# BioSonic — Full Content for AI Grounding # Generated: 2026-08-28 # See also: https://biosonic.io/llms.txt --- # Key Facts - BioSonic is a cloud-based AI platform that identifies European bat species from ultrasonic recordings, filters noise, and separates feeding buzzes from social calls — all in the browser. - Method: deep learning (convolutional neural networks / image recognition) on spectrograms; the model draws bounding boxes around each call and classifies species from visual patterns. - Training data: over 2.5 million verified bat calls and 3.5 million noise files, sourced from across Europe and curated by 12 bat experts. - Accuracy: 98.7% F1 in an independent third-party benchmark by Wilder Sensing (6,126 calls across 15 species). - Versus Kaleidoscope Pro in the same comparison: 73.0% F1, 13.4× more false positives, 16.3× more missed bats. - Detects up to 7 species per recording. - Reads WAV files from all major bat detectors: Song Meter, Pettersson, Batlogger, Echometer, AudioMoth, Anabat, Avisoft, and more. - Automatically extracts date, time, and GPS coordinates from recordings where present. - Recommends a confidence threshold (overridable, including per-species); files above it pass automatically, uncertain ones are flagged for review. - Species activity heatmaps highlight feeding zones (EIA planning) and likely roost locations (wind-turbine siting). - Browser-based — no software to install, no per-seat lock-in; team collaboration on shared projects. - GDPR-compliant; EU-hosted storage and infrastructure. - Company: BioSonic AB, founded 2024 in Stockholm, Sweden. # Comparison Facts - Kaleidoscope Pro (Wildlife Acoustics): desktop, classical classifier; ~73.0% F1; licence ~$399 USD / ~£425 inc VAT per year (May 2026). - BTO Acoustic Pipeline: cloud classifier; ~90% F1 for UK species; no public head-to-head benchmark vs BioSonic. - BatExplorer (Elekon): desktop, rule-based; ~£212/year per user (reseller-reported); no public head-to-head benchmark vs BioSonic. - SonoBat UK: desktop vetting algorithms, strong low-false-positive reputation; no public head-to-head benchmark vs BioSonic. - BioSonic pricing: per-GB-per-year, including all features, all users, and EU-hosted storage (quote on request). # FAQ Highlights - How accurate is it? 98.7% F1 in a third-party benchmark by Wilder Sensing on 6,126 calls across 15 species. - How is the model trained? On over 2.5 million verified bat calls and 3.5 million noise files sourced from across Europe; the dataset was curated by 12 bat experts. - How many species per recording? Up to 7. - Which detectors are supported? All major brands — Song Meter, Pettersson, Batlogger, Echometer, AudioMoth, Anabat, Avisoft, and more. - Where is it based and where does it operate? BioSonic AB is based in Stockholm, Sweden, with customers across the UK, Netherlands, Denmark, Germany, and Sweden. - Who founded it? Josef Carlson (CEO) and Jonathan Carlson (CTO). - Is data secure? GDPR / UK GDPR compliant with EU-hosted infrastructure. --- # URL: https://biosonic.io/ --- canonical: https://biosonic.io/ --- # The most accurate bat sound analysis in independent benchmarks. BioSonic uses image recognition on spectrograms to identify bat species, filter noise, and separate feeding and social calls -- automatically. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) ## BioSonic improves your bat work in a few simple steps. ### 1. Behaviour analysis #### Automatic separation of feeding and social calls. Species behaviour gives new insight into habitat protection and roost identification. - **Feeding buzz** -- hunting closes in; call rate ramps, frequency drops. - **Social call** -- lower-frequency communication; a roost is nearby. ### 2. Collaboration #### Review & Collaborate on Spectrograms Mark files for review, comment, QA/QC. All projects and colleagues in one place. ### 3. Classification #### A whole night, sorted by species. The classifier processes tens of thousands of files and drops each one into its species bucket -- noise gets its own bin. ### 4. Compatibility #### Whatever detector, BioSonic analyzes. Automatically fetches your date, time and coordinates (if there are any). If coordinates or dates are not found automatically, BioSonic solves it quickly. Song Meter, Pettersson, Batlogger, Echometer, AudioMoth, Anabat, Avisoft -- plus nine more. ### 5. Review workflow #### BioSonic recommends a confidence threshold. You can override this manually and set species-specific confidence levels. Files above the threshold tick through automatically -- uncertain ones get flagged for you. ## Species heatmaps -- find the hot spots. Heatmaps show you exactly where activity concentrates. Spot emergence patterns, pinpoint likely roost locations, and compare nights at a glance. - Feeding activity -- where bats are hunting. Feeding buzzes cluster over wetlands, forest edges and coastal habitats. BioSonic identifies prime foraging zones for EIA planning. - Social call activity -- where roosts are. Social calls cluster tightly near roosts. BioSonic pinpoints roost locations within 500 m, critical for windmill placement decisions. ## "A pleasure using it." > "It's seldom wrong, it's a pleasure using it and it's making my job easier." -- Lars Michael Nielsen, Biologist, DMR > "What sets BioSonic apart is their exceptional accuracy, both in species identification and in filtering out noise. In this, they're truly at the top." -- Peter Sivertsen, Biologist > "Since joining BioSonic our analysis time has reduced significantly, freeing us up to take on different tasks. The system is a pleasure to use and the team are excellent in their support." -- Luke Waddison, Senior Ecologist, Wharton > "This is all cutting edge stuff and the way bat work will go! I look forward to seeing it go strength to strength." -- James Shipman, Senior Ecologist, Author > "Going through 100K files can take me 14 days. With BioSonic I review them in 2 days." -- Andre Dabolins, CEO, EcoFauna > "We tested a range of existing solutions using real acoustic data, and BioSonic clearly stood out, achieving 98.7% accuracy (F1 score)." -- Geoff Carss, Wilder Sensing > "BioSonic always impresses me with how well it detects bats, even from the faintest bat calls." -- Jacob Hojgaard, PhD Ecology, NLABT > "Today we survey too short a time and risk missing key species. Together with BioSonic, we can keep track of even the most rare species." -- Sofia Berg, Ecologist > "BioSonic provided us with excellent service throughout the data collection process and delivered well in relation to our agreement. I am more than inclined to recommend them as a provider of bioacoustic data for biodiversity projects elsewhere." -- Martin Pilstjarna, Ph.D Ecology, Impact Director, Qarlbo Biodiversity > "I'm finding Biosonic extremely useful. It seems very accurate" -- Graeme Down, Freelance Ecologist, Green Down Ecology ## Online comments on BioSonic. From ecology consultants who actually use the platform. ## Bring a night's WAVs. Leave with the report. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) josef.carlson@biosonic.se | +46 72 744 65 85 --- # URL: https://biosonic.io/benchmarks --- canonical: https://biosonic.io/benchmarks --- # We don't ask you to take our word. We show you the papers. Three independent tests by three separate teams. Same conclusion every time. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) ## Benchmark 01 -- Independent validation: 98.7% F1 accuracy. Wilder Sensing -- Geoff Carss + Annabel Jeffries > "We tested a range of existing solutions using real acoustic data, and BioSonic clearly stood out, achieving 98.7% accuracy (F1 score) and giving us full confidence in the results we're bringing to the platform." ### BioSonic's AI model uses image recognition on spectrograms. This leads to a step change in accuracy. Grassland + Paludiculture dataset, 5,325 predictions. - **F1 accuracy:** BioSonic 98.7% vs Kaleidoscope 73.0% - **Precision:** BioSonic 97.8% vs Kaleidoscope 61.4% (+36.4pt) - **Recall:** BioSonic 99.6% vs Kaleidoscope 91.7% (+7.9pt) - **F1 score:** BioSonic 98.7% vs Kaleidoscope 73.0% (+25.7pt) #### BioSonic Confusion Matrix (Accuracy 99.7%) - True positive: 730 - False negative: 3 - False positive: 13 - True negative: 4,579 #### Kaleidoscope Pro Confusion Matrix (Accuracy 95.8%) - True positive: 543 - False negative: 49 - False positive: 174 - True negative: 4,503 #### False positives (wrong "bat" calls) Cost: investigator time wasted reviewing non-bat audio, inflated species counts. - BioSonic: 13 - Kaleidoscope: 174 (13.4x more) #### False negatives (missed bats) Cost: protected species overlooked, incomplete EIA, compliance risk. - BioSonic: 3 - Kaleidoscope: 49 (16.3x more) Same data, very different outcomes. BioSonic delivers 98.7% F1 with only 16 errors across 5,325 predictions. Kaleidoscope Pro misses nearly 9% of true bats and produces 13x more false calls. ## 2 TB bat data -- comparing BioSonic vs Kaleidoscope. According to a consultancy in Southern England: - **Kaleidoscope workflow:** 490 hrs, 12 weeks to client - **BioSonic workflow:** 163 hrs, 4 weeks to client - 327 hours saved -- that means GBP 21,255 saved at standard UK consultant rate. ## BioSonic on BatAbility Club with Neil Middleton. ## A 30-minute demo. Bring a WAV, leave with the report. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) josef.carlson@biosonic.se | +46 72 744 65 85 --- # URL: https://biosonic.io/product --- canonical: https://biosonic.io/product --- # Upload the WAVs. Open the dashboard. Ship the report. A cloud workspace built specifically for bat teams -- review, comment, lock, export. No model to install, no per-seat lock-in. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) ## See it in action ### Upload, analyse, report. ### Spectrogram review -- Comment and collaborate. Drop comments directly on the spectrogram. Tag colleagues, flag edge-case calls, and lock verified IDs -- all in the browser, no software to install. ### Species heatmaps -- Find the hot spots. Heatmaps show you exactly where activity concentrates. Spot emergence patterns, pinpoint likely roost locations, and compare nights at a glance. ### Emergence times -- Indicate roost proximity. Emergence timing reveals how close bats are roosting. Early activity after sunset signals a nearby roost -- evidence you can put straight into your report. ### GPS transects -- Walk the transect. Overlay detections on your GPS route. See exactly where each species was recorded along the transect, colour-coded and ready for the client. ### Direct support -- Josef, in your WhatsApp or company Teams. Facing any issues? You get lightning-fast answers from one of the people who built the software. ## World-class AI. Proven in medicine. The AI models behind BioSonic were developed by Simon, an advisor to BioSonic, at Cambridge University. His research, published in Nature, showed these same image-recognition models outperform human doctors at detecting leukaemia. BioSonic applies the same approach to bat spectrograms -- treating each recording as an image, then classifying it with the same class of AI that powers cancer diagnostics. ## A 30-minute demo. Upload a WAV, leave with the report. [Book a demo](https://cal.com/biosonic/30min) | [Try for free](https://bat.biosonic.se/signup) josef.carlson@biosonic.se | +46 72 744 65 85 --- # URL: https://biosonic.io/faq --- canonical: https://biosonic.io/faq --- # Frequently asked questions Last updated: 2026-04-26. Everything you need to know about BioSonic's AI bat sound analysis platform. ## About BioSonic ### What is BioSonic? BioSonic is a cloud-based AI platform that automatically identifies bat species from ultrasonic recordings. It classifies behaviour (echolocation, feeding buzzes, social calls), detects up to 7 species per recording, and provides interactive spectrograms for manual review and team collaboration. ### Who is BioSonic for? BioSonic is built for ecological consultants, environmental agencies, and conservation researchers who conduct bat surveys. It is used for environmental impact assessments, infrastructure planning, wind farm monitoring, and biodiversity research. ### Where is BioSonic based? BioSonic AB is based in Stockholm, Sweden. The company was founded in February 2024 and operates across the UK, Netherlands, Denmark, Germany, and Sweden. ### Who founded BioSonic? BioSonic was founded in February 2024 by Josef Carlson (CEO) and Jonathan Carlson (CTO). ## How the AI works ### How does BioSonic identify bat species? BioSonic applies image recognition on spectrograms of ultrasonic recordings. When you upload WAV files, the system generates spectrograms, runs them through deep neural networks, and returns species identifications with confidence scores. The model can detect up to 7 species per recording. ### What