Calibration tool for Headphone + Hearing capability

[Peter Verbeek]

It just doesn't EQ anything. It works (test section, GUI), but it doesn't seem to EQ anything. I am on windows 10.

The cause is that Equalizer APO isn't working. If it did before then a Windows update "broke" Equalizer APO. This is usually fixed by starting its Configurator. This app checks and fixes, if it can, things that got broken.

However, if Equalizer APO isn't work at all, so after installing it and Peace, then it depends on what device you need to EQ. At installation of Equalizer APO it installs itself (activates) only on the default output (speakers/headphones jacks) and input (microphone jack). If you're using a Bluetooth, USB, HDMI, etc. connection then you need to install Equalizer APO on that particular device in the Configurator.

If you have done the above and Equalizer APO still isn't working then you might select "SFX/EFX" (or if not available) "SFX/MFX" in the "Install as" dropdownbox. This is located at the right of the troubleshooting options. These options are coming available after clicking on the device name in the connector column at the left and then checking the troubleshooting options checkbox.

 

[Paolo Piaser]

The cause is that Equalizer APO isn't working. If it did before then a Windows update "broke" Equalizer APO. This is usually fixed by starting its Configurator. This app checks and fixes, if it can, things that got broken.

However, if Equalizer APO isn't work at all, so after installing it and Peace, then it depends on what device you need to EQ. At installation of Equalizer APO it installs itself (activates) only on the default output (speakers/headphones jacks) and input (microphone jack). If you're using a Bluetooth, USB, HDMI, etc. connection then you need to install Equalizer APO on that particular device in the Configurator.

If you have done the above and Equalizer APO still isn't working then you might select "SFX/EFX" (or if not available) "SFX/MFX" in the "Install as" dropdownbox. This is located at the right of the troubleshooting options. These options are coming available after clicking on the device name in the connector column at the left and then checking the troubleshooting options checkbox.

Thank you so much, it worked like a charm (the third option).

 

[Peter Verbeek]

Thank you so much, it worked like a charm (the third option).

Great to hear.

 

[mmdiss]

Hello everybody and thank you for this great software.

I don't want to bother you on how I came here, so I'll try to get straight to the point because I going crazy on this

On my laptop (HP) I recorded the Stereo Mix with Audacity while doing the hearing test (only the sine tone, without the noise). The results below are quite confusing.

I am not using any equalization (not in Peace, nor directly with APO equalizer, each graphic equalizer is turned off).

I am running on Windows 10.

What I was expecting is dB shown in the Hearing test was the same recorded in Audacity (to explain better: a difference of 10 dB in Peace should reflect in the same difference of 10 dB in Audacity, regardless of the absolute level). While this holds true for a single tone (ie. lowering or rising 10 dB in Peace I can see the same variation in Audacity) this seems not true between *different* tones.

See the screenshots attached.

Even if I'm not too familiar with the AutoIt language I went through the code (Peace1591.au3) but I couldn't find where to download the "DirectSound.au3" (a "DirectSound.7z" should exist but I ended up in a page in German where I can't find if there is a link to download that file).

Anyway, I do understand that this row "[3456] $SineBuffer.SetVolume($SineVolume)" sets the volume, and it appears to be taken in Decibel.

Can you please help me in understanding if I did something incorrectly? I can't figure out by myself if I'm misunderstanding something or if the output level is incorrect.

Thank you for reading!

 

[Peter Verbeek]

but I couldn't find where to download the "DirectSound.au3"

I forgot to include the DirectSound source files but now I have as you can see on the source code page. Sorry about that.

I can't figure out by myself if I'm misunderstanding something

Perhaps I can take away one misunderstanding (if this is one in the first place). The test environment dB isn't the same as the output dB of a sound. Of course the shape over the frequencies is the same. The resulting graph (of a test) doesn't include the set Windows volume. This is by design btw.
Anyway, after doing a sine sweep myself (recording the stereo mix output in Audacity as you did) I get a perfect flat db line (see attached image). Are you sure there isn't an EQ going on somewhere? Peace disables its EQ when going to the test interface. Do you notice a volume difference between the sines?

 

[Robbo99999]

Hello everybody,

I've been dreaming for such a tool for many years and now it is part of the Peace EQ interface.
View attachment 91510
Has anyone tried it yet? If you use a Windows PC as a music source, give it a go.
I was part of the development and so I used it the most. If you need more info than what's on my write up, ask:
https://www.beautifulaudio.biz/peace-equalizer-calibration-tool

Cheers

That is pretty interesting, but EQ'ing to Equal Loudness is not the ideal Target Curve, and nor is correcting for natural hearing loss, which is what this process would also do.

