Hearing EQ

[Tom C]

This work is absolutely outstanding. Well done!
I have an HTC10 phone I got a couple years ago that his a built-in function to test and calibrate the EQ for the listeners hearing and earphones. Different profiles can be saved for use with different headsets. I really like the function. Too bad HTC is struggling so.

 

[pozz]

It looks like OSHA only requires a noise of floor of 40dB for occupational testing.

OSHA works on the principle that hearing loss should occur only to the degree that a face-to-face conversation can be held at a reasonable volume at a distance of 1m. Nothing about hearing what's going on in the next room or listening to music.

 

[JoachimStrobel]

So a parametric EQ boost is not what's happening. The image below shows graphs for hearing loss of 0dB, 20dB, 40dB and 60dB HL.

View attachment 37760
With age, the effects are most noticeable at the threshold of hearing for the quietest sounds and with declining sensitivity to high frequencies (both of which are part of typical audiometric tests). But the changes are more fundamental, and have to do with perception of loudness overall (and this kind of testing is done only in research settings).

What's experienced is level-dependent sensitivity changes. As you get older the filter characteristics of the ear change and broaden. Because of this, more frequencies will fall into the same critical band and so certain ranges will feel louder than before. Once mapped, it will appear as if the equal-loudness curves have been compressed. The SPL-to-phon ratio is what's being affected.

Briefly, if you play bandpassed noise which stays within a single critical band you will experience a specific loudness level. You can broaden or narrow the bandwidth of the noise and compensate the amplitude per frequency such the overall power level and consequently the perception of loudness will remain the same. However, once the noise bandwidth is broadened beyond the critical band, even after compensating for amplitude, you will experience a subjective increase in loudness.

So with playback current hardware applies the same level of gain to each frequency without level compensation. Given the compression of loudness curves, it means that the dynamic changes in music are much too broad. This means that older folks may prefer more compressed music given that the level changes will stay within a more narrow band on average. This may also explain the preference for tube electronics (not because of added distortion, which can affect loudness perception as well, particularly for the highs) but because tubes dynamically compress signals.

@JoachimStrobel is entirely right when he says that the industry should move toward creating equipment that is able to compensate for subjective loudness at different volume levels, much of which will include modifying existing parts to make sure they are rugged enough to handle individual perceptual requirements.

What this all translates to is that you'll need greater overall boost at low levels to achieve the same relative loudness across the spectrum and experience relatively normal hearing at louder levels, although the distance between what constitutes "low" and "normal" has shrunk.

What you are writing agrees with the loudness curves that I found for the elder ones (plotted in my post). I saw more datA behind paywalls. Loudness perception versus age seems to be the most important data sets to get. Threshold measurements seem not to be so difficult to do, loudness measurements are.

 

[JoachimStrobel]

Mine takes my curve preference, and creates a static filter.

The FR is the same for volume changes at the signal level.

I have a Dynamic EQ available in a DEQ2496, but didn't find it useful. Could try again.



I set the DRC up using 80dB tone (pretty loud), and typically play music that registers (LEQ) a few dB below to a few dB above that average level, yielding 95 to 105dB peak.

View attachment 37778View attachment 37780


That, by way of "it just turns out that way", puts me in the "flat" zone of Fletcher Munson (blue), though I seem to be lacking per ISO (red).

I tried imitating the ISO curve and didn't like what I heard with music.

It may be because the bass (player's level) is already adjusted to be heard within the recording by the creators - just my impression.

I don't find the bass lacking at lower levels, and besides, wouldn't that would be a natural consequence of listening at a lower level?

I don't attempt to make any adjustment for my hearing loss, as it is fruitless. My deficiency is more like brickwall than rolloff.

---

Sounds natural to me, nobody complains, so, there you have it.

I could reproduce the ISO low frequency loudness curves easily and across multiple persons. The Fletcher curves did not fit there at all. For the high frequency part, I struggle with the hump which implies that 15 kHz is better audible than 10 kHz, strange. I saw that when tuning the head in a certain direction and keeping it there, but only then. The Stevens curve from 1957 seem good for high frequency, but it is no threshold curve.
The ISO curves are made with data from 20-30 year old people - I guess that is not majority around here.

 

[JoachimStrobel]

Interesting. Still, the DRC would have to be able to calibrated according to your individual loudness perception (which is uncorrelated to electric/digital signal level). It becomes kind of like EQing headphones at that point.

