Bernafon Juna 9 Programming

The amplification at three different input levels is standard. Simple digital aids only have two input levels. Three input levels mean compression with two kneepoints.

For music, the levels would need to be higher, though. You have almost no control in the area above 80 dB (which doesn´t matter for speech, above 80 dB is yelling, but it matters much for music).

Using no compression is bad. Consider an amplification of 20 db at 3 kHz. If your mpo is set at 105 dB, all input-signals above 85 dB will make your aid hit hard agains the mpo (which means: compression ratio infinity). The rule of thumb is: Use less amplification to allow for less compression. The values used for the live-music-plus program are a good starting point.

In my experience, compression ratio below 1.5 sounds pretty good for music. I have tried a program with no compression. Drums sound pretty hard then, but the sweet spot for listening is extremely narrow then.

Looking forward to your report!

Several years ago, when I got my first audiogram, I was shocked and confused. When I saw the amount of high frequency loss I thought there had to be a mistake…I was still hearing the cymbols, the brushes on the drums, and pretty much all the high frequency overtones in the instruments. Then a light bulb lit up in my head. I rigged up a way to sum pink noise with pure tones and do my own audiogram for the threshold of hearing just above various levels of pink noise background. With the noise up around 90 dB, my audiogram was much flatter, explaining why I was hearing the high frequencies in concert level music, while losing them at much lower levels. Then I remembered Fletcher-Munson and it all became clear.

Certainly level-dependent compression and frequency response is important in hearing aid programs for music. It would be good to know exactly how both are implemented in current high-end aids that claim improved performance for live music…or recorded music for that matter. Just my gut feel…I think Bernafon is on the right track with their single-channel approach (Chasin agrees) which preserves the relative fundamental/harmonic balances in instrument spectra. Certainly, they sound better on music than all the other digital aids I have tried, but that is partly because I am able to tailor their response based on what I hear, which no audiologist can do.

I will try to adjust the programming to more closely emulate the analog K-Amp in the next day or two and report back. I will have a chance to listen to live music Thursday night, so that should be a good test.


There is no second microphone, the output is just compared with the normal input to evaluate whether there is likely to be phase/harmonic issues in the canal. Some tuning occurs within the output stage to smooth the overall combined signal.

Manufacturers are aware that this effect occurs - especially as improved feedback control has increased the number of open fittings that allow a greater proportion of the signal to be heard directly through the venting.

Very interesting pcasper! I am about to enter the realm of self programming, and as an audio engineer reproduction of music is an important requirement. I will have a lot of fun finding what works best for me.

Um Bongo: I confess I still do not understand how the hearing aid output can be compared with the normal input and detect any comb-filtered and distorted composite sound field occurring at the eardrum. Seems to me there would have to be another microphone at the eardrum sensing what the ear is hearing for comparison with the input, resulting in speaker output with complementary distortion. I apologize for being so dense…perhaps you can provide more details.

GaryM: Welcome to the club for pursuit of high quality music reproduction with digital hearing aids! It is a difficult and exasperating exercise, but success can be achieved, as I am about to report.

Musician 72: I have reprogrammed my Bernafon Live Music program to emulate the level-dependent frequency response of the K-Amp , with complete disregard for my audiogram except for one item. I elevated the overall volume of the left ear to roughly equal the right ear volume. So far, I have had little time to do serious listening…BUT I HAVE NOT STOPPED SMILING!! Everything I have listened to at default aid volume and up to 95dBa SPL (slow) has been just wonderful. As an engineer I am leery of forming conclusions based on first impressions, but I cannot help feeling that this has been a breakthrough for me. Tonight I will get some exposure to fairly loud live music and we shall see how that sounds. Then, over the next few days I intend to listen seriously to several reference recordings I have used for sound system evaluation for years, and how they should sound is etched in my brain. I will use my theater sound system as the source, which is world class (it replaced my previous German Avantgarde Hornspeaker system, which you may be familiar with). Assuming success with that exercise, I will write up a report detailing exactly how I programmed the Bernafon aids, why I did what I did, what high quality recordings I used for evaluation, observations of interest, and numerous references to K-Amp design information. I will put the report online and post a link to it for anyone interested. It will probably be early next week before I can get all that done, but for now I am a happy (but exhausted) camper!


The comparison you mention to Um Bongo is done. It is called REM Real Ear Measurement. But it is done by the fitter. They insert a listening tube that compares the output to the prescription. It takes into account the shape of the canal etc. That makes what you should be hearing the same as the prescription.