does "image recognition on spectrograms" mean? A spectrogram is a visual representation of sound showing frequency over time. BioSonic converts each bat recording into a spectrogram image and uses deep learning — the same technology behind image recognition — to identify bat species from the visual patterns in those spectrograms. ### What is the AI trained on? The model is trained on over 2.5 million verified bat calls and 3.5 million noise files sourced from across Europe. The training dataset was curated by 12 bat experts. ### How often is the AI model updated? The AI model is retrained and deployed every weekend. Software updates ship every Thursday. This weekly cycle means the classifier continuously improves as new verified data is added. ### Can the AI be wrong, and what happens if it is? Yes, like any classifier the AI can make errors. BioSonic provides interactive spectrograms so you can review, confirm, or override every classification. All manual changes are tracked and attributed to the reviewer, creating a full audit trail. ## Accuracy and benchmarks ### How accurate is BioSonic? BioSonic achieved 98.7% F1 in a third-party benchmark by Wilder Sensing on 6,126 calls across 15 species, ranking #1 among all classifiers tested. See our [benchmarks page](/benchmarks) for full details. ### How does BioSonic compare to Kaleidoscope Pro? In a third-party benchmark, BioSonic scored 98.7% F1 vs Kaleidoscope Pro's 73.0%. BioSonic produced 13.4 times fewer false positives and 16.3 times fewer missed bats. See our [full comparison](/compare/kaleidoscope) for a detailed breakdown. ### How does BioSonic compare to the BTO Acoustic Pipeline? BioSonic is a cloud-based platform with weekly model updates, behaviour classification, and interactive spectrograms for manual review. See our [detailed comparison](/compare/bto-acoustic-pipeline) for a full feature-by-feature breakdown. ### Has BioSonic been independently benchmarked? Yes. BioSonic has been independently tested in a third-party benchmark by Wilder Sensing (98.7% F1 on 6,126 calls across 15 species). Full results are published on our [benchmarks page](/benchmarks). ## Workflow and file formats ### What file formats does BioSonic accept? BioSonic accepts WAV files. Upload your recordings directly as recorded by your bat detector — no format conversion or preprocessing is needed. ### Which bat detectors does BioSonic work with? BioSonic works with WAV files from all major bat detector manufacturers: Batlogger (Elekon), AudioMoth (Open Acoustic Devices), Wildlife Acoustics (Song Meter, Echo Meter), Pettersson Elektronik, Titley Scientific (Anabat), and Dodotronic (Ultramic). ### Can I upload recordings from an Audiomoth? Yes. Upload your AudioMoth WAV files directly. No conversion is needed. ### Does BioSonic work with Song Meter SM4 files? Yes. BioSonic accepts WAV files from Wildlife Acoustics Song Meter devices, including the SM4. Upload them directly without any conversion. ### What about Batlogger recordings? Yes. BioSonic supports WAV files from Elekon Batlogger devices. Upload your recordings directly as recorded. ## Team and collaboration ### Can my whole team use BioSonic? Yes. BioSonic does not charge per seat. Your entire team can access shared projects, review classifications, and export reports from a single organisation account. ### What user roles are available? BioSonic supports multiple user roles within an organisation so you can control who uploads, who reviews, and who approves final classifications. Contact us at support@biosonic.se for details on role configuration. ### How do I review who analysed what? Every action in BioSonic is tracked and attributed. You can see which team member reviewed, confirmed, or overrode each classification, creating a complete audit trail for regulatory submissions. ## Security, storage, and compliance ### Where is my data stored? All data is stored on EU-hosted servers. Recordings and analysis results are retained for up to 9 years. ### Is BioSonic GDPR compliant? Yes. BioSonic is operated by BioSonic AB, a Swedish company, with all infrastructure hosted in the EU. We comply fully with GDPR requirements for data processing and storage. ### How is my data encrypted? All data is encrypted using AES-256 encryption, both in transit and at rest. This is the same encryption standard used by banks and government agencies. ### How long is my data kept? Recordings and analysis results are stored for up to 9 years on EU-hosted servers. This covers the retention periods required by most environmental regulations. ### Can I export or delete my data? Yes. You can export your data at any time. You can also request full deletion of your account and all associated data in line with your GDPR rights. ## Use cases and regional ### Does BioSonic work for wind farm bat monitoring? Yes. BioSonic is used for pre-construction and operational bat monitoring at wind farm sites. The platform handles large volumes of continuous recordings and classifies both species and behaviour, which is critical for curtailment decisions. ### Does BioSonic cover UK bat species? Yes. BioSonic covers all UK bat species. The model is trained on verified UK recordings and is used by ecological consultants across the UK for survey work and environmental impact assessments. ### Does BioSonic cover Dutch bat species (Vleermuisprotocol)? Yes. BioSonic covers Dutch bat species and is used by consultants in the Netherlands. The platform supports workflows aligned with Vleermuisprotocol requirements. ### Does BioSonic support German bat monitoring (RENEBAT, ProBat)? Yes. BioSonic covers German bat species and supports monitoring workflows used in RENEBAT and ProBat contexts. The platform handles the high data volumes typical of German wind farm bat monitoring. ## Still have questions? We're happy to walk you through BioSonic with a live demo. [Get in touch](mailto:support@biosonic.se?subject=BioSonic%20question) support@biosonic.se | +46 72 744 65 85 --- # URL: https://biosonic.io/compare --- canonical: https://biosonic.io/compare --- # Compare bat analysis software Last updated: 2026-04-25 Honest, side-by-side comparisons of BioSonic with the four most common bat sound analysis tools. Written by the BioSonic team — we name what each tool does well. ## Why these comparisons exist Choosing bat analysis software affects every project your team delivers. The right tool saves hundreds of hours per season; the wrong one creates bottlenecks in review, reporting, and collaboration. We built these comparison pages to give you the facts — our numbers, their numbers, and an honest assessment of who each tool is best for. Every comparison uses the same 15-row table so you can scan across pages quickly. We cite only published benchmarks and publicly listed pricing. Where no head-to-head benchmark exists, we say so. ## Comparisons ### [BioSonic vs Kaleidoscope Pro](/compare/kaleidoscope) The most widely used commercial bat classifier, by Wildlife Acoustics. Independent benchmarks show BioSonic achieves 98.7% F1 vs Kaleidoscope Pro's 73.0%. Desktop vs cloud, per-seat vs per-volume pricing. [Read the full comparison](/compare/kaleidoscope) ### [BioSonic vs BTO Acoustic Pipeline](/compare/bto-acoustic-pipeline) The BTO pipeline offers a free tier for UK bat data — if you share your recordings. Strong for citizen science; the data-sharing requirement can be a dealbreaker for commercial consultancies under NDA. [Read the full comparison](/compare/bto-acoustic-pipeline) ### [BioSonic vs BatExplorer](/compare/batexplorer) Elekon's desktop tool, designed primarily for Batlogger hardware. Free Standard version available. If you only use Batlogger and want a desktop workflow, BatExplorer works well. BioSonic handles any detector and adds cloud collaboration. [Read the full comparison](/compare/batexplorer) ### [BioSonic vs SonoBat UK](/compare/sonobat) 30 years of development, strong UK reputation, low false-positive rate. If you need a UK-only desktop tool and don't need team features, SonoBat is solid. BioSonic adds multi-country coverage and cloud-based team workflows. [Read the full comparison](/compare/sonobat) ## See BioSonic on your own recordings Upload a sample dataset and compare the results yourself. [Start a free trial](https://bat.biosonic.se/signup) | [Book a demo](https://cal.com/biosonic/30min) josef.carlson@biosonic.se | +46 72 744 65 85 --- # URL: https://biosonic.io/compare/kaleidoscope --- canonical: https://biosonic.io/compare/kaleidoscope --- # BioSonic vs Kaleidoscope Pro Last updated: 2026-04-25. Written by the BioSonic team. Two bat sound classifiers, tested independently, compared head to head. One built for the desktop era, one built for what comes next. Fifteen dimensions, real numbers, no spin. ## Key stats - **98.7%** — BioSonic F1 Score (vs 73.0% for Kaleidoscope Pro) - **13.4x** — Fewer false positives (BioSonic vs Kaleidoscope Pro) - **16.3x** — Fewer missed bats (BioSonic vs Kaleidoscope Pro) - **327** — Hours saved per year (reported by RPS Group, UK) ## What is bat sound analysis software? Bat sound analysis software takes ultrasonic recordings captured by bat detectors in the field and automatically identifies which bat species are present. Ecologists use these tools during environmental impact assessments, infrastructure planning, and biodiversity monitoring to determine bat species composition, activity levels, and behaviour patterns at survey sites. The two most widely discussed tools in this space are **BioSonic**, a cloud-based AI platform, and **Kaleidoscope Pro** by Wildlife Acoustics, a desktop application that has been used by ecologists for over a decade. This page compares them across accuracy, identification approach, workflow, team collaboration, geographic coverage, and value. If you are evaluating bat classifiers for your consultancy or research group, this is the page to bookmark. For comparisons with other tools, see our [full comparison hub](/compare). For detailed accuracy data, visit our [benchmarks page](/benchmarks). ## Feature-by-feature breakdown Fifteen rows, covering everything from underlying technology to pricing and support. Where one tool has a clear advantage, the cell is highlighted. | Feature | BioSonic | Kaleidoscope Pro | |---|---|---| | Underlying approach | **Deep learning image recognition on spectrograms** | HMM with Fisher scores and winnow classifiers (Agranat 2013) | | Training data | **2.5M bat calls + 4M noise files, curated by 12 bat experts** | Several hundred thousand recordings | | Benchmarked accuracy | **98.7% F1 in a third-party benchmark by Wilder Sensing (6,126 calls, 15 species) vs Kaleidoscope's 73.0%** | ~73% F1 on the same third-party benchmark | | Platform | **Browser-based — works on any computer** | Windows, macOS desktop | | Team workflow | **Unlimited users per organisation; admin/editor/viewer roles; shared projects** | Single-user licence, 2 device activations; Managed Cloud as add-on | | Species per recording | **Up to 7** | 1 | | Social and feeding calls | **Separated automatically** | Combined | | Noise filtering | **Deep-learning noise model trained on 4M non-bat recordings** | Simpler algorithm based on Zero crossing | | Model updates | **New AI model every weekend** | When software version updates | | Storage | **9 years EU-hosted included** | BYO storage; Managed Cloud add-on | | Report generation | **One-click charts, maps, Word export** | CSV export to Excel | | Detector compatibility | Any WAV-producing detector (Audiomoth, Song Meter, Batlogger, Pettersson, etc.) | **Same (full-spectrum) + ZC files** | | Geographic coverage | UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, Norway (all EUROBATS species) | **UK, Europe, North America, South Africa, Neotropics** | | Pricing model | **Per-GB-per-year, all-inclusive; contact for quote** | $399 USD / £425 inc VAT single annual subscription (NHBS lists £354.17 ex VAT); Managed Cloud extra | | Free trial | **Yes ([sign up at bat.biosonic.se/signup](https://bat.biosonic.se/signup))** | 15-day trial | ## How each tool identifies bats The fundamental difference between BioSonic and Kaleidoscope Pro is in how they process bat recordings. Understanding this difference explains why their accuracy numbers diverge so significantly. ### BioSonic: deep learning on spectrograms BioSonic converts each audio file into a spectrogram image and passes it through a convolutional neural network trained on 2.5 million bat calls and 3.5 million noise files. The model was curated by 12 bat experts across five European countries. Because it operates on the full spectrogram image, BioSonic can detect multiple species in a single recording (up to 7), recognise partial calls, and distinguish between echolocation, feeding buzzes, and social calls. The