rest, with retention long enough to cover your reporting obligations.

FAQ

What ecologists ask first.

Does it replace manual verification?

No — and it shouldn't. BCT good-practice guidance expects automated identifications to be manually verified, and some species (such as whiskered bat, Myotis mystacinus) can't be confirmed from a recording at all.2 BioSonic does the heavy sorting and gives you a confidence-ranked shortlist, so your verification time goes to the calls that actually need judgement.

Which detectors does it work with?

Any full-spectrum WAV recordings — AudioMoth, Wildlife Acoustics Song Meter & Echo Meter, Elekon Batlogger, Titley Anabat and Pettersson. Upload them as recorded; no conversion needed.

Is it accurate enough for planning and EIA work?

It reached 98.7% F1 in an independent third-party benchmark (Wilder Sensing, 6,126 calls across 15 species), ranking first of all classifiers tested.5 Every classification carries a confidence score, and because Myotis and Plecotus are intrinsically hard to separate on call alone, those always go to manual verification — so limitations stay transparent in your reporting.

How is it different from Kaleidoscope Pro?

It runs in the browser with no install, flags feeding buzzes and social calls, and supports collaborative team review. In the same independent benchmark it scored 98.7% F1 to Kaleidoscope Pro's 73.0%, with 13.4× fewer false positives.5 See the full comparison ›

Sources

Every claim, referenced.

We write for an expert audience, so the figures on this page are sourced to the UK authorities and peer-reviewed literature.

  1. Bat Conservation Trust — UK bats: "18 species of bat in the UK, 17 of which are known to be breeding here." The 18th, greater mouse-eared bat (Myotis myotis), is a non-breeding vagrant.
  2. Bat Conservation Trust — Guidelines for passive acoustic surveys of bats in woodland: automated classifications should be manually verified ("ideally at least one recording of each species on each survey date"), and some species, such as Myotis mystacinus, cannot be conclusively verified from a recording.
  3. Collins, J. (ed.) (2023) Bat Surveys for Professional Ecologists: Good Practice Guidelines (4th edn). The Bat Conservation Trust, London. The 4th edition adds dedicated guidance on automated identification (§10.2).
  4. Rydell, J. et al. (2017) Testing the performance of automated identification of bat echolocation calls, Ecological Indicators 78, 416-420 — automated classifiers identified Plecotus auritus 0-89% and Myotis spp. 4-93% of the time; "a request for prudence." See also Walters et al. (2012), J. Appl. Ecol. — even iBatsID correctly identifies only 49-81% of Myotis calls.
  5. Independent third-party benchmark by Wilder Sensing — 6,126 calls across 15 species; BioSonic ranked #1 at 98.7% F1, with 13.4× fewer false positives and 16.3× fewer missed bats than Kaleidoscope Pro (73.0% F1). Full methodology on our benchmarks page.

See it on your own recordings.

Upload a batch of WAVs and watch a night sort itself by species — free, in the browser, no install.

hello@biosonic.se · +46 72 744 65 85