No matter what my audi does, my acoustic and classical guitars always sound awful when I get home after a session with her. I have done some reading (references 1, 2, 3, 4, 5) and devised a method to improve the musical experience. As a layman in the field, I would like to hear your comments and suggestions on things I got wrong and ways to improve this little experiment:
Spend some time experimenting with domes and molds. Find the perfect acoustic setup for you. My ear canals are too narrow for custom molds, so I choose domes that give me both good audibility and physical comfort.
Do a REM session or in-situ audiometry. There is a rich discussion on “REM vs in-situ audiometry” here (Real ear measurement vs. in-situ audiometry). Either way, it is important to input your conductive results, so you get the most optimized prescription.
Create a music program with a fitting formula you feel more comfortable with. If needed, test all available to you. I’ve settled with DSL, though I don’t really like it for speech comprehension.
Disable all digital features you can in the music program: feedback manager, digital processing, voice booster, etc. Leave this program as much closer to an “analog” program as possible.
Lower the target soft gain until you get a linear compression rate (1). The gain curve should have about the same general shape as the one you started with. Adjust for volume. I do that playing my guitars.
Get your “eq” right (if that is important to you): when playing your instrument, it is easy to get sidetracked by what you want it to sound rather than what it really sounds. This has been the most difficult part to me, by far. There are multiple threads and posts in this forum addressing the many ways things can go south between your hearing aid and your brain. Here is what I did: a. I created an audio to test frequencies using the “equal-loudness contour standard” (reference here)). Edit 2 (new frequencies): The frequencies in the audio are (Hz): 125, 250, 500, 625, 750, 1000, 1250, 1500, 1700, 2000, 2300, 2600, 3000, 3500, 4000, 4500, 5000, 5500, 6000, 7000, and 8000. Edit (new file): Here is the link for the audio. b. Play the audio and adjust the volume in your speaker so that 1 kHz plays at edit (new volume) 65 dB. c. Check for your perceived loudness across all frequencies. Your music program is optimized if this perception is even across the board. Otherwise, adjust the gains in your program until you perceive all frequencies at the same volume.
Throw back some compression accordingly. Edit:: Here Oticon recommends to: 1) “keep the compression ratio as low as possible while ensuring sounds are audible to the listener”; 2) “most compression is done between the first two knee point levels (soft and moderate)”; and 3) “The compression ratio is kept to 1.0 between moderate and loud knee points”. I have a purely conductive loss in my LE, so none there. Things are a bit more complicated with my RE, but I still manage to have compression as low as possible there.
Play music and hear how it sounds. I am happier with my music program now, though I still feel the low frequencies on my classical guitars are not quite there yet.
Getting the music program right is hard. I’d appreciate any feedback/comments, especially on step 6.
Do not input your conductive results unless you have a truly condictive or mixed loss. Sounds like you do, but lot of people who have sensorineural hearing loss might have little variations between the bone line and air line that don’t really represent conductive components rather than just the increased variability of bone conduction testing.
Clever. I assume you just adjusted the dB SPL of the presented tone relative to the equal loudness curve. But you linked to an equal loudness contour for 40 dB SPL and you are presenting at 65? Are you sure your values are correct, because it’s not a linear shift in perceived loudness across frequencies when you increase the SPL. I would think you’d need to present at 40 dB SPL or get values from a 65 curve. Also, those highs sound substantially louder to my essentially normal-hearing ears from about 1250 up (but my current speakers are probably also just crap–the 8000 Hz tone breaks the BT connection ).
I can’t see your audiogram today for whatever reason, but you really can’t have custom tips, even if they are embedded? Always a bit of a trade-off managing acoustic coupling for speech versus music, but closing the canal more will typically improve the lows. Have you already checked the fit of the oticon griptip sleeves versus their domes?
Thanks for taking the time to read and give me feedback. You can change the equal loudness curve to 65 dB, there is a box in that link where you input the level of your choice. I used the values it yields to create the audio.
The empirical assessment of the perceived loudness across the frequency range is by far the most challenging step. Acoustic is a very complicated field, hard to make sense of all the technical nuances. I would appreciate any help on that front, very much
@tenkan: I am able to do my own setup. However programming for music is a bit tricky. With the right setup you can make a cheap guitar sound much better than it actually sounds. I am having a hard time tuning my “eq”, if you know what I mean.
I am trying to figure out an external reference to rely on since I cannot count with my own ears for that. After decades of hearing loss I just don’t know how exactly musical instruments sound anymore.