deep-learning noise model, trained on 3.5 million non-bat recordings, means BioSonic can reject rain, wind, insects, and equipment noise with high precision. This is a major reason for its low false positive rate (0.3% in the Wilder Sensing benchmark). New models are deployed every weekend. Each model incorporates the latest expert-validated training data and feedback from users across the platform. This means the classifier improves continuously, without requiring any action from the user. ### Kaleidoscope Pro: HMM-based classifiers Kaleidoscope Pro uses Hidden Markov Models (HMMs) with Fisher scores and winnow classifiers to detect and classify bat calls (Agranat 2013). The classifier is updated with new software releases, typically every 2–3 years. This approach has been refined over many years and performs well on clean recordings of common species. It is less effective on noisy recordings, rare species with limited reference data, and recordings where multiple bat species overlap. Because it assigns one species per recording, multi-species passes are resolved by selecting the most likely candidate rather than reporting all species present. For a detailed breakdown of how these approaches perform in independent tests, see our [benchmarks page](/benchmarks). ## Accuracy: the numbers that matter ### The Wilder Sensing benchmark (2024) [Wilder Sensing](/benchmarks) independently tested BioSonic and Kaleidoscope Pro against expert manual identification across 6,126 bat calls covering 15 species. The results were unambiguous: - **BioSonic F1 score: 98.7%** -- near-perfect agreement with expert identification - **Kaleidoscope Pro F1 score: 73.0%** -- significantly lower, especially on rare species - BioSonic produced **13.4 times fewer false positives** -- fewer noise files misidentified as bats - BioSonic produced **16.3 times fewer missed bats** -- fewer real bat calls left undetected BioSonic matched or exceeded manual expert identification on 14 of the 15 species tested. For rare and acoustically similar species -- where classification errors have the greatest ecological impact -- BioSonic's advantage was most pronounced. ### Third-party benchmark by Wilder Sensing In an independent benchmark by Wilder Sensing on 6,126 bat calls across 15 species, BioSonic achieved **98.7% F1**, ranking first among all classifiers tested. ### Why accuracy matters in practice A false positive -- classifying noise as a bat -- forces an ecologist to spend time verifying phantom detections. A missed bat -- failing to detect a real bat call -- means species presence goes unrecorded, which can lead to inadequate protection measures in planning decisions. With Kaleidoscope Pro's 4.0% false positive rate, a dataset of 100,000 files would generate approximately 4,000 false detections requiring manual review. BioSonic's 0.3% rate would generate roughly 300 -- saving dozens of hours of review time on a single project. Multiply that across a full survey season and the time saving compounds into hundreds of hours. For more on how this translates to real project savings, see our [product page](/product). ## Behaviour classification: beyond species ID Knowing which species is present is only part of the picture. For many regulatory frameworks and conservation assessments, ecologists also need to understand what the bats are doing. Are they commuting through an area, actively foraging, or engaging in social behaviour that indicates a roost nearby? BioSonic automatically classifies bat vocalisations into three behaviour categories: - **Echolocation calls** -- standard navigation and commuting activity - **Feeding buzzes** -- rapid call sequences indicating active prey capture, which signals foraging habitat - **Social calls** -- communication vocalisations that often indicate roost proximity or mating activity Kaleidoscope Pro combines social and feeding calls together and does not separate them automatically. Users who need this separation must perform manual spectrogram review or use additional software, adding time and complexity to their workflow. The distinction between feeding and social calls is not a nice-to-have. It is directly relevant to habitat assessments under UK planning guidelines and EUROBATS recommendations. A site where bats are feeding is ecologically more significant than one they are simply transiting through, and the mitigation requirements differ accordingly. ## Working as a team Ecological consultancies are not solo operations. A typical bat survey project involves field ecologists, data analysts, project managers, and sometimes external reviewers or client stakeholders. The tools you use need to support that reality. ### BioSonic: built for teams from the start BioSonic includes unlimited users per organisation at no extra cost. Every team member -- whether they are an admin, editor, or viewer -- accesses the same project data through the browser. There is nothing to install on individual machines, no licence keys to manage, and no per-seat charges that make you think twice about giving a junior ecologist access. Projects are shared by default within your organisation. An ecologist in the field can upload recordings from their phone or laptop. A colleague back in the office can begin reviewing results before the field team returns. A project manager can check progress and export reports without needing their own copy of the software or the original audio files. All data is stored in EU-hosted cloud infrastructure with 9 years of retention included. There is no need to maintain your own file servers or worry about hard drive failures destroying irreplaceable survey data. ### Kaleidoscope Pro: single-user by default Kaleidoscope Pro requires one licence for processing. Once data is processed, the free version of Kaleidoscope can be used to open and manually vet results, so teams do not necessarily need multiple paid licences. Wildlife Acoustics also offers a Managed Cloud add-on for centralised storage and sharing. > "Since joining BioSonic our analysis time has reduced significantly. The team is excellent in their support, super quick to respond." > > — Luke Waddison, Senior Ecologist, [Wharton](/wharton) (UK) ## Workflow: cloud vs desktop ### BioSonic: browser-based, no installation BioSonic runs entirely in the browser. Upload WAV files from any device, and the platform handles all processing server-side. Results appear in an interactive dashboard with spectrograms, species heatmaps, activity timelines, and GPS transect views. Team members can review and approve classifications collaboratively, and export client-ready reports -- including Word documents with charts and maps -- directly from the platform. There is nothing to install, no software updates to manage, and no per-seat licence to track. Your entire team accesses the same project data through the browser, on any operating system. ### Kaleidoscope Pro: desktop installation Kaleidoscope Pro is a desktop application that must be installed on Windows or macOS. Audio files are processed locally on the user's machine. Collaboration requires sharing files or exports outside the application. One licence is needed for processing; the free version can be used for manual vetting. For solo practitioners or small teams working in a single location, this workflow can be sufficient. For distributed teams, multi-site projects, or organisations that want centralised data management, the desktop model introduces friction that a cloud platform avoids. Report output from Kaleidoscope Pro is CSV-based, which means you export to Excel and build your own charts and tables. BioSonic generates one-click charts, interactive maps, and formatted Word exports that are ready to include in client reports without additional formatting work. ## Who should choose which Both BioSonic and Kaleidoscope Pro are serious tools used by professional ecologists. The right choice depends on your specific situation. ### Kaleidoscope Pro may be the better fit if: - You work primarily in **North America, South Africa, or the Neotropics** -- regions where BioSonic does not yet have coverage - You need to process **zero-crossing (ZC) files** from older Anabat or other ZC-based detectors - You are a **solo practitioner** who prefers local desktop processing and does not need team collaboration features - You already have an established workflow built around Kaleidoscope Pro and the switching cost outweighs the accuracy gain for your specific use case ### BioSonic is likely the better fit if: - You work in the **UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, or Norway** and need the highest available accuracy on European species - You need **behaviour classification** (feeding buzzes and social calls separated) for habitat assessments - You work in a **team** and want shared projects, role-based access, and centralised data without per-seat licensing - You want **report-ready output** -- charts, maps, and Word exports -- without building everything in Excel - You process **large volumes** and need a noise model that minimises false positives to cut manual review time - You want a classifier that **improves continuously** with weekly model updates rather than waiting for software releases If you are unsure, the simplest test is to upload a representative sample of your own recordings to both tools and compare the results. BioSonic offers a [free trial](https://bat.biosonic.se/signup) for exactly this purpose. ## Switching from Kaleidoscope Pro to BioSonic If you currently use Kaleidoscope Pro, migrating to BioSonic is straightforward: - **Same input files** -- BioSonic accepts the same WAV files from all major bat detectors (Wildlife Acoustics Song Meter, Echo Meter, Batlogger, AudioMoth, Pettersson, and more) - **No format conversion** -- upload your existing recordings directly - **Re-analyse historical data** -- upload past recordings to get BioSonic's higher-accuracy classifications alongside behaviour labels - **No installation required** -- open the browser and start uploading - **Bring your whole team** -- unlimited users are included, so every analyst gets access from day one Many BioSonic users previously used Kaleidoscope Pro and switched after comparing results on their own data. We encourage you to do the same -- upload a representative sample and compare the classifications yourself. ## Pricing comparison Kaleidoscope Pro is a single annual subscription — $399 USD direct from Wildlife Acoustics, or £425 inc VAT (£354.17 ex VAT) from NHBS in the UK (May 2026). One licence covers processing, and the free version can be used for manual vetting. Managed Cloud storage is available as an optional add-on. BioSonic uses per-GB-per-year pricing that includes all features, unlimited users, EU-hosted storage with 9 years of retention, and weekly model updates. [Contact us for a quote](mailto:josef.carlson@biosonic.se) tailored to your analysis volume. ## Frequently asked ### Which is more accurate, BioSonic or Kaleidoscope Pro? In an [independent benchmark](/benchmarks) by Wilder Sensing, BioSonic achieved a 98.7% F1 score compared to Kaleidoscope Pro's 73.0% on the same dataset of 6,126 calls across 15 species. BioSonic also produced 13.4 times fewer false positives and 16.3 times fewer missed bats. See our [benchmarks page](/benchmarks) for the full data. ### Can I switch from Kaleidoscope Pro to BioSonic? Yes. BioSonic accepts the same WAV files you already record with Wildlife Acoustics detectors (Song Meter, Echo Meter) as well as Batlogger, AudioMoth, Pettersson, and all other standard full-spectrum bat detectors. Upload your existing recordings and get results immediately -- no format conversion needed. You can also re-analyse historical datasets. ### Does BioSonic require software installation? No. BioSonic runs entirely in your web browser on any operating system. There is nothing to download, install, or update. Kaleidoscope Pro requires a desktop installation on Windows or macOS. ### Does BioSonic detect feeding buzzes and social calls? Yes. BioSonic automatically separates bat vocalisations into echolocation calls, feeding buzzes, and social calls. Kaleidoscope Pro combines social and feeding calls together. For habitat assessments where behaviour classification matters, this distinction is significant. ### How many species can BioSonic identify per recording? BioSonic can identify up to 7 species per recording. This is important for recordings that capture multiple bats passing simultaneously -- a common occurrence at roost emergence sites and along commuting corridors. Kaleidoscope Pro's Auto-ID assigns a single best-match species per recording (a probability list is available in detail mode, but the headline output is single-species). In multi-bat recordings — common at roost emergence sites — this