Regarding the first point of Equal Loudness I do have some gaps in my understanding (I admit), but thinking about it superficially you don't want to EQ each frequency to equal loudness because we're not supposed to hear all frequencies to the same loudness in a sine sweep for instance. This is the Griesinger Method isn't it, I believe it's been proven to be fundamentally flawed.....I remember some esteemed and knowledgeable members here showing why this was so, I believe it was @Mad_Economist as one of them, but he doesn't post much on here anymore, but he always backed up his arguments with research papers.

My second point re Correcting for Hearing Loss. The way I understand is that you don't want to be correcting for hearing loss unless you have such bad hearing loss that you wear a hearing aid in your daily life. The thinking behind this is that you become accustomed to what sounds natural in your everyday life as you age & as your hearing declines gradually, so correcting for this hearing loss (mostly by boosting higher frequencies) will not make the music sound more natural, because you burn in your brain to what sounds normal just by living your life in the acoustic environment within which you live every second of your waking life......so I don't believe it's natural & good to correct for those subtle hearing degredations.....in fact I see that as unnatural & would produce a strange result. The caveat to that is if you wear a hearing aid, and then that means you expect certain frequencies to be boosted because you just can't understand speech even properly without one......and it's here where I completely agree that you should compensate for those deficiencies in your headphone EQ. I discussed this with Oratory one time and he agreed with my thinking on the points in this paragraph.

So, in conclusion, I'm not convinced by this process, although I definitely respect the time & dedication & positivity towards intended better audio for the community that is evident in the creation of this tool, but I don't really agree with the fundamental process on a good sound "scientific" basis. One use I could see of it though, in terms of my own usage, I could envisage using it to iron out local peaks (over a short range in the frequency response) in the frequency response after doing a Harman EQ......my thinking behind this is that I know roughly how a sine sweep sounds on my 2 anechoically flat reference speakers, in terms of there is no waviness in the response, it's gradual changes.....and one other important thing I've noticed which would be a problem with your method is that I know from doing sine sweeps on my speakers that my hearing has an absolute massive positive gain in sensitivity between 8-12kHz.....I remember trying to EQ that down in headphones and it required something like -16dB in that area to equalise it (& sounded super strange after doing so), but I shouldn't have done that because I have a natural increase in my hearing sensitivity in that range (so that's another problem with your process). Yes, so perhaps I could use the Test Button in your tool to just iron down local peaks that I know shouldn't be there? But I don't agree with the overall use of that tool to create a total Target Curve.

 

[Peter Verbeek]

But I don't agree with the overall use of that tool to create a total Target Curve.

After reading your post I understand your hesistation in using it and the questions you have about it. I agree this tool isn't very scientific mainly because the equal-loudness curve is a general curve and also quite arbitrary as measuring the audio capability of man is very difficult to do, at least accurately. Our brains are just too capable to reproduce and create stuff in this case audio that's beyond the biology (construction) of our ears (or any of the senses for that matter). I guess this is what you're trying to say. So the aim of the test interface isn't to be a particularly fine-mazed tool and method to create an ultimate eq for ones ears. It's main usage lies in the support to discover possible flaws in ones hearing. In other words, to give a tool to the people so they can tinker with their hearing.

The most common positive remark I get of Peace users is that its EQ abilities give back the music to persons with impaired hearing. Without the ability to eq, the hearing impaired enjoy music far less as they're telling often. I guess that's because they remember how nice music sounded in the past.

As a test result can automatically be translated to an eq, such an eq always needs to be corrected. I would go so far to say that it's better to determine the flaws of ones hearing then to create, or download for that matter, an eq. And talking about predefined eq's. On the internet there are a lot of eq's for specific brands and type of headphones. That's fine as the search of people for these eq's is large. But these eq's are like the equal-loudness curve general ones. They show the frequency response of headphones and as such one can determine where to correct these frequencies. But this doesn't take things into account like taste, frequency response on ones ears or position of the headphones on ones head. There are a lot factors. One experience of my own: After buying my Beyerdynamic DT770 32 Ohm headphones I downloaded a frequency response. The subsequencial eq totally flattened these headphones. Now I was left with a flat (dull) sound so I had to eq them myself. What I'm saying is that even such measured frequency responses available on AutoEQ for instance, don't have the result many expect them to have. So people start tinkering after using such eq. But still we want to give people tools so they are going through the process of enhancing their experience. That's the main reason I create tools and support persons on audio matters.