H(T)L is probably hearing (threshold) level. 0dB HL is different for each frequency, and is set according to the lowest curve of the well-known equal loudness curves, which are in dB SPL. 40dB HL is considered moderate hearing loss. If the audiometric curve is shaped like this the hearing loss is noise-induced:

View attachment 37775
Edit: I plucked that chart from Google but didn't notice at first that bone conduction (<, [ ) is similar to air conduction (X, O). This means that hearing loss is sensorineural, which happens with exposure to sudden extremely loud sounds. This particular audiogram was that of a patient who served in the infantry: https://www.aafp.org/afp/2013/0101/p41.html

I should have said that long term noise-induced hearing loss will show a clearer mismatch between air and bone conduction with that sort of shape.

Trouble is, that EQ boosts of more than 10 dB are tough on the equipment, this limits the use of a DSP grade EQ with common amplifiers and speakers.
With Ear-EQ in mind, it might be worth investing in heavy load amplifiers, and test speakers for their ability to boost frequencies above 3 kHz by up to 20 dB, above 10khz by 40 dB. A double-tweeter may need to complement the double bass. Or may be an Elderly-add-on active tweeter?

 

[MC_RME]

Just replace the tweeter (typical SPL 90 dB 1W/1m) with a Beyma CP21/F. Will blow free your ears at 105 dB.

 

[JoachimStrobel]

I finally got around and configured an ADC feeding my turntable via HDMI into my NanoAvr‘s Dirac processor. I un-dusted some mid 70s Direct to disc cuts which I heard mainly with young ears. I had listened to them every now and then during past years but was missing the wow factor of younger days. Dialing in my hearing loss target curve brought it back. A lot of nostalgia mixed with truly awesome sound like I experienced last 30 years ago. The current weakness is my cheap ADC, I am working on a Raspi 4 based upgrade aiming at adding corrections after the phono pre-amp too.

 

[JoachimStrobel]

I finally got a Hifiberry DAC ADC for Rasp4 to work with Volumio and feed the digitized vinyl via HDMI into my NanoAvr’s Dirac. The sound is now much better than with my cheap Amazon ADC.

 

[gene_stl]

I would mention something that happened to me at least twenty years ago. I was visiting an audio repair shop where I used to work. The proprietor and I are about the same age. We were 45 or 50 then. In his reception area he usually had equipment for sale or consignment on display.

On this particular day there were a pair of pretty good speakers and a pretty good power amp. I don't remember any brands. For some reason there also was an audio oscillator sitting there and we decided to play "How bad are our 50 year old , rock an roll listened to ears??"

The initial results were TERRIBLE with neither of us hearing much above 5 kHz. He was laughing about it, but I was really upset, not thinking of myself as hearing impaired. Then completely by accident my head moved in front of the tweeter. It was emitting high frequencies in a laser like beam. It was not a large tweeter either. So we rechecked ourselves and I could easily hear up to 19kHz. Subsequent professional tests have confirmed that my hearing is pretty good for an old Geez especially on the right side. But my point is that if you are using speakers to test make sure you are "on the beam" which may be more laser like than you expect.

I do like the idea of extra EQ for ones own hearing when listening alone. Nice work.

 

[Hipper]

I'm a bit simple so what I did was use test tones to check on my hearing abilities. At 60 or so I found when listening through my speakers with both ears that I was down from 4kHz upwards and couldn't hear anything above 10kHz. Furthermore when listening on headphones I discovered one ear couldn't hear 7kHz at all.

At first I tried (using my Behringer DEQ2496) to increase the relevant frequencies accordingly. It seemed to work to a degree (notably more detailed percussion) but there were penalties, such as increased sibilance and fatigue.

I also thought I'd cut off all sounds above 10kHz and I could hear no obvious difference with that.

My latest EQ arrangement just sorts out the bass region and I've left the high frequencies unadjusted. This results in satisfactory percussion.

Another big factor is that I've made efforts to reduce ambient noise by installing a third piece of glazing. This knocks about 6dB of noise off the 100 - 1000Hz section and maybe a couple of dB above 1000kHz.

I've also often wondered if having speakers ninety degrees to each ear would be a good idea, rather like giant headphones. I keep meaning to try it but have never got round to it. In theory anyway I'm guessing you should still get the formation of the stereo phantom image.

 

[Ivanovich]

This is great thread, thank you!

I’m 55, have tinnitus, don’t hear above 7k. (It was 9k a couple of years ago. )

I wish you luck as it didn’t work out for me. A couple of years ago I played around with the idea of EQ’ing for my hearing loss. What I found was that there is no way to correct for what I can no longer hear, and the corrections within my audible range (boosting 2-5kHz) made music sound awful.

I concluded that my brain must be compensating for the loss with normal sounds, like music. I’ve also started doing some home recording and when I mix I just work on making it sound good to my ears. I checked with my 25 yo nephew and he thought it was a decently balanced mix across the frequency spectrum.

(Since I don’t hear above 7 KHz and there’s little musical content above 10k, I just don’t fiddle with those freqs. I just look at the spectrum in the EQ to see what’s going on there visually only and generally leave it alone.)