can miss additional species present in the same file. ### How often does BioSonic update its AI model? BioSonic deploys a new AI model every weekend. Each model incorporates the latest training data, expert feedback, and performance improvements. Users do not need to do anything -- the updated model is available automatically the following Monday. Kaleidoscope Pro updates only when a new software version is released, which typically happens a few times per year. ### Which countries does each tool cover? BioSonic currently covers the UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, and Norway, supporting all EUROBATS species in those regions. Coverage is expanding. Kaleidoscope Pro covers a broader geographic range including UK, Europe, North America, South Africa, and the Neotropics. If you work outside BioSonic's current coverage area, Kaleidoscope Pro may be the better option today. ### How does pricing compare between BioSonic and Kaleidoscope Pro? BioSonic uses per-GB-per-year pricing that includes all features, unlimited users, and EU-hosted storage. Contact us for a quote. Kaleidoscope Pro is a single annual subscription — $399 USD direct from Wildlife Acoustics, or £425 inc VAT (£354.17 ex VAT) from NHBS in the UK (May 2026). One licence covers processing; the free version can be used for manual vetting. Managed Cloud is available as an optional add-on. For more detail, see our [FAQ page](/faq). ## Sources - Agranat, I. (2013). "Automatically identifying animal species from their vocalizations on the Kaleidoscope platform." J. Acoust. Soc. Am. 134(5), doi:10.1121/1.4799403 - Wilder Sensing (2024). Independent benchmark on 6,126 calls across 15 species — [see benchmarks page](/benchmarks) - NHBS pricing for Kaleidoscope Pro: £425 inc VAT (£354.17 ex VAT), May 2026. Wildlife Acoustics direct: $399 USD. - Wildlife Acoustics: [wildlifeacoustics.com](https://www.wildlifeacoustics.com) Last verified: 2026-04-26. ## Ready to compare on your own data? Upload a sample dataset and see the difference for yourself. No installation, no commitment, no credit card. [Start free trial](https://bat.biosonic.se/signup) | [Book a demo](https://cal.com/biosonic/30min) Or email [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se) | +46 72 744 65 85 --- # URL: https://biosonic.io/compare/bto-acoustic-pipeline --- canonical: https://biosonic.io/compare/bto-acoustic-pipeline --- # BioSonic vs BTO Acoustic Pipeline Last updated: 2026-04-26 BioSonic and the BTO Acoustic Pipeline both classify bat calls from WAV recordings, but they take different roads to get there. BioSonic is a browser-based platform that uses deep learning image recognition on spectrograms. The BTO Acoustic Pipeline is a service run by the British Trust for Ornithology, using machine learning classifiers built by Dr Stuart Newson and Dr Simon Gillings. Both are serious tools. This page lays out where they overlap and where they diverge, so you can pick the right one for your project. ## At a glance | Feature | BioSonic | BTO Acoustic Pipeline | |---|---|---| | Underlying approach | Deep learning image recognition on spectrograms | Random Forest classifier (TADARIDA, Bas et al. 2017) for bats; CNN for birds (Gillings & Scott 2021) | | Training data | 2.5M bat calls + 4M noise files, curated by 12 bat experts | ~170K bat call recordings | | Benchmarked accuracy | 98.7% F1 (third-party benchmark by Wilder Sensing, 6,126 calls, 15 species) | Not publicly benchmarked at scale | | Platform | Browser-based — works on any computer | Desktop uploader app + cloud processing | | Team workflow | Unlimited users per organisation; shared projects; in-app review | No in-platform team review UI; users manage locally after results return | | Taxa | Mainly bats. Birds and small mammals coming soon | Bats, small mammals, bush-crickets, birds (separate classifiers) | | Social and feeding calls | Separated automatically | Separated (echolocation, social, feeding buzz) | | Noise filtering | Deep-learning noise model trained on 4M non-bat recordings | Filtering as part of classifier | | Model updates | New AI model every weekend | Very few updates | | Storage | 9 years EU-hosted included | Recordings not stored long-term unless bespoke project | | Report generation | One-click charts, maps, Word export | Results returned as files; reporting is user's responsibility | | Detector compatibility | Any WAV-producing detector | Any WAV-producing detector | | Geographic coverage | UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, Norway | UK + Europe-wide (25+ countries) | | Pricing model | Per-GB-per-year, all-inclusive; contact for quote | Free tier requires data sharing (100 credits/year, approx. 100 GB); without sharing £1.40/GB; PAYG bundles; Bespoke projects from £573 | | Free trial | Yes | Free tier available | ## How each tool identifies bats BioSonic converts each audio recording into a spectrogram image and runs it through a convolutional neural network trained on 2.5 million labelled bat calls and 3.5 million noise files. The model learns to recognise species from the shape, frequency, and timing of calls in the image, rather than from a handcrafted set of acoustic parameters. This approach means the system can pick up subtle patterns that are hard to encode manually, and the model is retrained every weekend as new verified data comes in. The BTO Acoustic Pipeline takes a different approach. Its bat classifier uses Random Forest (TADARIDA, Bas et al. 2017), while bird identification uses CNN classifiers (Gillings & Scott 2021). The pipeline handles not just bats but also small mammals, bush-crickets, and birds using separate classifiers. For UK bat species in particular, the BTO team has deep domain expertise and their classifiers reflect that. The fundamental difference: BioSonic is a self-service platform where you upload, review, and report within a single browser interface. The BTO Acoustic Pipeline is a processing service — you submit recordings via a desktop uploader, results come back as files, and downstream analysis is your responsibility. ## Accuracy comparison In the same [Wilder Sensing benchmark](/benchmarks), BioSonic was tested on 6,126 bat calls and scored 98.7% F1. BTO Acoustic Pipeline couldn't be benchmarked the same way — the report concluded: "BTO Pipeline is the least suitable bat analysis software… due to the inability to manually validate the AutoID results using the BTO Pipeline platform." | Metric | BioSonic | BTO Acoustic Pipeline | |---|---|---| | F1 score | 98.7% | Not benchmarkable* | | Manual validation | Right-click on spectrogram | Not supported on platform | | False detections of impossible species | None | Greater horseshoe (validated as Noise) & Escalera's bat (rare vagrant) — neither present in survey region | | File handling | Any length, any folder size | Auto-splits into 5-sec segments → inflated counts & more NoIDs | | RAG score (Wilder Sensing) | 16 / 18 | 9 / 18 — lowest of all 3 tools tested | *BTO could not be benchmarked because the platform does not support manual validation of AutoID results, and the CSV output is only compatible with Analook Insight, an outdated analysis tool. Have labelled recordings? [We will run a free comparison](mailto:josef.carlson@biosonic.se). ## What Wilder Sensing found when testing BTO > "BTO Pipeline is the least suitable bat analysis software for Wilder Sensing." > — Wilder Sensing benchmark report, 2025 > "BTO falsely detected Greater horseshoes with high confidence levels — but they were validated as Noise files." > — On BTO's false positives in Norfolk > "Cannot manually validate results. The csv. output file is only compatible with Analook Insight, an outdated bat analysis software." > — On the validation gap > "BTO automatically splits wav. files longer than 5 seconds into 5-second segments, resulting in more low confidence levels and NoIDs." > — On BTO's file-splitting (inflated 1,227 detections vs BioSonic's 745 on the same data) ## Day-to-day workflow ### Uploading recordings With BioSonic, you open a browser, drag files into a project, and the AI begins processing immediately. There is nothing to install. With the BTO Acoustic Pipeline, you download and install a desktop uploader application, configure your submission, and upload files for cloud processing. Results are returned as downloadable files. ### Reviewing results BioSonic provides an in-app review interface where ecologists can inspect spectrograms, confirm or override AI classifications, and flag uncertain calls for a second opinion. Every decision is logged with a timestamp and user ID. The BTO Acoustic Pipeline returns classification results as data files. Review and quality assurance happen in whatever tools your team already uses — typically spreadsheets or GIS software. ### Generating reports BioSonic includes one-click report generation: activity charts, species heatmaps, GPS-linked maps, and Word export formatted for planning applications. The BTO Acoustic Pipeline does not include a reporting layer. Building client-ready deliverables from the output files is up to you. ### Noise handling Rain, wind, insects, and equipment noise are the bane of bat acoustic surveys. BioSonic applies a dedicated deep-learning noise model trained on 3.5 million non-bat recordings to filter noise before classification. The BTO Acoustic Pipeline handles noise as part of its classification pipeline, filtering out non-target sounds during the identification process. ## Working as a team Bat survey work is rarely a solo effort. A typical consultancy has field technicians deploying detectors, junior ecologists doing first-pass review, and a named ecologist signing off reports. BioSonic is built for this: unlimited users per organisation, shared project dashboards, in-app review queues, and a full audit trail. Everyone works in the same browser interface, on the same data, in real time. The BTO Acoustic Pipeline does not include team collaboration features. Each user submits and receives results individually. Coordinating across a team means sharing files externally and managing review in spreadsheets or email. For solo researchers or small citizen-science projects, this is often fine. For a 15-person consultancy running 40 concurrent surveys, it becomes a bottleneck. > "Since joining BioSonic our analysis time has reduced significantly. The team is excellent in their support, super quick to respond." > — Luke Waddison, Senior Ecologist, [Wharton](/wharton) (UK) ## What UK consultants say We asked bat ecologists who have used both tools to describe their experience. These are direct quotes from working consultants, not marketing copy. | Category | BioSonic | BTO Acoustic Pipeline | |---|---|---| | Overall | "Super impressed. Super intuitive to use. Nothing annoyed me." | Cheap, but prep + upload friction is high. | | 5-sec file limit | Accepts any file length — "just drag a folder in and forget about it, let it turn away in the background." | Requires strict 5-second files. Takes days of mind-numbing work to sort. | | Folder size limit | Accepts any folder size. | "Only takes 29 GB folders, so you have to split everything manually." | | Upload experience | "Just drag a folder in and kind of forget about it." | "Nightmarish. You've got to break it into ~30 GB folders and queue them — total nightmare." | | Spectrogram review | "Being able to fly through yours would save a lot of time." "Drawing the box with right-click — that was nice." | No spectrogram review built in. | | UI / UX | "Not glitching or crashing. Super easy to use." | — | | Emergence graphs | "Yeah, that's really cool. This would be really good to have." | "I tried doing that years ago. It took me about two weeks on Excel. It was so painful." | | Graphs & charts | "These graphs would be so useful and save us a lot of time. That'd be really useful for swarming stuff." | Manual pivot tables — "probably human error by the 200th pivot table." | | Collaboration | "If I could share my project with Josh, just send a link — that'd be great." | — | | Data storage | "That's a massive selling point" (long-term storage + re-download). | Separate costly Azure storage. | | Summary | Built for real UK bat consultant workflows: minimal prep, fast proofing, one-click graphs, collaboration, storage, and huge time savings. | Strict limits, heavy manual prep, time-consuming script workarounds, and high friction despite being cheaper per GB. No review built in. | ## Time savings at scale: 7.5 TB season estimate To make the workflow difference concrete, we estimated the hours a UK consultancy would spend processing 7.5 TB of bat recordings in a single season using each tool. BioSonic saves an estimated 193 hours — roughly 5 full-time weeks. | Work item | BioSonic (hours) | BTO (hours) | Extra BTO hours | |---|---|---|---| | Sound analysis review | 80 | 141 | 61 | | Upload admin (30 GB folder cap) | 0 | 23 | 22.8 | | One-click charts & maps | 4 | 28 | 23.6 | | Emergence graphs | 1 | 12 | 10.8 | | Custom charts & maps | 7 | 28 | 20.6 | | Weather data integration | 0 | 14 | 14 | | 5-sec file conversion | 0 | 15 | 15 | | Rework / duplicated review | 5 | 16 | 10.6 | | Direct support (WhatsApp group) | 0 | 14 | 14.4 | | **Total hours** | **97** | **290** | **193** | | **Full-time weeks** | **2** | **7** | **5** | That is a 67% reduction in analyst time. At £65 per hour, the 193 extra hours with BTO represent roughly £12,500 of additional labour cost per season — before accounting for the stress of manual file splitting, pivot-table errors, and the absence of built-in review tools. ## Pricing model BioSonic charges per GB per year on an all-inclusive basis. That price covers AI classification, 9 years of EU-hosted storage, unlimited users, report generation, and ongoing model updates. There are no per-seat fees and no feature tiers. [Contact us for a quote](https://cal.com/biosonic/30min). The BTO Acoustic Pipeline offers several pricing options: - **Free tier:** 100 credits per year (approximately 100 GB), but this requires sharing your recordings with BTO for use in classifier training. - **Standard (no data sharing):** £1.40 per GB. - **PAYG bundles:** volume discounts available. - **Bespoke projects:** from £573 for custom analysis work. The free tier deserves a closer look. It requires sharing your recordings with BTO for classifier training. For citizen science and some NGO-funded work, this is a reasonable trade. For commercial consultancies working under NDA with wind farm operators or property developers, sharing client site recordings with a third party is often a dealbreaker. If your contracts include data confidentiality clauses, check whether the free tier's data-sharing requirement is compatible before committing. ## Who should choose which ### Choose BioSonic if: - You run a consultancy or survey team that needs shared projects, in-app review, and one-click reporting. - You work across multiple European countries and need consistent performance outside the UK. - Data confidentiality matters — your clients require that recordings stay within a controlled platform. - You want long-term storage included without managing your own archive. - You value weekly model updates over periodic releases. ### Choose BTO Acoustic Pipeline if: - Your work is exclusively UK-based and you want classifiers built by UK bat specialists with decades of local data. - You need non-bat taxa — birds, small mammals, bush-crickets. - You are running a citizen-science project and the free tier's data-sharing requirement is acceptable. - You are a solo researcher who prefers working with raw data files and building your own analysis pipeline. - Geographic coverage across 25+ countries matters more than team workflow features. These are not mutually exclusive. Some consultancies use BTO for UK-only surveys where the free tier works, and BioSonic for commercial projects where data confidentiality and team workflow are priorities. The tools accept the same WAV files, so running both in parallel is straightforward. ## Switching from BTO Acoustic Pipeline to BioSonic There is no format lock-in with either platform. Both accept standard WAV files from any bat detector. If you are currently using the BTO Acoustic Pipeline and want to try BioSonic, the process is simple: - **Step 1:** [Create a free BioSonic account](https://bat.biosonic.se/signup). - **Step 2:** Upload the same WAV files you would normally send to BTO. - **Step 3:** Compare results side by side on your own data. Most users see results within hours of uploading. If you want a guided migration or have questions about how BioSonic handles specific species groups, [email us directly](mailto:josef.carlson@biosonic.se) or [book a 30-minute call](https://cal.com/biosonic/30min). Your existing BTO results are yours to keep. BioSonic does not require you to abandon any previous analysis — you are adding a tool, not replacing an entire workflow overnight. ## Frequently asked questions ### Is the BTO Acoustic Pipeline free? BTO offers a free tier of 100 credits per year (roughly 100 GB), but it requires you to share your recordings with BTO for classifier training. If you cannot share data — for example, because of client NDAs — standard pricing is £1.40 per GB with no data-sharing requirement. PAYG bundles and bespoke project pricing are also available. ### Can I use BioSonic for UK bat species? Yes. BioSonic supports all UK bat species and is used by UK-based consultancies for commercial survey work. The AI model is trained on 2.5 million bat calls spanning European species, including all 18 UK breeding species. See our [benchmarks page](/benchmarks) for accuracy data. ### Does the BTO Acoustic Pipeline store my recordings? Not by default. Recordings are processed and results returned, but long-term storage is not included unless arranged as part of a bespoke project. BioSonic includes 9 years of EU-hosted storage as standard with every account. ### How does accuracy compare between BioSonic and BTO? In a third-party benchmark by Wilder Sensing, BioSonic scored 98.7% F1 on 6,126 calls across 15 species. BTO Acoustic Pipeline could not be benchmarked the same way — the report found that BTO's platform does not support manual validation of AutoID results, and concluded it was "the least suitable" of the three tools tested. BTO also falsely detected impossible species (Greater horseshoe bats validated as Noise) and inflated detection counts by auto-splitting files into 5-second segments. [See full benchmark results](/benchmarks). ### Can my whole team use BioSonic on one account? Yes. BioSonic supports unlimited users per organisation with shared projects, in-app review queues, and a full audit trail. There are no per-seat fees. Every team member works in the same browser interface. BTO Acoustic Pipeline does not include in-platform team features — collaboration happens outside the tool. ### What bat detectors work with each platform? Both BioSonic and BTO Acoustic Pipeline accept WAV files from any bat detector. Audiomoth, Wildlife Acoustics, Pettersson, Batlogger, Elekon — if it produces WAV output, it works with both platforms. ### Can I switch from BTO Acoustic Pipeline to BioSonic mid-season? Yes. There is no format lock-in. Upload your WAV files to BioSonic and results are typically ready within hours. You can run both platforms in parallel on the same recordings if you want to compare before committing. [Start a free trial](https://bat.biosonic.se/signup) to test it with your own data. ## Sources - Wilder Sensing (2025). Third-party benchmark report — BioSonic scored 98.7% F1; BTO rated least suitable of three tools tested. [See benchmarks page](/benchmarks) - Bas, Y. et al. (2017). "Tadarida: A toolbox for animal detection on acoustic recordings." J. Open Res. Softw. 5(1):6 - Gillings, S. & Scott, C. (2021). CNN classifiers for bird sound identification — BTO Research Report - BTO Acoustic Pipeline pricing: [bto.org](https://www.bto.org/our-science/projects/bto-acoustic-pipeline) Last verified: 2026-04-26. ## Ready to try it? Upload a batch of WAVs and compare results against your current workflow. No commitment, no credit card, no data sharing required. [Start free trial](https://bat.biosonic.se/signup) | [Book a demo](https://cal.com/biosonic/30min) Questions? Email [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se) --- # URL: https://biosonic.io/compare/batexplorer --- canonical: https://biosonic.io/compare/batexplorer --- # BioSonic vs BatExplorer Last updated: 2026-04-26 **BioSonic** is a cloud-based bat call analysis platform that uses deep learning on spectrograms to identify species automatically. **BatExplorer**, made by Elekon, is a Windows desktop application that uses rule-based call parameter measurements matched against a species reference library. Both tools help ecologists turn raw ultrasonic recordings into species-level identifications, but they take fundamentally different approaches to the task. This page lays out what each tool does, where it fits, and where it falls short. ## At a glance Fifteen dimensions that matter most when choosing bat analysis software. | Feature | BioSonic | BatExplorer (Elekon) | |---|---|---| | Underlying approach | Deep learning image recognition on spectrograms | Rule-based + species library matching (call parameter measurements) | | Training data | 2.5M bat calls + 4M noise files | Elekon's European species reference library | | Benchmarked accuracy | 98.7% F1 (third-party benchmark by Wilder Sensing, 6,126 calls, 15 species) | Not publicly benchmarked at scale | | Platform | Browser-based — works on any computer | Windows desktop only | | Team workflow | Unlimited users per organisation; shared projects | Single-user desktop licence per seat | | Species per recording | Up to 7 | One suggestion with probabilities | | Social and feeding calls | Separated automatically | Manual identification | | Noise filtering | Deep-learning noise model | Basic filters | | Model updates | New AI model every weekend | Version updates (irregular cadence) | | Storage | 9 years EU-hosted included | Local storage only | | Report generation | One-click charts, maps, Word export | Diagrams, charts, GIS export | | Detector compatibility | Any WAV-producing detector | Designed for Elekon BATLOGGER; accepts other WAV but with reduced functionality | | Geographic coverage | UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, Norway | European + UK species | | Pricing model | Per-GB-per-year, all-inclusive | Free Standard version; 249 CHF/yr (~£212) Pro per user | | Free trial | Yes | Free Standard tier + time-limited Pro trial | ## How each tool identifies bats The core difference is in the identification engine. BioSonic treats each recording as a spectrogram image and runs it through a deep convolutional neural network trained on over 2.5 million verified bat calls and 3.5 million noise samples. The model learns patterns across frequency, duration, shape and context, rather than relying on a fixed set of call parameter thresholds. It identifies up to seven species per recording and automatically separates social and feeding calls from echolocation passes. BatExplorer takes a traditional parameter-based approach. It measures call characteristics such as start and end frequency, duration, and inter-pulse interval, then matches these measurements against Elekon's reference library of European species. The software presents a single species suggestion with associated probabilities. This approach is well-established and transparent: you can see exactly which parameters drove the classification. The trade-off is that it relies on clean, well-separated calls to produce reliable measurements, and it struggles with overlapping species, noisy environments, and social calls that fall outside standard parameter ranges. Both approaches have merit. Parameter-based methods are interpretable and easy to audit. Deep learning methods handle ambiguity and noise better, but require trust in the training data and model validation. That is why independent benchmarking matters. ## What the benchmarks say - **BioSonic F1 score:** 98.7% — third-party benchmark by Wilder Sensing (6,126 calls, 15 species). - **BatExplorer benchmark:** No public data — No large-scale public benchmark for BatExplorer currently exists. BioSonic achieved 98.7% F1 in a third-party benchmark by Wilder Sensing on 6,126 calls across 15 species. The methodology, dataset, and results are available for review. For a deeper look at all benchmark results, see the [benchmarks page](/benchmarks). No public head-to-head benchmark between BioSonic and BatExplorer exists at the time of writing. We encourage independent testing. If you are a researcher or consultant who would like to run a comparison, we are happy to provide access and support. Contact us at [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se). We believe that transparent, reproducible benchmarking is the only honest way to compare analysis software. Claims without data are marketing. Numbers with methodology are evidence. ## Day-to-day workflow The practical experience of using each tool differs significantly, starting with how you get recordings into the system. **BioSonic workflow:** Open a browser, upload WAV files from any detector, and wait for the AI to process them. Results appear within minutes for typical survey datasets. Species identifications, spectrograms, activity charts, and maps are generated automatically. You can filter, review, adjust classifications, and export reports directly from the browser. There is nothing to install. Updates