[yticolev]

I also need a quieter space for testing. Still, the results shouldn’t be too different, right?

My (quiet) furnace kicked in first test with Mimi. Considering the low volume of the test, that was enough to impact the results. I also used a wired pair of Earpods, and the microphonic effect with the wires touching my collar was too much. Next test with furnace off and collared jacket off was much different, subjectively and objectively.

Frankly, I thought I would do better than just above average considering how much I've babied my ears my whole life. I wear earplugs in many situations I've not seen others do, like airplanes. Not fond of loud sounds which may make my complaints of not enough volume with the APMs more credible.

 

[tkddans]

I would recommend headphones with more like Harman Target. Something with greater uppermids like Sennheisers. This will equilize the area you are reduced in sensitivity. This may also be why you turn up volume for headphones that the uppermids presence is low. To hear the uppermids more.

You seem to be treble sensitive. Which shows your ears are young. Any headphones you tried that had too much treble?

Edit: I think your headphone could have skewed the test. What headphone were you using for the test?

What headphones do audiologiests recommend for these types of tests I wonder? Also it's best to test in sound isolated booth.

My bet is that any headphone shown in Mimi’s choices are equally good for testing, more or less. The app does recommend using headphones that have been calibrated. The app makers probably accounted for frequency response in their playback, if I had to guess what “calibrated for Mimi” would entail behind the scenes.

When I ran the test, it only showed AirPods for me. Maybe they just didn’t calibrate it for the Pro’s. I will rerun with non-pro AirPods when I can soon enough.

 

[SilverEars]

Yep sounds wonky to me. I listen at about 50% and generally, volume stays right around 75 DB. I could do louder for sure, and I’ll probably need to turn it up for quieter recordings, but that works for me. I also use Balanced/Slight and that seems to increase sound level automatically. Not sure how that might skew with the Health app’s decibel reporting.

True. I know with Apple products they can know the dB of intended output, but yes the real dB output still depends on the actual loudness ouputed as the tracks have different loudness levels and replaygains. How do they figure out the actual loudness? I'm thinking not so precise. Can they extrapolate from content's information combined with the information about the output of the hardware?

Real precision would be a mic calibrated for the ear position inside the cups where the ear sits. I think dB is touchy to positioning of the ear.

 

[tinyman392]

I should add that I'm using with the telescoping headband adjusted at maximum length. Although I'm likely on extreme end of the populus, I have had headphones with more adjustment length. Perhaps there are people that require longer.

I forget which review it was, but a reviewer said the same thing. We big heads demand more flexability! Lol

Crinicle is the one who had this issue.

Great data found there. Thanks for pointing it out!

Looks like I’ve had about 87 dB on average for many more hours than I should this week. Only because of extreme amount of testing I’m sure, but darn. Looks like my normal listening is usually high, but considered ok in total exposure time.

It’s also pretty cool how the data for each separate headphone is parsed out. So it looks like I listened to Panda and K361 at similar peak levels.

EDIT:: I listen to my headphones at over 80%. It’s loud on most music. On classical I go 100% or close to it. Despite listening to it so loud, it’s still hard to hit unsafe limits.

For people thinking Apple is protecting hearing, you may be right. But if you go to 100% volume, you WILL break the healthy weekly exposure limits within just a few hours. I know this because I got a notification and the volume went lower (I had a setting turned on that would lower my volume for me when I reached exposure limits; even if I could raise it again right after, I wanted to know when it happened).

You've hit nearly 100dB quite a few times... Note that with other headphones, Apple can't be nearly as precise as it can with their own AirPods. With their own AirPods they know how loud they produce given a sound signal and volume level. With others they'd have to guess a lot of this.

 

[SilverEars]

My bet is that any headphone shown in Mimi’s choices are equally good for testing, more or less. The app does recommend using headphones that have been calibrated. The app makers probably accounted for frequency response in their playback, if I had to guess what “calibrated for Mimi” would entail behind the scenes.

When I ran the test, it only showed AirPods for me. Maybe they just didn’t calibrate it for the Pro’s. I will rerun with non-pro AirPods when I can in soon enough.

Ok, makes sense to calibrate the headphone because with earphones, there is no pinna gain like circumaural. I wonder what an audiometer headphone response is like?

Here is one by sennheiser, in which being closed makes sense for sound isolation. https://www.bhphotovideo.com/c/prod...da_300_audiometers_headphones.html?ap=y&smp=y

 

[tinyman392]

True. I know with Apple products they can know the dB of intended output, but yes the real dB output still depends on the actual loudness ouputed as the tracks have different loudness levels and replaygains. How do they figure out the actual loudness? I'm thinking not so precise. Can they extrapolate from content's information combined with the information about the output of the hardware?

Real precision would be a mic calibrated for the ear position inside the cups where the ear sits. I think dB is touchy to positioning of the ear.

They have waveform of what's being sent to the headphones and based on that know what's coming out of the headphones. So they can be very precise with their AirPods as the volume level basically depends on the waveform that's being set to the headphones. With others they'd have to estimate it so it won't be as precise and more of a ballpark.

 

[tkddans]

You've hit nearly 100dB quite a few times

Yes but that’s instantaneous peaks only, not sustained or average.

