How to Choose an Acoustic Camera for Industrial Gas Leak Detection
A practical buyer's guide to selecting the right acoustic camera based on performance, safety, and real industrial use cases.
Start with your objective: what problem are you solving?
Before comparing acoustic cameras, clarify what you want to achieve.
The “right” camera depends less on the device itself than on the impact of the leaks you are trying to detect. In industrial environments, gas leaks typically affect four areas:
- Operating cost: compressed air and industrial gases represent a continuous expense. Even moderate leaks can generate significant annual losses.
- Reliability: leaks can reduce system performance, increase equipment load, and contribute to unplanned downtime.
- Safety: depending on the gas, leaks may create explosion risks, asphyxiation hazards, or unsafe working conditions.
- Environmental compliance: methane and other emissions are increasingly regulated, with requirements for detection, documentation, and sometimes quantification.
These objectives are not equivalent. A camera that is sufficient for reducing compressed air losses may not provide the performance required for process gases, safety-critical leaks, or emissions monitoring.
Defining your primary objective is the first step in selecting the appropriate level of performance.
Understand your operating constraints
Once your objective is defined, the next step is to understand the constraints of your operating environment.
These constraints determine the level of detection performance, accuracy, and reliability required from the acoustic camera.
| Application context | Key constraints | What it means for your camera |
| Compressed air and utilities (incl. atmospheric gases) | Leaks of interest are typically loud relative to background noise in most environments. Systems are often accessible and inspections are routine. | Detection is generally not the limiting factor. Leak rate quantification accuracy, reporting capabilities, and ease of use become the main differentiators. |
| Higher-value gases (e.g. Ar, Kr, Xe) | Same environments as utilities, but gas losses are significantly more costly, increasing the importance of detecting smaller leaks and quantifying them reliably. | Higher detection performance and robust quantification models become more important to prioritize repairs and estimate losses accurately. |
| Industrial gases (e.g. methane, hydrogen, syngas, ethylene) | Complex systems, variable pressures, congested installations, and often high background noise. Leaks may impact safety and reliability. | Requires strong detection performance in real conditions, robustness to noise, and repeatable, operator-independent inspections. |
| Methane and LDAR programs | Large-scale inspections, low signal levels, regulatory requirements, and need for traceable and auditable results. | Requires high sensitivity, wide-area coverage, and validated performance (case studies, independent testing). |
| Hazardous (Ex) environments | Presence of explosive atmospheres, strict safety procedures, and operational constraints. | ATEX / IECEx / UL certification is mandatory, along with safe and practical usability in the field. |
| Large or high-noise industrial sites | High ambient noise (compressors, turbines), long distances, difficult access, and complex layouts. | Requires high dynamic range, effective noise rejection, and reliable detection at distance. |
In practice, many industrial sites combine several of these conditions. The required performance should therefore be evaluated based on the most demanding use case, not the simplest one.
What acoustic cameras are designed for... and where they are not the right tool
Acoustic cameras are powerful tools for locating pressurized gas leaks in industrial environments. They detect the ultrasound generated by turbulent gas flow and allow rapid inspection of large areas from a distance. However, they are not designed for every type of leak detection problem.
8 criteria to evaluate acoustic camera performance in practice
1 Detection performance in real conditions (beyond datasheet specifications)
Detection performance is the primary factor when evaluating an acoustic camera.
In practice, performance is defined by the ability to detect leaks reliably in real industrial environments, not only by sensitivity under ideal conditions.
Industrial sites introduce challenges such as high background noise, multiple simultaneous sources, complex geometries, and varying distances. These factors can significantly impact detection quality.
These aspects can be observed during a demonstration, but do not require a full evaluation campaign. Vendors should be able to demonstrate performance clearly through a combination of:
- documented results in environments similar to your own
- field references or case studies
- demonstrations under realistic conditions
When evaluating solutions, focus on the following points:
Dynamic range: detecting weak leaks near strong sources
In industrial environments, strong ultrasound sources (e.g. large leaks or mechanical noise) can mask weaker ones. This is one of the main reasons why two cameras with similar microphone technologies and microphone count can behave very differently in practice.