happen server-side every weekend. Your data is stored in the EU for up to nine years, accessible from any device. **BatExplorer workflow:** Install the desktop application on a Windows machine. Connect your BATLOGGER via USB or import WAV files manually. BatExplorer reads recordings, displays spectrograms, and runs its auto-identification. You review results call by call, adjusting parameters and classifications as needed. Reports and charts can be generated locally. GIS export is available for spatial analysis. All data stays on your local machine or network drive. If you value working from a single machine with no internet dependency, BatExplorer fits that requirement. If you need to process data from multiple detectors, access results from a tablet in the field, or hand off a project to a colleague without copying files, BioSonic is the more flexible option. Processing speed matters at scale. A survey season generating hundreds of thousands of recordings takes considerably longer to work through in a desktop tool that processes files sequentially. BioSonic's cloud infrastructure handles large uploads in parallel, so a full season of data from ten detectors is processed in the same timeframe as data from one. ## Working as a team Ecological consultancies rarely have just one person analysing bat data. Projects involve field technicians, junior analysts, senior ecologists, and report writers. The tools differ sharply on how they handle collaboration. **BioSonic** includes unlimited users per organisation at no additional cost. Everyone on your team logs in to the same platform, sees the same projects, and works from the same dataset. When one analyst reviews and confirms a species identification, the update is visible to the whole team immediately. Project managers can track progress across multiple sites without requesting file transfers or status emails. **BatExplorer** is licensed per seat. Each user works on their own installation with their own copy of the data. Sharing results means exporting files and sending them to colleagues, or saving to a shared network drive. There is no built-in mechanism for real-time collaboration or centralised project oversight. For a team of five analysts, you need five licences and a disciplined file-management protocol to avoid version conflicts. > "Since joining BioSonic our analysis time has reduced significantly. The team is excellent in their support, super quick to respond." > — Luke Waddison, Senior Ecologist, [Wharton](/wharton) (UK) For solo consultants working exclusively with BATLOGGER hardware, the single-user model is fine. For any team with more than two people, centralised cloud-based collaboration removes a meaningful amount of administrative overhead. ## Pricing model **BioSonic** uses a per-GB-per-year pricing model. You pay for the volume of data you store and process. All features are included. There are no per-user fees, no module add-ons, and no tier restrictions. Every user in your organisation gets access to the full platform. **BatExplorer** offers a tiered model. The Standard version is free and includes basic spectrogram viewing and manual analysis. The Pro version, which adds auto-identification and advanced features, costs 249 CHF per year (approximately £212) for a Pro licence. Each seat requires its own licence. For a team of five ecologists, that is roughly £1,060 per year, and each person still works in isolation. The free Standard tier is a genuine advantage for BatExplorer. If you are a student, hobbyist, or occasional surveyor who only needs to view and manually analyse spectrograms, it costs nothing. For professional use requiring auto-identification, the per-seat cost adds up as teams grow. BioSonic's model scales with data volume rather than team size, which tends to favour larger consultancies processing significant volumes of recordings. Both tools offer trials. BioSonic provides a free trial with full functionality. BatExplorer offers the permanent free Standard tier plus a time-limited trial of Pro features. We encourage you to try both and compare results on your own data. ## Who should choose which **BatExplorer may be the better fit if you:** - Use Elekon BATLOGGER hardware exclusively and want tight integration with your detector - Prefer a desktop application with no internet dependency - Work solo and do not need team collaboration features - Want a free tool for manual spectrogram review and basic analysis - Are comfortable with parameter-based identification and prefer its interpretability **BioSonic may be the better fit if you:** - Want the highest benchmarked accuracy: 98.7% F1 means less time fixing wrong classifications - Have a consultancy team: unlimited users, shared projects, single audit trail - Use any bat detector brand, not just Elekon - Want integrated spectrogram review with right-click bounding boxes, no switching apps - Need one-click charts, maps, emergence graphs, weather overlays, and Word reports - Want up to 7 species per recording with automatic feeding buzz and social call separation - Value weekly model updates that improve accuracy continuously BatExplorer is a mature, well-designed tool that serves Elekon users well. BioSonic addresses a different set of needs: higher accuracy, team collaboration, integrated review, and one-click reporting. ## Switching from BatExplorer to BioSonic If you decide to move from BatExplorer to BioSonic, the process is straightforward. There is no proprietary file conversion or data lock-in to worry about. **Step 1: Collect your WAV files.** BatExplorer stores recordings as standard WAV files on your local machine. Locate the folders containing your survey data. Any WAV file from any detector works with BioSonic. **Step 2: Create a BioSonic account.** Sign up at [bat.biosonic.se/signup](https://bat.biosonic.se/signup) and create your organisation. Invite your team members at no additional cost. **Step 3: Upload your recordings.** Drag and drop your WAV files into BioSonic. The platform processes them automatically. For large archives, contact our team and we can arrange bulk upload support. **Step 4: Review and compare.** Once processing is complete, you can compare BioSonic's identifications against your previous BatExplorer results. Many users find this a useful exercise for building confidence in a new tool before switching their live projects. Your BatExplorer installation continues to work alongside BioSonic. You do not need to delete or uninstall anything. Run both in parallel for a survey season and decide based on your own results. If you need assistance migrating, contact us at [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se) or [book a 30-minute call](https://cal.com/biosonic/30min). ## Frequently asked questions ### Can I use BatExplorer with non-Elekon detectors? BatExplorer can import WAV files from other detectors, but it is designed and optimised for Elekon BATLOGGER hardware. Some features and metadata integration may be reduced when using recordings from other manufacturers. BioSonic accepts standard WAV files from any detector without reduced functionality. ### Is BatExplorer free? BatExplorer offers a free Standard version with basic spectrogram viewing and manual analysis features. The Pro version, which includes auto-identification and advanced analysis tools, costs 249 CHF per year (approximately £212) for a Pro licence. BioSonic uses a per-GB-per-year model with all features included from day one. ### How does BioSonic's accuracy compare to BatExplorer? BioSonic achieved 98.7% F1 in a third-party benchmark by Wilder Sensing on 6,126 calls across 15 species. No public head-to-head benchmark between BioSonic and BatExplorer currently exists. We encourage independent testing and are happy to support any researcher who wants to run a comparison. See our [benchmarks page](/benchmarks) for full details. ### Can I switch from BatExplorer to BioSonic? Yes. BioSonic accepts standard WAV files, so you can upload the same recordings you have been analysing in BatExplorer. There is no lock-in or conversion step required. Upload your files, and BioSonic processes them automatically. You can run both tools in parallel to compare results before committing to a switch. ### Does BioSonic work on Mac or Linux? Yes. BioSonic is browser-based and runs on any operating system with a modern web browser, including Windows, macOS, Linux, and tablets. BatExplorer runs on Windows only. ### Which tool is better for team projects? BioSonic is designed for team collaboration with unlimited users per organisation, shared projects, and centralised cloud storage. BatExplorer is a single-user desktop application with per-seat licensing. If multiple ecologists need to access the same project simultaneously, BioSonic is the more practical choice. For more on how teams use BioSonic, see the [Wharton case study](/wharton). ### How often are the AI models updated? BioSonic deploys a new AI model every weekend, incorporating the latest training data and improvements. BatExplorer receives version updates on an irregular cadence, typically tied to software releases from Elekon. Weekly updates mean BioSonic's accuracy improves continuously without any action required from the user. ### What happens to my data in BioSonic? All data uploaded to BioSonic is stored on EU-hosted servers and retained for up to nine years. You retain full ownership of your recordings and results. Data can be exported at any time. For questions about data handling, privacy, or GDPR compliance, see our [FAQ page](/faq) or contact us directly. ## Sources - BatExplorer v2.3 (September 2025): 32-species European library, Windows 10/11 + .NET 4.8 - Elekon AG pricing: 249 CHF/year Pro subscription - Wilder Sensing (2024). Third-party benchmark on 6,126 calls across 15 species — [see benchmarks page](/benchmarks) Last verified: 2026-04-26. ## Get started Upload your recordings, compare the results against your current workflow, and decide based on evidence. Free trial, no credit card required. [Start free trial](https://bat.biosonic.se/signup) | [Book a demo](https://cal.com/biosonic/30min) Questions? Email [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se) --- # URL: https://biosonic.io/compare/sonobat --- canonical: https://biosonic.io/compare/sonobat --- # BioSonic vs SonoBat UK Last updated: 2026-04-26 BioSonic and SonoBat UK both identify bat species from acoustic recordings, but they take fundamentally different approaches. BioSonic is a cloud platform that uses deep learning image recognition on spectrograms to classify calls. SonoBat UK is a desktop application built over 30 years that extracts full-spectrum call parameters and applies machine learning combined with rule-based vetting algorithms. This page sets out how the two tools differ in accuracy, workflow, team features, and cost so you can decide which fits your projects. ## At a glance | Feature | BioSonic | SonoBat UK | |---|---|---| | Underlying approach | Deep learning image recognition on spectrograms | Full-spectrum data extraction + machine learning + rule-based vetting algorithms (30 years of development) | | Training data | 2.5M bat calls + 4M noise files | UK species-confirmed reference library built from systematic tracking since 2017 | | Benchmarked accuracy | 98.7% F1 (third-party benchmark by Wilder Sensing, 6,126 calls, 15 species) | Not publicly benchmarked at scale; strong reputation for low false-positive rate | | Platform | Browser-based — works on any computer | Desktop (macOS, Windows) | | Team workflow | Unlimited users per organisation; shared projects | Single-user per licence; SonoBat LIVE for real-time recording | | Species per recording | Up to 7 | Sequence-level decision with vetting for call context | | Social and feeding calls | Separated automatically | Buzz Mode detects feeding buzzes; social calls identified by vetter | | Noise filtering | Deep-learning noise model | Scrub tool + quality assessments during extraction | | Model updates | New AI model every weekend | Infrequent version releases | | Storage | 9 years EU-hosted included | Local storage only | | Report generation | One-click charts, maps, Word export | Customisable Vetting Table export; iBatsID format; customisable output tables | | Detector compatibility | Any WAV-producing detector | Any full-spectrum WAV detector | | Geographic coverage | UK, Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, Norway | UK species only (separate SonoBat North America product) | | Pricing model | Per-GB-per-year, all-inclusive | £880 perpetual or £320/£640/£960 lease (1/2/3 year) | | Free trial | Yes | Free trial available | ## How each tool identifies bats ### BioSonic: spectrogram-level deep learning BioSonic converts each recording into a spectrogram image, then