 

[tinyman392]

Yes but that’s instantaneous peaks only, not sustained or average.

This is true. Keep in mind that the average isn't a sustained either, it's just average. Your iPhone will actually limit the volume after you go too high for too long though. Happened to me when I had it connected to a BT DAC/Amp (iPhone set to 100%, BT amp set to like 30%), so it thought I was listening to music at above 120 dB.

My average is sitting at around 67 for the week.

Edit: you have to have the setting enabled for the iPhone to limit you automatically after you go past the acceptable levels. I think it's on by default (I never actually turned it on).

 

[yticolev]

I would suspect that calibrating one's headphones to correct for falling frequency response to a hearing test will sound very artificial and unpleasant. After all, that is not how you hear the world when you are not wearing earphones. Your brain is compensating all the time - a natural sound processor. Adjusting for a small blip may or may not sound better, just as random fiddling with EQ may yield personally pleasing results. But raising FR at 8 kHz by 30 dB as Mimi implies for me? No way.

 

[SilverEars]

I would suspect that calibrating one's headphones to correct for falling frequency response to a hearing test will sound very artificial and unpleasant. After all, that is not how you hear the world when you are not wearing earphones. Your brain is compensating all the time - a natural sound processor. Adjusting for a small blip may or may not sound better, just as random fiddling with EQ may yield personally pleasing results. But raising FR at 8 kHz by 30 dB as Mimi implies for me? No way.

Based on the test, does it seem to correlate to your headphone taste?

 

[MayaTlab]

True. I know with Apple products they can know the dB of intended output, but yes the real dB output still depends on the actual loudness ouputed as the tracks have different loudness levels and replaygains. How do they figure out the actual loudness? I'm thinking not so precise. Can they extrapolate from content's information combined with the information about the output of the hardware?

I don't see why they couldn't. In fact it's probably trivial to do. Analyse the track's loudness, and evaluate it in regards to the Airpods' volume setting. That should already make for a decent estimation.

Real precision would be a mic calibrated for the ear position inside the cups where the ear sits. I think dB is touchy to positioning of the ear.

Most current ANC headphones have a mic directed towards the ear (feedback mechanism in addition to the feedforward mics directed towards the outside), Apple's seem to have full(-ish ?) range MEMS throughout which would enable them to know the output at higher frequencies than the ones commonly used for ANC.
The remaining imprecision would be down to variations in how headphones interact with the user's anatomy past 2000hz or so but may still be quite accurate with a lot of smoothing (IDK).

 

[tinyman392]

I would suspect that calibrating one's headphones to correct for falling frequency response to a hearing test will sound very artificial and unpleasant. After all, that is not how you hear the world when you are not wearing earphones. Your brain is compensating all the time - a natural sound processor. Adjusting for a small blip may or may not sound better, just as random fiddling with EQ may yield personally pleasing results. But raising FR at 8 kHz by 30 dB as Mimi implies for me? No way.

This is the equivalent of doing an EQ, so it actually could make headphones sound less artificial. Raising by 30 dB could create distortion, but you could also do a reverse EQ where you raise everything, then drop it so the max EQ is set to 0dB.

 

[tinyman392]

I don't see why they couldn't. In fact it's probably trivial to do. Analyse the track's loudness, and evaluate it in regards to the Airpods' volume setting. That should already make for a decent estimation.



Most current ANC headphones have a mic directed towards the ear (feedback mechanism in addition to the feedforward mics directed towards the outside), Apple's seem to have full(-ish ?) range MEMS throughout which would enable them to know the output at higher frequencies than the ones commonly used for ANC.
The remaining imprecision would be down to variations in how headphones interact with the user's anatomy past 2000hz or so but may still be quite accurate with a lot of smoothing (IDK).

That mic is also used for Adaptive EQ too. But yeah, Apple knows how loud their products are, and not only based on the microphone reading either. They could easily just collect data using a dummy head to create a very strong approximation (like within 1 dB) of how loud it is given a waveform.

 

[pnoble]

Do you mind sharing your age? I knew somebody that turns up volume real loud probably to hearing damage level, but when you converse with him he seems to hear fine. Weird.

I got to say this is a very useful feature. This is the benefit of Apple ecosystem.

For what it’s worth, I tested as slight hearing loss using the uncalibrated APM. I switched to first gen APs (calibrated) and a VERY quiet environment and my results improved to “no hearing loss.”

The latter calibrated test is the more accurate one! Similar result initially using the APP v the AP.

 

[ricksastro]

They have waveform of what's being sent to the headphones and based on that know what's coming out of the headphones. So they can be very precise with their AirPods as the volume level basically depends on the waveform that's being set to the headphones. With others they'd have to estimate it so it won't be as precise and more of a ballpark.

The audiogram compensation using an uncalibrated headphone should be accurate for your ears paired with that headphone. i.e. if the headphone is weak at a certain frequency you will likely have a harder time detecting it so it will show up as a weakness in your hearing and audiogram compensation will boost it slightly. Now this isn't an accurate measurement of your hearing in real terms and shouldn't be used on a different set of headphones, but the net effect is good for the pairing of you and that set.