This capability is linked to the overall dynamic range of the imaging system, but is rarely described in a meaningful or comparable way in datasheets.
When reviewing vendor demonstrations or recorded examples, look for simple visual indicators:
- Can several leaks be distinguished clearly in the same scene, or do they merge into a single spot?
- Does a weaker leak disappear when a stronger one appears?
- Does the acoustic image show meaningful structure, or does every source appear as the same round blob?
A system that can separate multiple sources and preserve meaningful structure in the acoustic image generally provides more usable information in real inspections.
Absence of artifacts (“ghost” sources)
Acoustic images are generated by signal processing algorithms. Depending on the algorithm design, artificial acoustic spots (“ghosts”) may appear that do not correspond to real sound sources.
These artifacts are typically caused by:
- limitations in the imaging algorithm
- spatially extended sources
- correlated or complex acoustic fields
They do not correspond to real physical sources, but can appear similar to leaks in the acoustic image
What to look for when evaluating solutions:
- Do acoustic spots appear in locations where no physical source is present? For example, in empty areas such as the sky or open space
- Do some acoustic indications have structured shapes (e.g. rings or extended patterns) that do not correspond to a physical source? These can be signs of algorithm-generated artifacts rather than real leaks.
- Is there a clear correlation between acoustic spots and the optical scene? Persistent spots that cannot be linked to any physical element are likely artifacts (note: reflections are not artifacts).
In practice, a reliable system should produce acoustic images where each indication can be clearly linked to a physical source.
Time synchronization between acoustic and optical images
Acoustic cameras combine two data streams, the optical image and the acoustic image
These must be synchronized in time to ensure that the acoustic indication remains correctly aligned with the visual scene during inspection.
In practice, the operator is constantly moving the camera. If synchronization is not precise, the acoustic image may lag behind or shift relative to the optical image making it complicated to identify a faulty component, especially at distance..
What to look for when evaluating solutions:
- Observe how the acoustic indication moves relative to the optical image when the camera is in motion The acoustic spot should remain clearly attached to the source.
Even small misalignments become significant at distance. A system that is not well synchronized may appear usable at close range, but will make leak localization unreliable when inspecting from afar.
Usability and inspection repeatability
Even with strong detection performance, an acoustic camera must be usable in real conditions and produce consistent, repeatable results.
In practice, inspections are performed:
- in complex environments
- over long periods of time
- by operators with varying levels of experience
If the system depends too much on operator technique or manual adjustments, results can vary significantly.
What to look for when evaluating solutions
- Field of view (FOV)
A wide field of view allows the operator to inspect larger areas at once and reduces the risk of missing leaks, especially in complex installations.
- One-handed operation and ergonomics
The camera should be usable with one hand, allowing the operator to maintain stability (e.g. holding a railing) and operate safely in industrial environments.
- Limited reliance on manual settings
Systems requiring manual focus or frequent adjustments increase operator dependency. Automatic or assisted settings improve consistency and reduce user error.
- Clear identification of the leaking component
The acoustic indication should clearly correspond to the optical image, allowing unambiguous identification of the faulty element, even in dense installations.
Systems that rely heavily on user expertise or manual adjustments may perform well in expert hands, but often lead to variability and missed leaks in practice.
Inspection reliability and completeness
A key question for inspection programs is:
How do we know that the inspection was complete, and that no relevant leaks were missed?
In practice, detection conditions vary during an inspection:
- background noise changes
- distance to the source varies
As a result, the ability to detect a given leak is not constant.
Without clear feedback, operators may unknowingly perform inspections under conditions where certain leaks are not detectable, leading to incomplete or inconsistent results.