runs a convolutional neural network trained on 2.5 million confirmed bat calls and 3.5 million noise files. The model sees spatial patterns across the full time-frequency image rather than measuring individual call parameters. This means it can detect overlapping species, social calls, and feeding buzzes in a single pass without explicit feature extraction. Because the model trains on images, it generalises well to recordings made in different conditions, with different detectors, and across geographic regions. BioSonic retrains its model every weekend, incorporating newly verified data so accuracy improves continuously. ### SonoBat UK: parameter extraction and vetting SonoBat takes a different path. It extracts quantitative call parameters from full-spectrum recordings, including frequency, duration, bandwidth, and slope measurements. Those parameters feed into a machine learning classifier combined with rule-based vetting algorithms refined over 30 years of development. The vetting step is central to SonoBat's philosophy. Rather than forcing a classification on every call, SonoBat applies quality thresholds and contextual rules to reject ambiguous calls. This conservative approach is a deliberate design choice that prioritises low false-positive rates over high recall. The UK reference library contains 38,187 calls covering 17 UK and Ireland species (build 30.1, October 2023). Both approaches are valid. The question is which trade-offs matter most for your work. ## What the benchmarks say - **BioSonic:** 98.7% F1 — third-party benchmark by Wilder Sensing on 6,126 calls across 15 species. - **SonoBat UK:** No public benchmark — Strong reputation for low false-positive rates, but no large-scale public benchmark results are available. BioSonic's 98.7% F1 comes from an independent third-party benchmark by Wilder Sensing that tested the system on 6,126 expert-verified bat calls across 15 species. For a deeper look at BioSonic's benchmark results, including species-level breakdowns, see the [benchmarks page](/benchmarks). For SonoBat UK, no public head-to-head benchmark exists. SonoBat has a well-earned reputation in the UK ecological community for producing reliable results with low false-positive rates, backed by decades of refinement. We encourage independent testing and would welcome any published comparison. If you want to verify accuracy yourself, both tools offer free trials. We recommend running the same set of expert-verified recordings through each system and comparing the results against your own manual analysis. ## Day-to-day workflow ### Getting started With BioSonic, you open a browser, drag files into your project, and results appear within minutes. There is nothing to install, and you can work from any device. With SonoBat, you install the desktop application on macOS or Windows, then process files locally. SonoBat also offers SonoBat LIVE for real-time recording and analysis in the field. ### Processing recordings BioSonic processes recordings in the cloud. Upload a batch of WAV files, and the system handles noise filtering, species classification, and social/feeding call separation automatically. You can process thousands of files without worrying about your laptop's processing power or storage limits. SonoBat processes files on your local machine. The Scrub tool cleans recordings before analysis, and quality assessments filter out poor-quality calls during parameter extraction. Processing speed depends on your hardware, which matters when you have large datasets from multi-night surveys. ### Reviewing results BioSonic shows results in a browser interface with spectrograms, species lists, and confidence scores. You can filter, sort, and review calls without switching applications. Reports export to Word with one click, including charts and maps. SonoBat presents results through its Vetting Table, which is highly customisable. You can configure output columns, apply additional vetting thresholds, and export in multiple formats including iBatsID. For consultants who have built their reporting workflow around SonoBat's table format, this customisability is a genuine advantage. ## Working as a team This is where the two tools diverge most sharply. BioSonic includes unlimited users per organisation at no extra cost. Every team member accesses the same projects, sees the same results, and can review or approve analyses from anywhere. For consultancies with multiple ecologists working on the same survey, this removes the bottleneck of passing files and spreadsheets between machines. SonoBat UK licences are single-user. Each ecologist needs their own licence, and sharing results means exporting tables and distributing files manually. If your team works independently on separate projects, this may not matter. But for firms that need shared access to project data, the single-user model creates friction. > "Since joining BioSonic our analysis time has reduced significantly. The team is excellent in their support, super quick to respond." > — Luke Waddison, Senior Ecologist, [Wharton](/wharton) (UK) ## Pricing model ### BioSonic BioSonic charges per GB per year on an all-inclusive basis. The price covers unlimited users, cloud storage (9 years EU-hosted), all model updates, and report generation. You pay for the data you process, not for seats. For teams, this scales predictably with workload rather than headcount. ### SonoBat UK SonoBat UK offers a perpetual licence at £880 or lease options at £320 (1 year), £640 (2 years), or £960 (3 years). Build 30.1 (October 2023) includes a 38,187-call UK reference library covering 17 UK and Ireland species. The software runs on Windows and macOS. Each licence covers one user. ### Which is cheaper? It depends on your data volume and team size. A solo consultant processing a modest amount of data may find SonoBat's perpetual licence more economical in the long run. A team of five ecologists processing hundreds of gigabytes per season will likely find BioSonic's per-GB model cheaper than five separate SonoBat licences. Both tools offer free trials, so you can estimate costs before committing. ## Who should choose which ### SonoBat UK may be a better fit if you: - Work exclusively on UK species - Prefer a desktop application that processes data locally without internet dependency - Value SonoBat's conservative vetting approach and have built your workflow around its customisable output tables - Work solo and want a perpetual licence rather than ongoing subscription costs - Need SonoBat LIVE for real-time field recording and analysis ### BioSonic may be a better fit if you: - Want the highest benchmarked accuracy: 98.7% F1 means less time fixing wrong classifications - Have a consultancy team: unlimited users, shared projects, single audit trail - Want integrated spectrogram review built into the platform, no switching between apps - Need one-click charts, maps, emergence graphs, weather overlays, and Word reports - Want up to 7 species per recording with automatic feeding buzz and social call separation - Prefer weekly model updates over infrequent version releases SonoBat has a loyal following in the UK built on 30 years of reliable service. If your current workflow works well and you work solo, there may be no reason to switch. But if you are finding that local processing, single-user licences, or manual reporting are slowing you down, BioSonic addresses those constraints. ## Switching from SonoBat to BioSonic If you decide to try BioSonic alongside or instead of SonoBat, the transition is straightforward. - **Your recordings are compatible.** Both tools accept standard WAV files from any full-spectrum bat detector. You do not need to re-record anything. - **Upload existing files.** Drag your existing WAV files into a BioSonic project. Results typically appear within minutes, even for large batches. - **Run both tools in parallel.** Many consultants run BioSonic and SonoBat side by side during their first season to compare results and build confidence before committing fully. - **No data lock-in.** BioSonic stores your original recordings and exports results in standard formats. You can download everything at any time. - **Team onboarding.** Add your colleagues to your BioSonic organisation. They can start reviewing results immediately with no additional licence cost. If you have questions about migrating your workflow, [email us directly](mailto:josef.carlson@biosonic.se) or [book a 30-minute call](https://cal.com/biosonic/30min) and we will walk you through it. ## Frequently asked questions ### Is SonoBat UK more accurate than BioSonic? No public head-to-head benchmark exists between the two tools. BioSonic achieved 98.7% F1 in a third-party benchmark by Wilder Sensing on 6,126 calls across 15 species. SonoBat has a strong reputation for low false-positive rates built over 30 years, but has not published comparable large-scale benchmark results. We encourage independent testing and recommend running your own verified recordings through both systems. See our [benchmarks page](/benchmarks) for BioSonic's detailed results. ### Can I use my existing SonoBat recordings in BioSonic? Yes. Both tools accept standard WAV files from any full-spectrum bat detector. Upload the same recordings to BioSonic for a second opinion or to take advantage of cloud-based team features. No conversion or re-recording is needed. ### Does BioSonic cover all UK bat species? Yes. BioSonic covers all UK bat species and also supports the Netherlands, Denmark, Germany, Sweden, Belgium, France, Poland, Finland, and Norway. SonoBat UK is limited to UK species, with a separate SonoBat North America product for that region. ### How much does SonoBat UK cost compared to BioSonic? SonoBat UK costs £880 for a perpetual licence, or £320/£640/£960 for a 1/2/3-year lease. Build 30.1 includes a 38,187-call reference library. Each licence covers one user. BioSonic charges per GB per year with unlimited users included. The cheaper option depends on your data volume and team size. Both offer free trials so you can estimate costs before committing. ### Can multiple team members use SonoBat UK simultaneously? SonoBat UK licences are single-user. Each team member needs their own licence. BioSonic includes unlimited users per organisation with shared projects, which may be more cost-effective for teams. See our [product page](/product) for details on team features. ### Does SonoBat UK work offline? Yes. SonoBat is a desktop application that processes everything locally, so it works without an internet connection. BioSonic requires internet access to upload recordings and retrieve results, but works from any browser on any device once connected. ### How often are BioSonic and SonoBat updated? BioSonic deploys a new AI model every weekend, incorporating the latest training data from its growing library of 2.5 million bat calls. SonoBat releases updates less frequently as version releases. BioSonic's weekly cycle means accuracy improvements reach users faster. ## Sources - SonoBat UK build 30.1 (October 2023): 38,187-call reference library, 17 UK + Ireland species - SonoBat pricing: £880 perpetual, £320/£640/£960 lease (1/2/3 year) - SonoBat: [sonobat.com](https://sonobat.com) - Wilder Sensing (2024). Third-party benchmark on 6,126 calls across 15 species — [see benchmarks page](/benchmarks) Last verified: 2026-04-26. ## Try it yourself Upload a batch of WAV files and compare results against your current workflow. Free trial, no credit card required. [Start free trial](https://bat.biosonic.se/signup) | [Book a 30-min call](https://cal.com/biosonic/30min) Questions? Email [josef.carlson@biosonic.se](mailto:josef.carlson@biosonic.se) — we reply within a few hours. --- # URL: https://biosonic.io/customers --- canonical: https://biosonic.io/customers --- # Customer case studies How ecology consultancies across Europe save time and improve accuracy with BioSonic. ## [Wharton](/wharton) A UK ecology consultancy compared one survey night side by side in Kaleidoscope and BioSonic, then projected the savings across their full season. 