 

[Robbo99999]

After reading your post I understand your hesistation in using it and the questions you have about it. I agree this tool isn't very scientific mainly because the equal-loudness curve is a general curve and also quite arbitrary as measuring the audio capability of man is very difficult to do, at least accurately. Our brains are just too capable to reproduce and create stuff in this case audio that's beyond the biology (construction) of our ears (or any of the senses for that matter). I guess this is what you're trying to say. So the aim of the test interface isn't to be a particularly fine-mazed tool and method to create an ultimate eq for ones ears. It's main usage lies in the support to discover possible flaws in ones hearing. In other words, to give a tool to the people so they can tinker with their hearing.

The most common positive remark I get of Peace users is that its EQ abilities give back the music to persons with impaired hearing. Without the ability to eq, the hearing impaired enjoy music far less as they're telling often. I guess that's because they remember how nice music sounded in the past.

As a test result can automatically be translated to an eq, such an eq always needs to be corrected. I would go so far to say that it's better to determine the flaws of ones hearing then to create, or download for that matter, an eq. And talking about predefined eq's. On the internet there are a lot of eq's for specific brands and type of headphones. That's fine as the search of people for these eq's is large. But these eq's are like the equal-loudness curve general ones. They show the frequency response of headphones and as such one can determine where to correct these frequencies. But this doesn't take things into account like taste, frequency response on ones ears or position of the headphones on ones head. There are a lot factors. One experience of my own: After buying my Beyerdynamic DT770 32 Ohm headphones I downloaded a frequency response. The subsequencial eq totally flattened these headphones. Now I was left with a flat (dull) sound so I had to eq them myself. What I'm saying is that even such measured frequency responses available on AutoEQ for instance, don't have the result many expect them to have. So people start tinkering after using such eq. But still we want to give people tools so they are going through the process of enhancing their experience. That's the main reason I create tools and support persons on audio matters.

Hi man, had to give you a Like because I think this tool is useful for people with proper hearing impairment who have hearing aids for instance, and that's quite beautiful for those people to experience music like they used to (as I guess hearing aids don't interact particularly well with music from speakers let alone the enclosed cup of a headphone (but I'm intuitively guessing on the last point)). So credit where's credit's due. I'll still sit on the doubtful side for those without hearing impairment.

 

[Jumbotron]

Hello everybody,

I discovered this thread very recently. I installed the Peace GUI in order to try this headphone and hearing calibration tool with my pair of Takstar HF 580, because they definitely need equalization, there are no EQs widely available for them and I do not know how to create a really good EQ preset just by ear. My hearing is normal for my age by the way (I am in my forties).

Well, I passed the headphone and hearing test for both the left and right channels and then I let the tool create a graphic EQ for my HF 580. This was yesterday. The sound of these cans has definitely improved with the created EQ preset. Thank you.

 

[mmdiss]

Are you sure there isn't an EQ going on somewhere?

Thanks Peter, you are right and now I'm feeling very dumb: it occurred to me only now to generate different tones with the same amplitude using Audacity and then playing back to see how they are recorded by the Stereo Mix... Guess what? I've found the same, unexpected, 9-10dB difference between 200hz and 3200hz. It looks like there is some EQ somewhere else, not from Equalizer APO, I suspect something related to bloatware that comes within my HP laptop (and that probably I can get rid of using APO, that's just crazy).

Do you notice a volume difference between the sines?

Here comes the trouble... I don't know, as I'm trying to calibrate my headphones. And it only gets worse when you came to know that Stereo Mix works only when headphones are disconnected...

Anyway, I've just discovered I can record the audio using this https://github.com/rdp/screen-capture-recorder-to-video-windows-free/ instead of Stereo Mix (useful when headphones are plugged). With this, I found a flat response for headphones (good! Maybe the hidden equalization applies only to the internal speakers of the laptop) but I am not sure if this is a reliable result since I get a different delta with loudspeakers (ie. without plugging headphones) than with the Stereo Mix (I get 3dB of difference, instead of 9-10dB measured with the Stereo Mix).

I guess this is "problem solved", thanks

 

[Peter Verbeek]

Hi man, had to give you a Like

Thanks

I'll still sit on the doubtful side for those without hearing impairment.

