Commissioning and verification: proof that the system works

Design software is very good at physics and knows nothing about your building. A model assumes the metal loss someone measured or estimated, wire exactly where it was drawn, and a clean audio feed. The room adds rebar nobody drew, a run moved a foot by a flooring crew, a driver setting nobody wrote down. A predicted coverage map, however good, doesn’t prove an installation passes. Measurements in the finished room do.

Commissioning is part of every hearing loop we design and install. We also commission loops other firms installed: for owners and general contractors before final payment, for integrators before handover, and for consultants and public bodies who need a record for their files.

What IEC 60118-4 measures, in plain words

IEC 60118-4 is the international performance standard for hearing loops. It asks three questions about every listening position: is the signal strong enough, is it the same everywhere, and does speech come through with its full range? It also asks whether the room is quiet enough, magnetically, when the loop is off.

The 2017 edition of ICC A117.1, the accessibility standard behind many building codes, references IEC 60118-4 for induction loops; check which edition your jurisdiction has adopted. Owners and consultants often write IEC 60118-4 into specifications too, which makes the measurements, not the equipment list, the thing that matters at handover. Here’s a guide specification that does that.

How we measure

We measure on a grid across the whole listening area, at the height listeners’ heads will be, with the test signal coming from one device at the rack and the meter on another in the room. Then we set the driver so the average reading sits at 0 dB, check the response and the noise, and listen to the real microphones.

1. Check the system against the design. Driver model, settings, feed and wiring are confirmed before any reading is trusted. On phased arrays we test each channel alone, then both together.
2. Measure background noise with the loop off, walking the area with the meter and noting any sources.
3. Play the reference from one device, measure on another. A generator at the rack feeds a steady 1 kHz tone into the driver input, the way program audio arrives, while the meter goes from seat to seat. We never generate and measure on the same device.
4. Read the grid at the design listening height, covering the seating, aisles, front and back, including the highest and lowest readings. The outer 2 ft next to the wire stays out of the average.
5. Set the level. We scale the driver so the average reads 0 dB. If the average reads −3 dB with 2 A in the loop, the driver goes to about 2.83 A, because 3 dB is a factor of about 1.41 in current.
6. Check frequency response at 100 Hz, 1 kHz and 5 kHz across representative seats, and set the driver’s metal-loss correction, a treble boost, where metal has taken the highs.
7. Check spill at any boundary the project names.
8. Listen. A loop receiver in T mode, with speech from the actual microphones: no hum, no distortion, no missing microphones.

Our instruments: a wide range of professional test equipment, including calibrated field-strength meters, with frequency-response analysis referenced at 1 kHz, and for Auracast, a 2.4 GHz spectrum analyzer, venue Auracast receivers and a Bluetooth LE Audio test receiver that reads a broadcast’s settings and lost packets. Every report names the instrument and its serial number. Corrections replace earlier readings, so the record holds one final version of the truth.

What the commissioning report contains

A commissioning report is the record you keep: what was installed, how it was set, where each reading was taken, what it read, and whether it met IEC 60118-4. It’s signed by the person who took the measurements. Below is a sample summary page with illustrative values.

Every report covers the same ground:

1. System description: venue type, room, layout, install method and the design listening height.
2. Equipment and settings: driver model, input gain, loop current and metal-loss correction, written down so anyone can restore them.
3. Measurement map: every point on the floor plan, with its reading.
4. Results table: field strength, variation across the area, frequency response, background noise and spill, each against its criterion.
5. Frequency response plots at representative seats.
6. Listening check: what we heard through a receiver, from the real microphones.
7. A verdict in plain words, listing every position outside the window and anything else that fell short, and why.
8. Recommendations, ranked by value.
9. The date, the instrument and its serial number, and the signature of the person who measured.

With it come as-built drawings showing every wire position, a one-page guide for whoever runs the sound desk, and the Auracast walk-test record where there’s Auracast. Together they mean any qualified installer can service the system later.

Auracast™ broadcast audio verification: walk-tested and documented

Auracast has no installed-performance standard like IEC 60118-4 yet, so we verify it the way people use it. We read the transmitter’s actual on-air settings, walk the room joining the broadcast at marked listening positions, test with a hearing aid, a phone and a loan receiver, and listen for dropouts and delay. Then we write it all down.

• On-air settings. Stream quality, retransmissions, presentation delay, power, broadcast name and any encryption. Most hearing aids join only Standard Quality streams (16 or 24 kHz), so a broadcast that offers only 48 kHz can lock them out.
• Walk test. We join at numbered listening positions, including the edges and the back. A full room is harder than an empty one: one person between listener and transmitter costs about 10 dB, two about 15 dB, so coverage is judged with that margin in mind.
• Audio, not just signal. Good signal strength doesn’t mean good audio. At each stop we check that the stream syncs, how long joining takes, whether it drops out, and how it sounds.
• Join tests with an Auracast hearing aid through its app, a compatible phone, and a loan receiver with a neckloop.
• Delay and level. Auracast runs a few tens of milliseconds, depending on the settings behind the room, so we measure the delay, confirm it’s fed before any loudspeaker delay, and check it sits at a sensible level.
• Names and signs. A clear broadcast name, join cards and signage, and the receivers and neckloops the ADA calls for (see the ADA receiver requirements).

The handover record lists every stop, what joined, what we heard and the final settings. More on how we design and install Auracast systems.

Loop and Auracast together: one record

For most rooms our default is a loop and Auracast fed from the same assistive listening mix. We commission the two together and document them in one record: IEC 60118-4 results for the loop, the walk test for Auracast, and the shared settings that tie them together. We check that both carry the same mix, that Auracast is fed before any loudspeaker delay, and that levels match, so whichever system a listener uses, they hear the same service.

Predicted vs measured

Every design we do starts as a model in Hearing Loop Designer or BroadcastPlan Pro, the tools David built. Commissioning checks the model against the room. Our reports show both, labeled: the prediction as a model, the readings as field measurements with a date and an instrument.

When the two differ, the report says why: metal that wasn’t on the drawings, a wire run that moved, furniture or a crowd in an Auracast path. That’s why we treat predictions as a starting point and never as proof. BroadcastPlan Pro’s radio model is planning-grade, so we walk-test every Auracast design on site and judge coverage with a full room in mind. Read more about Hearing Loop Designer and BroadcastPlan Pro.

For integrators and consultants

We commission loops on other firms’ projects: for integrators before handover, and for consultants, owners and general contractors before final payment. You get the readings, the report and a plain account of anything that needs fixing. Your client relationship stays yours.

If a loop falls short, the report shows where and by how much, what we think the cause is, and what would fix it, and we’re glad to measure again after corrections. Commissioning a loop someone else installed is from $950 per room, plus travel. See how we work with integrators, consultants and architects.

Re-verification: once a year, and after building work

Loops rarely drift on their own; buildings change around them. New flooring, a renovated ceiling, new lighting, added steel or a replaced mixer can all move the results. We recommend a check every year and after any renovation, and a full re-commissioning when the building or the sound system changes substantially.

The yearly check is our Loop Health Check visit: spot measurements against IEC 60118-4 criteria at representative seats, a look at the driver and feed, and a short report. It isn’t a full commissioning, and we say so. If something has stopped working, start with how we diagnose and repair loops.