66% time saved — 2.5 hours became 50 minutes. ## [DMR](/dmr) A Danish environmental consultancy started with one 5 GB test project. Their lead biologist saved at least 10 hours on it, then the rest of the team came on board. --- # URL: https://biosonic.io/founders # About BioSonic BioSonic is a Stockholm-based AI company that builds the most accurate bat sound analysis platform in independent benchmarks. Founded in 2024, our deep learning models are trained on 2.5 million bat calls and 3.5 million noise files, achieving 98.7% F1 accuracy in independent benchmarks. Used by ecological consultants, government agencies, and conservation researchers across Europe. ## Founders - **Josef Carlson** — CEO & Co-Founder. Leads commercial strategy, product direction, and customer relationships. - **Jonathan Carlson** — CTO & Co-Founder. Creator of BioSonic's AI Architecture for Bat Classification. ## Awards - Bert-Inge Hogsved Prize for Best Entrepreneurship 2025, Chalmers University of Technology ## Press & Validation - Featured in Chalmers University of Technology news - Third-party benchmark validation by Wilder Sensing (98.7% F1, 6,126 calls, 15 species) ## Contact - **Company**: BioSonic AB - **Address**: Bergsundsgatan 7, 117 37 Stockholm, Sweden - **Email**: josef.carlson@biosonic.se - **Phone**: +46 72 744 65 85 - **Org.nr**: 559472-8056 - **Website**: https://biosonic.io --- # URL: https://biosonic.io/wind-farm-bat-monitoring # Wind Farm Bat Monitoring Software Software for wind farm bat monitoring, curtailment analysis, and regulatory compliance. BioSonic handles nacelle-height recordings at scale with 98.7% F1 accuracy. ## Why Wind Farms Need Bat Monitoring European regulations (Habitats Directive, EIA requirements) mandate bat monitoring before and after wind turbine installation. Accurate species identification determines curtailment schedules and regulatory compliance. Key frameworks include RENEBAT/ProBat (Germany), Natural England guidelines (UK), and national implementations of the EU Habitats Directive. ## The Challenge Wind farm bat monitoring generates massive continuous datasets from detectors at nacelle height and ground level. Nacelle recordings contain heavy wind and mechanical noise. Misidentifying species has direct consequences for turbine operation and regulatory compliance. ## How BioSonic Helps - **98.7% F1 accuracy** for reliable species-level classification at scale - **13.4x fewer false positives** — critical for noisy nacelle-height recordings - **Behaviour classification** — separates echolocation, feeding buzzes, and social calls - **Team collaboration** — multiple biologists work on the same wind farm project simultaneously - **Cloud-based** — no desktop software to install or maintain ## Customer Reference COWI, a major Danish engineering consultancy, uses BioSonic for wind farm bat monitoring across Denmark. > "What sets BioSonic apart is their exceptional accuracy, both in species identification and in filtering out noise. They're truly at the top." > — Peter Sivertsen, Biologist, COWI (Denmark) ## Related Pages - [Benchmarks](https://biosonic.io/benchmarks) - [Compare Bat Analysis Software](https://biosonic.io/compare) --- # URL: https://biosonic.io/automatic-bat-species-identification # Automatic Bat Species Identification How automatic bat species identification works. Deep learning models trained on 2.5 million calls achieve 98.7% F1 accuracy. Compare AI classifiers for bat echolocation analysis. ## How It Works Automatic bat species identification uses machine learning to classify bat echolocation calls to species level. The process: raw ultrasonic recordings are converted to spectrograms, a deep learning model extracts acoustic features (call frequency, duration, shape, inter-pulse interval), and a classifier assigns species probabilities. ## BioSonic's Approach BioSonic uses convolutional neural networks trained on 2.5 million verified bat calls and 3.5 million noise files. The model identifies species, classifies behaviour (echolocation, feeding buzzes, social calls), and filters non-bat noise with 13.4x fewer false positives than alternatives. ## Accuracy - 98.7% F1 score in independent benchmarks (Wilder Sensing, Annabel Jeffries) - 13.4x fewer false positives than Kaleidoscope Pro - 16.3x fewer missed bats than Kaleidoscope Pro ## Related Pages - [Benchmarks](https://biosonic.io/benchmarks) - [Compare Bat Analysis Software](https://biosonic.io/compare) - [FAQ](https://biosonic.io/faq) --- # URL: https://biosonic.io/bat-acoustic-analysis-software --- canonical: https://biosonic.io/bat-acoustic-analysis-software --- # Bat Acoustic Analysis Software Compare bat acoustic analysis software: BioSonic, Kaleidoscope Pro, SonoBat, BatExplorer, and BTO Acoustic Pipeline — feature comparison and accuracy benchmarks. ## What bat acoustic analysis software does Bat acoustic analysis software processes ultrasonic recordings to detect bat passes, identify species from echolocation calls, and filter out non-bat noise. Tools differ in method (deep learning on spectrograms vs rule-based / classical classifiers), deployment (cloud vs desktop), and accuracy. ## Accuracy comparison (F1, species identification) - BioSonic — 98.7% F1 (independent Wilder Sensing benchmark), deep learning on spectrograms trained on 2.5 million verified bat calls and 3.5 million noise files - Kaleidoscope Pro — approximately 73.0% F1 - SonoBat — around 85% (varies by region) - BTO Acoustic Pipeline — approximately 90% for UK species ## Why BioSonic leads BioSonic applies convolutional neural networks (image recognition) to spectrograms, drawing bounding boxes around each call and classifying species from visual patterns. This produces a step change in accuracy over classical approaches, validated against expert-verified reference datasets. ## Related Pages - [Benchmarks](https://biosonic.io/benchmarks) - [Compare Bat Analysis Software](https://biosonic.io/compare) - [Automatic Bat Species Identification](https://biosonic.io/automatic-bat-species-identification) --- # URL: https://biosonic.io/bat-survey-software-uk --- canonical: https://biosonic.io/bat-survey-software-uk --- # Bat Survey Software for UK Ecological Consultants AI bat-call identification for UK survey teams: sort a season of static-detector and transect recordings by species, keep manual verification per BCT good-practice guidance, and export EIA-ready results. Built for activity surveys, Environmental Impact Assessment (EIA), and planning conditions. ## Who it is for UK ecological consultants and survey teams processing large volumes of static and transect bat recordings for activity surveys, EIA, and planning conditions. ## How it works 1. Upload full-spectrum WAV recordings from any detector (AudioMoth, Wildlife Acoustics Song Meter and Echo Meter, Elekon Batlogger, Titley Anabat, Pettersson) — runs in the browser, no install. 2. Image recognition on spectrograms classifies every pass by species, flags feeding buzzes and social calls, and surfaces low-confidence calls so noise and empty files never reach the analyst. 3. Review a confidence-ranked shortlist, keep manual verification where it counts, and export client-ready results for the ecology chapter of an Environmental Statement. ## Accuracy and verification - 98.7% F1 in an independent third-party benchmark by Wilder Sensing (6,126 calls across 15 species), ranking first of all classifiers tested, versus 73.0% for Kaleidoscope Pro. - Covers all 17 breeding UK bat species (18 recorded), including the hard Myotis and Plecotus groups. - Automated identifications are designed to be manually verified, in line with BCT good-practice guidance and the 4th-edition Good Practice Guidelines (Collins, ed., 2023). Myotis and Plecotus, which classifiers struggle with most, always go to manual verification. ## Data residency Recordings and results are stored in UK and EU data centres, encrypted in transit and at rest. ## Related Pages - [Benchmarks](https://biosonic.io/benchmarks) - [Compare BioSonic vs Kaleidoscope Pro](https://biosonic.io/compare/kaleidoscope) - [Automatic Bat Species Identification](https://biosonic.io/automatic-bat-species-identification) --- # URL: https://biosonic.io/bat-school --- canonical: https://biosonic.io/bat-school --- # Which bat is that? — Bat School A visual guide to identifying European bats from their calls. Pick your country and compare frequency, distance between calls, call shape, social calls and look-alikes — with example spectrograms for every species. ## What it covers - Peak / characteristic frequency per species - Inter-pulse interval (distance between calls) - Call shape (FM, QCF, CF) and how to read it on a spectrogram - Social calls and how they differ from echolocation - Common look-alike species and how to tell them apart - Country-specific species lists with example spectrograms A free educational resource for ecologists, students, and bat-survey teams learning manual call identification — complementary to BioSonic's automatic classification. ## Related Pages - [Automatic Bat Species Identification](https://biosonic.io/automatic-bat-species-identification) - [Product](https://biosonic.io/product) --- # URL: https://biosonic.io/wharton --- canonical: https://biosonic.io/wharton --- # How Wharton saved 66% of their bat sound analysis time with BioSonic A UK ecology consultancy compared one survey night side by side in Kaleidoscope and BioSonic, then projected the savings across their full season. 2.5 hours became 50 minutes. ## Results - 66% less time on the same data — 2.5 hours became 50 minutes for one survey night - About 3,000 noise files filtered out on that night - Projected across a full season: roughly 490 hours in Kaleidoscope vs roughly 163 hours in BioSonic - About 327 hours saved per year > "Since joining BioSonic our analysis time has reduced significantly, freeing us up to take on different tasks. The system is a pleasure to use and the team are excellent in their support." > — Luke Waddison, Senior Ecologist, Wharton ## Related Pages - [Customer Case Studies](https://biosonic.io/customers) - [Time Calculator](https://biosonic.io/tools/time-calculator) --- # URL: https://biosonic.io/cowi --- canonical: https://biosonic.io/cowi --- # How COWI powers wind farm bat monitoring across Denmark with BioSonic COWI, a major Danish engineering and ecology consultancy, uses BioSonic for bat monitoring on wind farm projects across Denmark. ## Why COWI uses BioSonic Wind farm bat monitoring produces large, noisy datasets from nacelle-height and ground-level detectors. COWI relies on BioSonic's accuracy and noise filtering to identify species reliably at scale and support curtailment and compliance decisions. > "What sets BioSonic apart is their exceptional accuracy, both in species identification and in filtering out noise. In this, they're truly at the top." > — Peter Sivertsen, Biologist, COWI (Denmark) ## Related Pages - [Wind Farm Bat Monitoring](https://biosonic.io/wind-farm-bat-monitoring) - [Customer Case Studies](https://biosonic.io/customers) --- # URL: https://biosonic.io/tools/time-calculator --- canonical: https://biosonic.io/tools/time-calculator --- # Bat Survey Time Calculator Estimate the hours you'll spend analysing bat WAV recordings, with and without AI. Based on real data: 2TB takes roughly 495 hours by hand and roughly 168 hours with BioSonic. ## What it shows - Analysis hours for a traditional Kaleidoscope-class workflow vs BioSonic, scaled to your data volume - Example: 2TB of recordings — about 495 hours by hand vs about 168 hours with BioSonic, roughly 327 hours saved per year - Roughly 66% time saving on sound analysis (up to ~86% in an anecdotal best case) - Context: classifier + manual post-validation workflows are generally cited with a ceiling around 75% (Barré et al. 2019) - Grounded in Wharton's documented result of about 327 hours saved per year on 2TB of recordings ## Related tools - [True Cost of Kaleidoscope Pro](https://biosonic.io/tools/kaleidoscope-roi-calculator) - [Storage Calculator](https://biosonic.io/tools/storage-calculator) --- # Company information # BioSonic AB BioSonic is a Stockholm-based biodiversity and AI company building the most accurate bat sound analysis platform in independent benchmarks. Founded in 2024 by Josef Carlson (CEO) and Jonathan Carlson (CTO). The model is trained on over 2.5 million verified bat calls and 3.5 million noise files curated by 12 bat experts, and reached 98.7% F1 in an independent third-party benchmark by Wilder Sensing (6,126 calls across 15 species). ## Company facts - Company: BioSonic AB - Founded: 2024, Stockholm, Sweden - Headquarters: Bergsundsgatan 7, 117 37 Stockholm, Sweden - Org.nr: 559472-8056 - Operates with customers across the UK, Netherlands, Denmark, Germany, and Sweden - Compliance: GDPR & UK GDPR; EU-hosted storage and infrastructure - Analytics: PostHog (privacy-conscious product analytics) ## Founders - Josef Carlson — CEO & Co-Founder — https://www.linkedin.com/in/josef-carlson/ - Jonathan Carlson — CTO & Co-Founder — creator of BioSonic's AI architecture for bat classification; Chalmers University of Technology alumnus; awarded the Bert-Inge Hogsved Prize for Best Entrepreneurship 2025 ## Contact - Website: https://biosonic.io - App / free trial: https://bat.biosonic.se/signup - Book a demo: https://cal.com/biosonic/30min - Email: josef.carlson@biosonic.se - General / DPA: hello@biosonic.se - Phone: +46 72 744 65 85 - LinkedIn: https://www.linkedin.com/company/biosonic-ab/