What to look for when evaluating solutions
- Does the system provide feedback on whether current conditions are sufficient to detect leaks of interest? For example, based on distance and real-time background noise conditions.
- Is the operator guided to adjust their position when conditions are not adequate? The system should help ensure that inspections are performed under suitable conditions.
Why it matters
If detection conditions are not controlled:
- some leaks may be missed
- inspection results may vary depending on how the inspection was performed
- confidence in the results is reduced
Key takeaway
A reliable system should not only detect leaks, but also help ensure that inspections are performed under conditions where leaks are actually detectable.
Leak rate quantification (if required)
Some acoustic cameras provide leak rate estimates (L/h, SCFM, kg/h) based on acoustic measurements. This can support prioritisation of maintenance actions, monitoring of leak evolution over time, estimation of financial losses, and emissions reporting.
However, not all quantification approaches are equivalent. Accurate estimation requires models that take into account: accurate distance between the camera and the leak, gas type, system pressure, and leak geometry.
If these parameters are not properly considered, leak rate estimates can be highly inaccurate, with errors potentially reaching orders of magnitude.
The geometry and size of the acoustic sensor array also play a role. Smaller arrays typically provide lower accuracy of distance estimation, especially at larger distances. Since distance is a key parameter in leak rate estimation, inaccuracies in distance measurement can directly impact the reliability of the quantified leak rate.
When evaluating solutions, it is therefore important to understand which physical parameters are used in the model, how the model has been validated, and whether results are consistent in real industrial conditions.
4.3 Safety and certifications (ATEX, IECEx, UL)
In many plants (refineries, petrochemical sites, gas storage, hydrogen facilities), operators may enter explosive atmospheres while performing inspections.
Key questions:
- Do you have explosive gases in the facility you want to inspect?
- Do you intend to use the camera inside these classified areas?
- Does your company require ATEX, IECEx, UL or other certifications?
Explosion-proof design ensures that the camera cannot become a source of ignition. ATEX and IECEx classify areas into Zones (0, 1, 2) depending on how often explosive gas atmospheres are present, while other systems (like UL Divisions in the USA) use a different classification but similar concepts.
Beyond explosion safety, ultrasound cameras are digitally enabled products. Your company will likely ask:
- How is inspection data stored and transmitted?
- Where are servers located (on-prem, EU, US)?
- Which cyber-security standards does the vendor comply with?
4.5 Ease of use and operator dependency
Even the most powerful ultrasound camera is ineffective if operators struggle to use it in the field.
- One-handed vs two-handed operation: a one-handed camera allows the operator to keep one hand on the railing. One-handed devices are safer and more comfortable when climbing stairs, using ladders, or working in confined spaces.
- Usability with PPE: consider how the device is operated when wearing gloves, helmets, or other protective equipment.
- Reliance on manual settings: manual focus requires the operator to estimate the distance to the leak. Acoustic autofocus automatically adjusts to the sound source distance, removing guesswork and reducing operator dependency.
- Consistency across users: the more settings require manual adjustment, the more results depend on individual operator expertise. Systems that minimise manual configuration produce more consistent results across users.
Lightweight designs reduce fatigue during long inspections.
4.6 Imaging capability: field of view, focus, and detection coverage
- Optical field of view (FOV): a narrow FOV (e.g. 60–70°) may be acceptable for small, controlled areas. For large industrial plants, a wide FOV (around 150°) makes it easier to see more equipment at once, detect several leaks simultaneously, and reduce blind spots and human error.
- Acoustic focus: manual focus requires the operator to estimate the distance to the leak. Acoustic autofocus automatically adjusts to the sound source distance, removing guesswork and helping to avoid accidentally "blurring" leaks out of the acoustic image.
- Lighting: integrated LEDs are extremely helpful in dark or partially lit areas, making the optical image usable for documentation.