I assume this compensation is similar to when you play a high pitch sine wave it changes as you move your head from side to side, but on music you don’t notice any such effect.

Many highly respected music industry pros are older and have hearing loss from past traumas, and still manage to produce great sounding music (loudness wars notwithstanding).

To me, all of this seemed to validate the brain EQ process theory and now I just carry on, trying not to worry about my lost sensory acuity for something I’m so passionate about. I’m not religious but if were, I’d be praying doctors will make progress on a treatment to regrow those little hairs in my inner ear that I’ve blasted away with searing electric guitar for years!

\m/

 

[JoachimStrobel]

I would mention something that happened to me at least twenty years ago. I was visiting an audio repair shop where I used to work. The proprietor and I are about the same age. We were 45 or 50 then. In his reception area he usually had equipment for sale or consignment on display.

On this particular day there were a pair of pretty good speakers and a pretty good power amp. I don't remember any brands. For some reason there also was an audio oscillator sitting there and we decided to play "How bad are our 50 year old , rock an roll listened to ears??"

The initial results were TERRIBLE with neither of us hearing much above 5 kHz. He was laughing about it, but I was really upset, not thinking of myself as hearing impaired. Then completely by accident my head moved in front of the tweeter. It was emitting high frequencies in a laser like beam. It was not a large tweeter either. So we rechecked ourselves and I could easily hear up to 19kHz. Subsequent professional tests have confirmed that my hearing is pretty good for an old Geez especially on the right side. But my point is that if you are using speakers to test make sure you are "on the beam" which may be more laser like than you expect.

I do like the idea of extra EQ for ones own hearing when listening alone. Nice work.

As you said “for hearing alone”. My wife does not like my EQ as she still has a better hearing. I read it is a sign of hearing impairment when the high frequency response differs a lot for different head positions.

 

[JoachimStrobel]

This is great thread, thank you!

I’m 55, have tinnitus, don’t hear above 7k. (It was 9k a couple of years ago. )

I wish you luck as it didn’t work out for me. A couple of years ago I played around with the idea of EQ’ing for my hearing loss. What I found was that there is no way to correct for what I can no longer hear, and the corrections within my audible range (boosting 2-5kHz) made music sound awful.

I concluded that my brain must be compensating for the loss with normal sounds, like music. I’ve also started doing some home recording and when I mix I just work on making it sound good to my ears. I checked with my 25 yo nephew and he thought it was a decently balanced mix across the frequency spectrum.

(Since I don’t hear above 7 KHz and there’s little musical content above 10k, I just don’t fiddle with those freqs. I just look at the spectrum in the EQ to see what’s going on there visually only and generally leave it alone.)

I assume this compensation is similar to when you play a high pitch sine wave it changes as you move your head from side to side, but on music you don’t notice any such effect.

Many highly respected music industry pros are older and have hearing loss from past traumas, and still manage to produce great sounding music (loudness wars notwithstanding).

To me, all of this seemed to validate the brain EQ process theory and now I just carry on, trying not to worry about my lost sensory acuity for something I’m so passionate about. I’m not religious but if were, I’d be praying doctors will make progress on a treatment to regrow those little hairs in my inner ear that I’ve blasted away with searing electric guitar for years!

\m/

I believe that for a successful hearing EQ you need to benchmark yourself vs a younger person on your specific system, everything else is too theoretical. Measure a young’s person hearing on your system. Then measure your hearing. And then compensate for the difference. Be aware, that this testing was then done on thresholds, i.e 20db or so, while you listen to music at 80db. The differences are not the same any more and you have to enlarge them according to one of these Iso curves. The boost will reach ridiculous magnitudes above 10-12 kHz, reaching to 30-40 dB. Even if you could EQ that, it will blow your tweeter at loud playing levels. The old German HiFi “norm” required 12.5 kHz as max reproducible frequency - stick to that.

 

[Ivanovich]

I believe that for a successful hearing EQ you need to benchmark yourself vs a younger person on your specific system, everything else is too theoretical. Measure a young’s person hearing on your system. Then measure your hearing. And then compensate for the difference. Be aware, that this testing was then done on thresholds, i.e 20db or so, while you listen to music at 80db. The differences are not the same any more and you have to enlarge them according to one of these Iso curves. The boost will reach ridiculous magnitudes above 10-12 kHz, reaching to 30-40 dB. Even if you could EQ that, it will blow your tweeter at loud playing levels. The old German HiFi “norm” required 12.5 kHz as max reproducible frequency - stick to that.

I’m good without any of that. I know I have loss at 2k and 4K. If I EQ up by that much or even a fraction of that I don’t like how it sounds. It sounds wrong and unnatural. My brain seems to compensate just fine with normal hearing EQ.