OK, let me see if I understand how a REM measurement is done. A microscope probe is inserted in the ear canal, along side the aid speaker (if it is a RITE aid). A loudspeaker in front of the patient is used to create an acoustic swept frequency tone, which enters the open dome direct path as well as the time delayed aid path. This results in a coherent interference pattern at the ear drum, which the probe microphone detects. If the resulting REM frequency response graph is not smoothed, the comb filtering distortion will be visible in the amplitude crossover zone. If this is correct, how is the aid programming modified to create a complementary pattern to null the interference pattern, or is it? I’m treading in areas where I have little knowledge here, but I am curious and eager to learn.

My audiologist showed me the trace of the probe measurement on the fitting graph and I’m pretty sure she used it to make some steady state adjustments in the amplitude response of the correction. I doubt there is the ability or processing power to do Realtime dsp to deal with any comb filtering. I guess the in canal aids are going to be better at this.

1 Like

Musician 72,

I have written my Bernafon reprogramming report, but before posting an open link to it online, I would like to request you to have a look at it first. If you could send me your email address to, I will send you the link.

Thank you, Paul

No, there’s no R.E.M. being done, that’s for the fitting. The aid sets itself during the ‘feedback manager’ sweep. This allows an accurate in-situ assessment of the canal resonance to be made. It’s just a matter of maths, long term averaging and assumptions to work out where the issues are likely to occur. If you also apply the before and after processing logic that I was talking about above you can work out the actual uSec delay and phase effects at different frequencies. So you can tell whether the signal is going to be additive, subtractive or cancelling relative to the incoming signal and apply the appropriate filter to the output.

If done properly the signal is ‘smoothed’ back to the original waveform with the desired amplification.

**Um Bongo,

Thank you for that explanation, now it’s starting to make sense. So the relative phase is predicted based on the known aid delay at any given frequency and the delay is adjusted for neither peaking nor cancellation. I assume the process does not have to have the feedback cancellation turned on after the initial sweep. It should be turned off in the all hearing aid music programs. Also, I expect it is only marginally effective with significant direct path amplitude and frequency changes during the several millisecond aid delay interval.

I wonder how many aid models include this process. Certainly my Rexton Quintras do not…the interference pattern in the crossover zone looks terrible. I’ll have to test the Bernafons when I get a chance.

Still waiting on Musician 72 to critique my Bernafon programming report. I don’t want to post it and find I have to make a series of corrections based on imperfect knowledge of how the software works in the Live Music program.


[quote=“pcasper, post:30, topic:26049”] I assume the process does not have to have the feedback cancellation turned on after the initial sweep. It should be turned off in the all hearing aid music programs. Also, I expect it is only marginally effective with significant direct path amplitude and frequency changes during the several millisecond aid delay interval

You assume correctly. The adjustment forms part of the gain prescription, not the feedback manager. The efficacy is pretty high and instantaneous, the calc has already been done, it’s just a case of modifying the gain appropriately vs. the input level and frequency, which the aid is sampling a few thousand times a second anyway.

Hello All,

I was hoping to get a critique from Musician 72 before I exposed my Bernafon Juna 9 music reprogramming report to the forum, but I have received no response in several days, so I assume he is occupied with more important matters at the moment. With some reluctance, since my report has not had an ‘expert review’, here is the link to it for anyone curious enough to have a read:

I have thick skin, so all critical and constructive comments are welcome. I still have a lot to learn. I am pleased to report that after some 2 weeks listening to mostly recorded music, with some live thrown in, I am still very happy with the results. My analog K-Amp aids are now officially retired. The only thing I still feel needs adjustment is the left ear volume, which needs to go down 2 or 3 dB .

It is interesting to experience the level dependent frequency response which is maintained in the Bernafon Live Music program. With the music above 80 dB average, the aids have virtually no effect (the Bernafons have a nice mute feature for testing), and none is needed. At lower levels the treble frequencies progressively increase down to about 50 dB, then stay constant. So I hear the treble clearly at all levels, which is wonderful. The Fletcher-Munson effect still operates, even with degraded hearing.

I hope someone benefits from this information, and is encouraged to pursue self-programming. Alternately, if you have an agreeable audiologist willing to experiment, it would be extremely interesting to see if K-Amp emulation programming works for anyone else, and with other high dynamic range aids such as the Widex Dreams.


Link doesn’t work for me. Don’t know about others.

Link points to Gmail something-or-other requiring viewer to logon.

@pcasper, Take a look at this thread. It might give you another way to upload your info here.

OK, I have uploaded the report to the forum. You can find it here:

Reprogramming My Bernafon Juna 9 Aids for Music.pdf (1.1 MB)

I hope it works this time.


I just tried it…I got two copies…anyone else get the same result?


I tried it and it worked like a charm (one copy) Great writeup! I was really impressed.

Did you have any luck with self programming the Quintra 2cs? I’m trying to determine if it’s possible or not.


No, to tell the truth I haven’t even got around to trying. I’ve been happy with my Bernafons, so I have put the Rextons on the back burner. I WILL reprogram them at some point, just not sure when. I will post the results when I do.