Setting a Real-Ear Target for Hearing in High-Level Background Noise

A few years ago audiologists spent a lot of time arguing about which real-ear target was best. Now, we don’t usually discuss targets any more, as most of us rely on the manufacturer’s software to give us an estimate of the sound produced by the hearing aid. (We recognize that this is a mistake, but we do it anyway.)

We do still tend to use real-ear targets when we run into difficult patients, people who are unable to hear well with their hearing aids.

Today I want to discuss an unusual target. It is a target I use when I’m programming hearing aids for use in an especially difficult listening situation: high-level background noise. This is a unique idea, I believe, but one that has worked well for my patients. The basic formula for this concept is: REUR + CHBTE + a little. Allow me to explain.



The real-ear unaided response (REUR) is the natural boost your open ear gives your hearing in the higher frequencies. When you do an REUR measurement you see little gain in the lower frequencies and about an 18-dB extra “boost” in the higher frequencies near 3000 Hz. When you put your cupped hand behind your ear (CHBTE = cupped hand behind the ear) you add 5-10 dB to the REUR.

I have taught many continuing education classes on this topic, and whenever I give one I like to include a demonstration. I tell all the class participants to start talking simultaneously to the person next to them. Then, while they are talking, I have them switch between two conditions: no cupped hands behind their ears, and both cupped hands behind both ears. The class members all agree they hear much better, much more clearly when they cup their hands behind their ears.

Now think about this idea for a second. We have a technique we all know produces marked improvement in word understanding in loud background noise. We just need to incorporate this idea into our hearing aid fittings.

With any real-ear system, you can easily measure the CHBTE response curve. This curve provides an excellent starting point, a guide for your hearing in noise listening program. I prefer the word “guide” to “target” because this response is not the best curve for the patient. Rather, it is a good place to start. You can usually make the fitting better by adding “a little” more sound to this approach depending on: (1) the degree of hearing loss, (2) the sound needed to produce good word understanding, and (3) the amount of sound the patient is comfortable listening to.

This “little” amount more sound may be 5-10 dB if the patient does not have a severe hearing loss, or it may be more than 5-10 dB if the patient’s hearing loss is substantial. An important note: I usually give patients a remote control (or volume wheel) so they can adjust the level of the sound. That way, if a patient needs more sound, he or she can simply turn up the volume without need for any technical adjustments in AGC, bandwidth, or frequency response.



This idea needs to be modified depending on the patient’s degree of hearing loss. If we list hearing thresholds at the standard five test frequencies (250, 500, 1000, 2000, and 4000 Hz) consider the following six patients:


Patient #1: 15, 20, 35, 55, 65

Patient #2: 30, 45, 60, 70, 80

Patient #3: 65, 55, 50, 55, 65

Patient #4: 80, 65, 55, 50, 45

Patient #5: 80, 85, 95, 105, nr

Patient #6: 45, 75, 105, nr, nr


Patients 1-4 will do well with the “REUR + CHBTE + a little” target discussed above. Patient 5 and 6 will not. They hear so little in the higher frequencies that they cannot take advantage of amplification in that zone.



One last point. In creating this “target” I am assuming the patient is using this strategy in a noisy restaurant seated directly across from the person they want to hear. The speaker-listener distance is 4-6 feet maximum. You can’t use this idea and give a patient great hearing at long distance.


1 Comment

  1. From practical experience of CHBTE and adjusting high-frequency peaking in ALDs, I agree with the views expressed.

Leave a Reply

Your email address will not be published.