4.7 Reporting, data handling, and IT integration
Finding leaks is only half the story. In many organizations, other teams (maintenance, operations, HSE, management) will rely on clear, traceable reporting.
- Photo and video quality: colour images with a wide angle of view help quickly identify the exact leaky element and read asset tags. Embedding gas type and estimated leak rate directly in the image or video makes it "self-explanatory".
- File formats and transfer: standard formats (JPEG, MP4) avoid IT headaches. USB, Wi-Fi or Bluetooth must comply with your IT/security policies. Ideally, the camera won't delete files automatically after transfer.
- Reporting tools: simple workflows to generate reports with images, leak rates and annotations (tags, PNID references, equipment IDs).
- Cybersecurity: how is inspection data stored and transmitted? Where are servers located? Which cyber-security standards does the vendor comply with?
4.8 Calibration, training, and long-term support
- Calibration: opto-acoustic calibration (alignment between optical and acoustic images) over the whole field of view and microphone array calibration are critical for accuracy and comparability over time. Regular calibration may be required by your metrology or quality department, especially if you rely on leak rate quantification.
- Training levels: availability of training material and online courses may be sufficient for simple use cases such as compressed air leak surveys. On-site training is often recommended for complex environments, regulatory applications, or service providers, where consistent and reproducible results are required.
- Expertise of vendor: access to specialized support is particularly important for challenging use cases (noisy environments, leak quantification, emissions reporting). The level of expertise behind the solution can have a significant impact on the quality and reliability of the results. When evaluating vendors, it is worth considering whether support is provided by specialists in acoustic imaging, or as part of a broader portfolio of inspection technologies.
When a low-cost acoustic camera may be sufficient — and when performance becomes critical
Low-cost devices are often suitable when:
- the application is limited to compressed air
- environments are simple and relatively quiet
- no Ex certification is required
- leak rate quantification is not needed
- use is occasional
Higher-performance systems are typically required when:
- leaks affect uptime or safety
- gas losses are costly
- methane / emissions reporting matters
- hazardous areas are involved
- inspections are frequent
- results must be consistent and traceable
In these cases, detection performance, reliability, and usability in real conditions have a direct operational and financial impact.
Detection
Low-cost devices are often suitable
- the application is limited to compressed air
- environments are simple and relatively quiet
- no Ex certification is required
- leak rate quantification is not needed
- use is occasional
Measurement
Higher-performance systems are typically required
- leaks affect uptime or safety
- gas losses are costly
- methane / emissions reporting matters
- hazardous areas are involved
- inspections are frequent
- results must be consistent and traceable
How to evaluate vendor claims: what evidence actually matters
Ask for:
- detection limits with conditions (distance, noise, gas, pressure)
- independent vs internal testing
- real industrial case studies
- methane/process gas performance (not only compressed air)
- repeatability across operators
- validation of leak rate quantification
- demonstration in your environment
A practical checklist to shortlist acoustic cameras
Use this checklist when comparing vendors:
- ❑ Do we need an Ex-certified (ATEX, IECEx, UL) ultrasound camera?
- ❑ Has the detection performance been tested by independent laboratories?
- ❑ Does the camera provide realistic, documented leak rate quantification?
- ❑ Is the device single-handed and light enough for long inspections?
- ❑ What is the optical field of view? Are we comfortable with potential blind spots?
- ❑ Is there acoustic autofocus, and how is dynamic range handled?
- ❑ Does the camera include integrated LEDs for low-light inspections?
- ❑ Are images/videos in standard formats? How are files transferred and stored?
- ❑ What calibration processes and intervals are recommended?
- ❑ What training and support packages are available (online, on-site, refreshers)?
Frequently Asked Questions
Choosing the right acoustic camera is about fit — not just price
The right acoustic camera is not simply the one with the lowest purchase price. It is the one that reliably detects the leaks that matter in your environment, can be used safely and efficiently by your teams, and produces results you can trust for maintenance, safety, and emissions decisions.