To be honest we (Silvian and especially I) had some doubts whilst developing this tool (which is mainly for headphones btw.). But after we were using it during the development we discovered that it can be useful. The resulting eq's we got were useful in contrast to my doubts. I was quite surprised knowing the issues of the equal-loudness contour. So we pressed on to fully develop a first version. How helpful this tool is for those without hearing impairment, still remains to be seen. We had some good feedback though. Some were very pleased, also people without impairments. Anyway, it helped me discover and correct some flaws over the frequencies. However, nowadays I tend to use headphones as they are, so with their flaws and my flaws. I do have to say I'm not listening to music much so that's a huge difference between me and others. Btw. Thanks for your remarks. Always good to know what others think. Tools can only be improved if others can have their saying on them. I also remain in some doubts about our test tool, even now as it's proven to be useful.

 

[Peter Verbeek]

The sound of these cans has definitely improved with the created EQ preset.

Great to hear

 

[Peter Verbeek]

Thanks Peter, you are right and now I'm feeling very dumb

Don't be. Over the years many good feature developments and bug repairs happened as a result of some mistake or misconception. In other words, any feedback is good.

but I am not sure if this is a reliable result since I get a different delta with loudspeakers (ie. without plugging headphones) than with the Stereo Mix (I get 3dB of difference, instead of 9-10dB measured with the Stereo Mix).

I've no clue where this eq is coming from, especially knowing that Stereo mix is in the equation. Usually it's an eq due to some eq setting in the audio software or some other audio enhancement. However, I don't know what's going on your laptop. Perhaps it is some bloatware or perhaps some crappy audio driver. If you found out more, could you let me know?

 

[Mad_Economist]

Regarding the first point of Equal Loudness I do have some gaps in my understanding (I admit), but thinking about it superficially you don't want to EQ each frequency to equal loudness because we're not supposed to hear all frequencies to the same loudness in a sine sweep for instance. This is the Griesinger Method isn't it, I believe it's been proven to be fundamentally flawed.....I remember some esteemed and knowledgeable members here showing why this was so, I believe it was @Mad_Economist as one of them, but he doesn't post much on here anymore, but he always backed up his arguments with research papers.

Your timing is impeccable, I just started posting here more again

Griesinger's method may be defined as "subjective loudness comparison of headphones and speakers" - he has the listener match a bandlimited noise sample's loudness between the headphones on their head, and the speaker in front of them, to attempt to derive HRTF. This is improper, because the apparent acoustic source position influences perceived loudness - see Thiele 1986, 2016, he refers to it as the "SLD effect". Griesinger's method is also flawed because his closely-positioned, frontal speaker approximates a free field HRTF (because direct sound massively dominates), and the free field HRTF is radically unpreferred relative to both the diffuse field HRTF, and the equalized approximation of DF which the Harman target represents.

This method does not use a reference in the listener's real room - the only comparison being made is between frequencies at threshold of hearing, so superficially it might seem to be immune. However, the ISO equal loudness contours which it uses as references are based on a frontally located loudspeaker in a free field, so the target remains unsuitable, compounded by the fact that listener HRTF differs (and that this test would attempt to force HRTFs to "converge" on the average, at least net of its impact on threshold of audibility), so the effective target response this aims for is "a generic human head, in an anechoic chamber, with a frontally located loudspeaker".

Because the differences between FF and DF HRTFs for an average human are known, and in turn the delta between DFHRTF and the Harman target for a given fixture, it would be possible to improve the target response that this process aims for, or give users control over it.

As I understand @Peter Verbeek and @Silvian's goals here (disclaimer: I have not read this whole thread), however, the main purpose is to deal with substantial listener hearing loss. Because hearing loss is a grossly large adjustment to our perceived "frequency response" of the world, this process as stands probably does represent an improvement for a meaningfully hearing impaired listener, although I would suggest revising its targets as said.

 

[Robbo99999]

Your timing is impeccable, I just started posting here more again

Griesinger's method may be defined as "subjective loudness comparison of headphones and speakers" - he has the listener match a bandlimited noise sample's loudness between the headphones on their head, and the speaker in front of them, to attempt to derive HRTF. This is improper, because the apparent acoustic source position influences perceived loudness - see Thiele 1986, 2016, he refers to it as the "SLD effect". Griesinger's method is also flawed because his closely-positioned, frontal speaker approximates a free field HRTF (because direct sound massively dominates), and the free field HRTF is radically unpreferred relative to both the diffuse field HRTF, and the equalized approximation of DF which the Harman target represents.