 

[jrobbins50]

Interesting discussion, but somewhat complicated methods to reach a solution. I'd like to simply take the standard audiology dB HL hearing chart covering 500Hz-8,000Hz and develop basic adjustments to an Audiolense target curve to reasonably account for my deficient hearing. For example, attached is a recent hearing test I ran using the Etymotic Home Hearing Test https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6710072/ and my current target curve in Audiolense. Thoughts on how this might be done? Cheers. JCR

 

[Theriverlethe]

Interesting discussion, but somewhat complicated methods to reach a solution. I'd like to simply take the standard audiology dB HL hearing chart covering 500Hz-8,000Hz and develop basic adjustments to an Audiolense target curve to reasonably account for my deficient hearing. For example, attached is a recent hearing test I ran using the Etymotic Home Hearing Test https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6710072/ and my current target curve in Audiolense. Thoughts on how this might be done? Cheers. JCR

I’d strongly recommend seeing a a real audiologist, if you haven’t already. It won’t be possible to compensate for 50dB loss between 1KHZ to 4KHz, for example. Every 3dB increase requires twice the amplifier power.
As the previous poster mentions, your brain will perform its own “EQ” if the frequencies in question cross the audible threshold at all for you. Maybe compare your chart to an equal loudness contour and start by boosting the 2KHz region to see if it helps.

 

[JoachimStrobel]

I initially thought about using my medical hearing test. But my goal is not to design a hearing aid. I simply want to see, what I do not hear from my loudspeakers that others do hear.
Hence, I believe it is mandatory to measure the level of a frequency sine wave at the place where I am listening. And then see at what level I hear it. Every other approach would need a calibrated headphone. And even then it will not yield a loudspeaker-ear response as the headphone shines more or less onto your eardrums.
What you suggest is better done with a hearing aid.
What you suggest would need some sort of feedback mechanism to check that the correction works. That what you dialed in has accomplished it’s goal.
Hint: Use an old vinyl record that you remember well from younger years, assuming CDs were not there yet. Hear it with and without hearing EQ. For me, the difference was like night and day (This involves an ADC which is another discussion..). With newer CDs, that I never had heard with young ears, it is difficult to verify the success.

 

[Ivanovich]

My perception/intuitive guess is that when you have dips in your hearing response, you can’t hear tones at those frequencies, but for music, speech, etc. the brain has psychoacoustics going on and provides some sort of processing gain that fills in all the blanks.

Similarly, at high frequencies moving your head makes the tone cancel and reinforce as you change the distance to the source because the 1/2 wavelengths are so short. When you’re listening to music, you don’t hear these changes. (I’m not sure about pink or white noise though, correlated/ uncorrelated.)

Anyway, I guess/will assume once your hearing loss reaches the point where the brain is not getting enough information, the make up gain fails. I suppose this is where hearing aids come in.

It’s fascinating to me that so many old rocker dudes who must have some hearing loss still produce very good sounding music (loudness wars notwithstanding). The brain seems to compensate and has years of experience and training to fill in the blanks. Or, maybe, just fills in what makes sense! See McGurk effect!

 

[jrobbins50]

Thanks all for the thoughtful responses. I do have an otolaryngologist and that is what led me to hearing aids several years ago for general speech use. A forum participant at Roon Labs mentioned EQing in Roon to compensate for hearing loss. He just adjusted to taste. I thought maybe there was a more analytical approach and I can now appreciate the limitations. I’m guessing that those with better hearing discern the vocals easier and don’t strain to hear In particular the female vocalist, even if soundstaged perfectly center. JCR

 

[JoachimStrobel]

The general hearing test aims at clearing your employer of any fault and hence is limited to 8 kHz and calls for a hearing aid when you can no longer follow a conversation. The ‚Stereophile’ grade of loss starts much earlier, with the doctor saying all is fine. There are some exceptions and they have been highlighted elsewhere.
Believing in the brain’s compensation scheme helps, but only as so far as one does not feel the loss. However, when I switch on my HearinEQ Dirac setting, I do hear things I did not hear before, and I pass every hearing test easily. In the big band trumpet section I hear different phrasings, the high-hat comes through ..... It is annoying for my wife as here hearing is better and she therefore finds my adjustments too hard, but then she appreciates the quality of her good hearing much more. It could be an illusion, but then I lately played my direct-to-disc recording of Woody Hermann‘s Thundering Heard (1977), and there I did hear it again what I had lost: The shivering if my eardrums during one of the trumpet solos that can only be played with a 2gram weight on the pickup.
So for inbetweens, that do not need a hearing aid as such but suffer from losses, EQ driven hearing compensation is the way to go. In 2125 every AVR will have one as part of the room loudspeaker setup.