I own pretty decent guitars. But too often I get an acoustic sounding “right” just to hear the classical sounding awful, and vice versa. I think the way out is getting the music program “flat”.
In this regard, I think the audio I made helped. However, I am still not sure I got everything right or if this is a reasonable way to tweak the music program.
I recently had the vent in my right (worst) ear plugged with a removable plug (easy). The bass is boosted a lot and the sound is better overall. However the same plug in my left aid makes things worse. I have custom mods, obviously.
I can place and remove the plug easily at home. It’s possible to re program the aids if one likes this option, but I haven’t. I play classical guitar. This is all in terms of hearing for live music.
@jeffrey: Thanks for your comment. I am getting a good volume of low frequencies. The problem has been the quality of the sound. I feel the bass is sort of “plasticky”, for lack of a better term.
I think it has to do with the amount of compression I need for my right ear (the one I did a stapedectomy a couple of decades ago). It is a game of finding the sweet spot between tightness of sound and comfort.
For a good music program, compression has to be eliminated or at least greatly dialed back. It’s a dedicated program, so I just switch back to my compressed speech program for usual circumstances. Maybe you need to have a lot of compression due to the nature of your loss?
@jeffrey: I had to change the dome in my right ear one more time yesterday. That soft pain that builds up over time was getting unbearable by the evenings. Anyways, with the new dome I also fine tuned the program so it is a bit better now. As you mentioned, it is a dedicated program with the minimum digital processing I was allowed to have by the fitting software. I do need compression on my right ear, but usually not higher than a factor of 2. I have no compression in my left ear, in this one I have a purely conductive loss.
Quick report on results: so far the best music program I ever had. Actually, the two best because I tweaked my original one based on DSL and created a new based on the standard MyMusic (Oticon More). Both programs sound really good and a bit similar, though I hear a tinny difference: DSL sounds more raw, MyMusic sounds like some digital processing is still going on. Regardless, I am very pleased with the more natural resonance of my acoustic and classical guitars.
Below are screenshots of the MyMusic program before and after the “EQ” I did with the new audio I created. Unfortunately my right dome won’t allow me to have all the gain I need in the high frequencies (due to feedback). It is good enough the way it is. Before:
Edit: Note that there is minimum to no compression in my program. That is what I feel ideal for live music. When I decide between DSL and MyMusic, I will have another program with some compression for streamed music.
If your music program sounds good then that’s all that really matters. But something in that audio still sounds off to me. The higher frequencies are loud. What measure does audacity use for volume?
But again, I’m traveling and I just listened on my phone speaker, which may be so poor as to not offer useful assessment.
Audacity inputs amplitude between 0 and 1. That is why I need to normalize things, which is just another step adding up to the overall uncertainty of this little “experiment”. Actually, there are a few variables that I just can’t really control, so I do the best approximation.
One of these variable is the calibration of the volume: the audio will only work when 1kHz is played at 60 dB. However, even doing so, who knows the frequency responses of the speakers? I guess I would have to play the audio in a high-end audio monitor for best results. I don’t own one, so that is not happening .
Edit: Audio now uses an EQ instead of normalization between 0 and 1, as described above.
I think this post describes a qualitative tool to help with the fine tuning of the music program. It seems it worked for me but that might be as well just confirmation bias… lol
I just started to read this thread. Hopefully, it will me, as an audiophile, get a good program going with my Widex Moment 440’s.
I haven’t really delved into the thread yet, but in terms of the audio test tones…would just using an SPL meter with your not-so-great home speakers allow you to be sure of the loudness accuracy? I shouldn’t even really mention this yet without really reading the thread, but…
Also, with my HA’s, as far as I can tell, there is no way for me get the Widex program that the audiologist uses without subscribing and purchasing it? I believe it has to connect to Widex for it to be usable. Is this correct?
That is a fair point. I am aware that with the gear I own I am getting just a passable approximation, at most. However, that seemed to be good enough for what I needed. My guitars sound great with my new music program, which was my initial goal. For other musical purposes I still need to figure out a compression rationale, but that I will do when I have the time.
Sounds like you’d be a good candidate for DIY, it’s not that hard and all you need is the Noahlink wireless and Widex compass software, check out a couple of these websites for help in setting your HAs for music.
I decided to ditch the MyMusic program. My tweaked DSL music program is superior (I feel like my guitars resonate “right” with it).
I played my new audio in a decent sound system (mostly flat frequency response) and tweaked my DSL music program a bit more. I’m very, very happy with it.
).
.