This method does not use a reference in the listener's real room - the only comparison being made is between frequencies at threshold of hearing, so superficially it might seem to be immune. However, the ISO equal loudness contours which it uses as references are based on a frontally located loudspeaker in a free field, so the target remains unsuitable, compounded by the fact that listener HRTF differs (and that this test would attempt to force HRTFs to "converge" on the average, at least net of its impact on threshold of audibility), so the effective target response this aims for is "a generic human head, in an anechoic chamber, with a frontally located loudspeaker".

Because the differences between FF and DF HRTFs for an average human are known, and in turn the delta between DFHRTF and the Harman target for a given fixture, it would be possible to improve the target response that this process aims for, or give users control over it.

As I understand @Peter Verbeek and @Silvian's goals here (disclaimer: I have not read this whole thread), however, the main purpose is to deal with substantial listener hearing loss. Because hearing loss is a grossly large adjustment to our perceived "frequency response" of the world, this process as stands probably does represent an improvement for a meaningfully hearing impaired listener, although I would suggest revising its targets as said.

Nice, good to see you're back once in a while! I understand some of what you've put down there, perhaps if there's room for improvement of @Peter Verbeek 's method then maybe they can tweak it with your help, as an idea. I agree that their process would help the hearing impaired, so I think it's for sure worthwhile, but if it can be improved then great. I remember you posting on this topic before so thought it would be helpful to mention you here.

 

[Mad_Economist]

Nice, good to see you're back once in a while! I understand some of what you've put down there, perhaps if there's room for improvement of @Peter Verbeek 's method then maybe they can tweak it with your help, as an idea. I agree that their process would help the hearing impaired, so I think it's for sure worthwhile, but if it can be improved then great. I remember you posting on this topic before so thought it would be helpful to mention you here.

I appreciate you bringing this to my attention! I'm a bit busy these days, but I'd be happy to contribute if @Peter Verbeek wants some input on this project - of all the subjective loudness matching approaches I've seen, it seems like it has the best potential (if primarily because it's exponentially more convenient for end users...).

Edit: FWIW, for collaboration, if you would like input, I'm easier to get ahold of on Discord. Here's a neat little audio science server I'm a part of: https://discord.gg/2ZjhBfRS4A

 

[Robbo99999]

I appreciate you bringing this to my attention! I'm a bit busy these days, but I'd be happy to contribute if @Peter Verbeek wants some input on this project - of all the subjective loudness matching approaches I've seen, it seems like it has the best potential (if primarily because it's exponentially more convenient for end users...).

Edit: FWIW, for collaboration, if you would like input, I'm easier to get ahold of on Discord. Here's a neat little audio science server I'm a part of: https://discord.gg/2ZjhBfRS4A

Ha, I know, what was I thinking, I meant like if it was as simple as including a few formulas or a slight change of process. (I've never used Discord).

 

[Peter Verbeek]

@Mad_Economist First of all thanks for sharing your insights. They are very welcome. For instance, I didn't know about Griesinger. I guess many try to "measure" the "general" hearing of our ears/brains. As said this procedure will always have inherent flaws. You've also named a few like unstationary head in a general room whilst having a front speaker.

this process as stands probably does represent an improvement for a meaningfully hearing impaired listener, although I would suggest revising its targets as said.

Yes, and mainly for headphones as the combination headphones and hearing differs from brand/type/person to brand/type/person. Using it for speakers in ones room could be troublesome although we haven't had any feedback on that yet. In other words, the test tool might be of some usage for speakers measurement but we don't know yet. I feel when using it for speakers it needs some improvements somehow, at least in the procedure (description).

I'd be happy to contribute

Great! There's a catch however. As the test tool, well Peace as a program, is written in the AutoIt programming language, there isn't much room for complex algoritme calculations in realtime, for instance creating some complex test sounds on the fly with the mic as an input. This could be useful for speakers testing like REW does. Anyway, I usually determine (and try to fix) obstacles that may arise from the slow AutoIt programming. All depends what suggestions/improvements you may have. Of course the workflow/procedure and interface elements can be improved upon. As you're implying, the test interface and procedure, although more complex than I want it to be, is still quite easy. Therefore people might benefit by doing tests. This is already being proven by feedback, also for the non-hearing impaired. A bit to my surprise. Silvian was more convinced but he has a lot of experience compared to me. It seems to me you have too.

