8.1 The probe-tube microphone and the REUR

The fundamental problem REM solves is what arrives at the eardrum. The audiogram tells us the patient’s threshold in terms of dB HL — a reference frame that bakes in a standardised average pinna, ear canal, and tympanic membrane impedance (see Lesson 2.2). The hearing-aid coupler gain tells us the device’s amplification in terms of a 2 cc coupler. Neither tells us what the patient’s eardrum actually sees when the device is in the patient’s actual ear.

The probe-tube microphone bridges this gap. It is a small silicone or soft-plastic tube, typically 1 mm outer diameter and 75 mm long, with one end inserted into the patient’s ear canal close to the eardrum and the other end connected to a calibrated microphone. The microphone sees a low-passed version of the SPL at the tube tip (the tube has frequency response losses above ~4 kHz that the instrument corrects for). With the probe in place, all subsequent measurements report SPL at the position of the tube tip — within a few millimetres of the tympanic membrane.

Inserting the probe tube

The clinical procedure for inserting the probe tube is the rate-limiting and skill-determining step of REM. The tube needs to terminate within ~5 mm of the eardrum (closer is better — the standing-wave pattern between the tip and the drum varies most rapidly near the drum). It needs to not contact the eardrum (uncomfortable and risk of perforation in pathological drums). It needs to be stable enough not to move during the test.

The two standard insertion techniques:

• Marked-tube depth method: pre-measure the patient’s ear-canal length (otoscopically — the distance from the canal opening to the visible drum), subtract 5 mm, mark the tube at that length from the tip, and insert until the mark is at the canal opening. Most reliable in adults with normal anatomy.
• Otoscopic-confirmation method: insert the tube under direct otoscopic visualisation, advancing until the tip is visible just short of the drum. More reliable but requires a video otoscope or direct visualisation that not every clinic has.

For pediatric patients, the canal length is shorter (15–20 mm in infants, 25 mm in school-age, 25–30 mm in adults); the probe-tube placement must be correspondingly shorter. Modern REM systems include automatic real-time spectrum analysis that detects standing-wave notches in the probe response and prompts the audiologist to advance the tube if a notch indicates the tip is too far from the drum.

The three baseline measurements

REM produces three baseline real-ear measurements, each with its own clinical meaning:

REUR — real-ear unaided response

Measured with no device in the ear and a calibrated speaker delivering a known signal (typically pink noise or a swept tone) from a position 0.5–1 m from the patient’s ear. The probe records the SPL at the eardrum. The REUR is the patient’s natural ear’s transfer function — the passive amplification provided by the head/pinna/canal resonance (see Hearing 2.4 — The ear canal and 2.5 — HRTF refresher →).

A typical adult REUR has:

• A broad concha resonance at ~2.5 kHz boosting the SPL by ~5–8 dB.
• A sharper canal resonance at ~2.7 kHz boosting by another 12–17 dB.
• Combined: a peak gain of 15–25 dB in the 2.5–4 kHz region, falling off above and below.

The REUR is the reason humans hear high frequencies as well as we do despite the mostly-flat acoustic environment: the head and ear canal collaborate as a passive amplifier exactly in the frequency region where speech-relevant high-frequency consonants live.

REOR — real-ear occluded response

Measured with the eartip in place but the device turned off. The probe tube runs alongside the eartip. The REOR represents what the canal does when the device’s eartip occludes the canal — removing the natural canal resonance, blocking the direct sound path, but introducing the device’s own occlusion effects.

For a fully occluded ear (CIC or closed-fit BTE), the REOR is dramatically reduced from the REUR — the canal resonance peak is lost and replaced by a lossy enclosure with the device sitting in it. For a vented or open-fit ear, the REOR retains some of the REUR character but is altered by the vent dynamics.

The difference between REUR and REOR — the insertion-loss of the unaided ear — defines the headroom the hearing aid needs to make up just to restore natural audibility. For an unvented ITE this insertion loss can be 15 dB at 2–3 kHz, all of which must be recovered before the aid begins to provide additional useful amplification.

REAR — real-ear aided response

Measured with the device in place and turned on, with the same calibrated speaker signal. This is the actual SPL at the eardrum produced by the combination of acoustic input + device gain + canal-acoustics. The REAR is the clinically relevant outcome measurement — it’s what the patient actually hears.

The REAR is then compared to the prescription target (Lesson 8.2) at each audiometric frequency. The difference is the fitting error.

Why the 2 cc coupler is not enough

The manufacturer’s specification of a hearing aid’s gain is measured in a 2 cc coupler (HA-1 or HA-2 type per ANSI S3.7), a rigid metal cavity of 2 cm³ internal volume terminated by a calibrated microphone. The 2 cc coupler approximates an adult average ear-canal-plus-residual-canal volume well enough for design and quality-control purposes, but it differs from a real ear in three important ways:

1. Volume. Real adult ear canals contain 0.5–1.5 mL of air with the hearing aid inserted; the coupler always presents 2 mL. Smaller volumes produce higher SPL for the same volume velocity, so devices typically produce more real-ear output than coupler output, especially at high frequencies where the wavelength approaches the canal dimensions.
2. Termination impedance. The coupler microphone has a fixed (and approximately rigid) acoustic impedance; the human tympanic membrane has a complex frequency-dependent impedance (see Hearing 3.2). The reflection coefficient at the canal end differs, especially below 1 kHz, causing standing-wave differences.
3. Pediatric ears. Infant ears can have canal volumes of 0.2–0.5 mL — a 4 to 10× difference from the 2 cc coupler. Pediatric devices fitted to coupler specs alone can be 10–15 dB louder at the eardrum than the spec implies — a clinically meaningful overshoot.

The RECD (real-ear-to-coupler difference) is the per-frequency dB difference between real-ear SPL and coupler SPL for the same input. RECD is patient-specific and measurable; modern REM systems can either measure RECD directly (placing the probe tube in the ear with the eartip in place and a calibrated signal delivered via the eartip) or estimate it from age and ear-canal dimensions. RECD is the formal device-side correction factor that converts coupler-gain specs into expected real-ear gain.

Coupling RECD to the audiogram

Audiometric thresholds are measured with supra-aural earphones or insert earphones, which deliver their stimulus into the patient’s actual canal. The threshold-in-canal SPL is not the same as the threshold the audiometer reports in dB HL — the audiometer converts using RETSPL (the reference threshold SPL of the calibration coupler). For pediatric patients, where the actual canal volume is much smaller than the calibration coupler, the threshold-in-canal SPL is higher than the dB HL reading would suggest — and the prescription target should reflect the in-canal SPL, not the dB HL value.

This is why DSL v5 in particular uses measured RECD as an input to the target calculation, especially in pediatric fittings. The clinical reasoning is: the audiogram tells us threshold referenced to a standardised ear; the RECD tells us how much that standardised ear differs from this patient’s ear; the prescription target adjusts accordingly to deliver the right SPL to this patient’s eardrum.

Next lesson: the prescription target itself — REIG, NAL-NL2 and DSL v5, and the ±5 dB match criterion that defines a verified fitting.