So if you have suggestions (and have the time) you're welcome

 

[Mad_Economist]

You've also named a few like unstationary head in a general room whilst having a front speaker.

FWIW the stationary or not ness of the head isn't necessarily an issue - it's the ratio of indirect to direct sound. It isn't too hard to stay within a say 5 degree window of a speaker with your head for a measurement, but because the response varies based on the incidence of sound, the ratio of direct sound to indirect sound impacts total response. For example, here are the ISO11904 standard free and diffuse (pressure) field HRTFs for different measurement points

The difference between them is not precisely the same at all measurement points, but given that the largest outlier is the eardrum data, which Hammershøi & Møller (whose data is the source of the ISO standard) found to be most variable, this is likely random variation.

As Olive, Welti, & McMullin 2013 found that the in-room baseline of the Harman target approximated the DFHRTF, this would be a more suitable baseline for a target.

Fig. 17 shows the frequency response of a circumaural headphone measured at DRP when equalized to the preferred target response (black). Also shown is the frequency response of the same headphone equalized to match the flat in-room response of a loudspeaker (dotted), which approximates the commonly practiced diffuse-field (DF) headphone calibration applied to many commercial headphones. Compared to the preferred headphone target curve, the DF calibration would make the headphone sound too thin and bright due to the lower bass and higher treble levels. This was reported in a previous study [6], and has been confirmed again in the current study.

Great! There's a catch however. As the test tool, well Peace as a program, is written in the AutoIt programming language, there isn't much room for complex algoritme calculations in realtime, for instance creating some complex test sounds on the fly with the mic as an input. This could be useful for speakers testing like REW does. Anyway, I usually determine (and try to fix) obstacles that may arise from the slow AutoIt programming. All depends what suggestions/improvements you may have. Of course the workflow/procedure and interface elements can be improved upon. As you're implying, the test interface and procedure, although more complex than I want it to be, is still quite easy. Therefore people might benefit by doing tests. This is already being proven by feedback, also for the non-hearing impaired. A bit to my surprise. Silvian was more convinced but he has a lot of experience compared to me. It seems to me you have too.

Fortuitously, for a target response, there is no need for dynamic adjustments. There are cases where you benefit by dynamic adjustment - e.g. head tracking for spatializing algorithms - and this can meaningfully improve subjective experience, but for frequency response correction, a static equalization may be applied.

Specifically, my recommendation would be using a diffuse field equal loudness - either a directly sourced one from the lit (there's one somewhere in the ISO standards, I believe, maybe even the same work as the FF equal loudness), or by applying the delta of FF > DF HRTF to the FF equal loudness (although this may have issues, due to the aforementioned influence of perceived acoustic source position) - and, following the work of Olive, Welti, McMullin, and Khansaripour, applying a generally "downward sloping" adjustment. You could allow listeners to adjust a "slope" parameter (e.g. the stylized form of the original Harman adjustments from 2009 are 1dB/octave from 20-20000hz), give listeners the obligatory two 2nd order shelf filters at 105 and 2500hz from the linked paper, or directly apply the average preferred adjustments from said filters that Olive, Welti, and McMullin found:

Because the test is still based on subjective loudness matching, it will likely be vulnerable to meaningful errors - it's possible that using the threshold of hearing will decrease this somewhat, but I'm not aware of research supporting that so that's pure conjecture - so I would encourage you to include a note suggesting that users adjust the result, because people often take these sorts of curves as "canonically correct" because they're based in "science".

 

[Bama214]

Not meaning to overly deflect this thread (which is great by the way!), but I'd like to bring up a similar capability for the iPhone.

Under "Accessibility" in the settings, one can navigate to "Audio/Visual" then to "Headphone Accomodations" and then to "Custom Audio Setup". At that point one can select from several presets, but more importantly, one can choose to use a personal audiogram as a basis.

I"ve used the app "Mimi" to generate an audiogram of my aging hearing and confirmed that I have what is defined as a "Slight" hearing loss. Mimi can then save the audiogram to the Apple "Health" app. This audiogram then serves as the basis for a custom EQ for my Airpods Pros.

My listening with this adjustment is just starting, but I'd have to say my initial reaction is quite positive! So far, I've just used this process to compensate the Airpods Pro for mobile use, but I'm now tempted to use the Mimi generated audiogram as a basis for an EQ under APO/Peace for home use on my Focal Clears.