Safe listening levels and headphone voltage/power requirements

[F1308]

When looking at efficiency numbers (dB/mW) the impedance is involved.
When looking at sensitivity (dB/V) it is not.... that is ... IF we have a source with near 0 ohm output resistance * and assuming the source in question is not current limited # and assuming the FR is the same ~ or very, very similar.

* When the output resistance is above say... 1 ohm or so (many interfaces and music instruments are, phones etc are not) then there is voltage division at play.
Say we have a 10 ohm output R and 2 headphones with the exact same sensitivity (dB/V) but one is 10 ohm and the other one is 32 ohm.
On a low output R source they will play equally loud (assuming FR is the same as well).
On a 10 ohm output R the 10ohm headphone will be playing 6dB softer and the 32 ohm headphone will play 2.4dB softer.

So in the end, not only sensitivity matters but also impedance combined with output resistance of the source.

# When a source is current limited (most are) there can be a difference in max SPL that can be reached.
When 2 headphones have the same sensitivity (dB/V) and the same FR but one is say... 12 ohm and the other one 300 ohm and the source is used that can supply 10V (in 300 ohm) but only 30mA then the 300 ohm headphone can reach 10V but the 16 ohm headphone can only reach 0.5V that is a whopping 26dB difference in max output level as the dB/V would be the same.
So... max output power in specific impedances also plays a role when looking for max. SPL

~ When a headphone is listed as having say 105dB/V then this is only valid at a specific frequency or frequency band. This depends on the measurement method.
It could be at 400Hz, at 1kHz or even averaged over a narrow band noise. Usually this is not specified.
A headphone with 105dB/V could well have 120dB/V in the bass (very bassy headphone) or be 95dB at 30Hz (bass-light headphone).
Of course with the right EQ you can effectively change the FR to flat or a specific target. When done digital and applying a boost you will need negative pre-amp.

Thank you very much.
It therefore goes without saying that one cannot compute much, since there is not enough data provided to we the users. Just plug and see as I did, finding what I explained before, exactly the opposite of what thought to be obvious.

 

[solderdude]

Calculating exact SPL may not be exact but when all aspects are taken into consideration you can get a pretty good idea.
When one is a calculus nerd it may be fun to do all the calculations.

I believe @Robbo99999 once attempted to estimate how loud one was listening by using sensitivity numbers, a DAC with a known output voltage and amp with 1x gain.

Now and then play something loud (impressively loud) for the duration of a song.
In general, when you get the urge of turning the volume down you are probably playing too loud.

The danger lies is the 'automatic gain control' in your hearing. When one has visited a loud (pop/rock) concert one knows it is loud yet stays there for hours.
Once you left the venue you might have noticed traffic is not 'loud' and the world has gotten a bit 'quieter' around you.
Likewise, late in the evening or early in the morning you can hear a clock tick but not during the day.

Therein lies the problem with 'sensible levels' based on the urge to dial down.
In the morning or late in the evening (while having a quiet evening) you automatically do not play as loud as during the day.

Still... in general when you get the urge to dial down the volume (after 1 or 2 songs) you are playing above 80-85 dB SPL average.

Note that a sensitive headphone 120dB/V or higher can reach non-sensible levels even from a phone/tablet/PC or player.
Sometimes when I am next to some young bloke and I can clearly hear the IEMs I wonder if they even get the urge to dial down the volume.


Ultimately measuring the output voltage under actual load (not easy to do, requires some knowledge) or measuring it with suitable calibrated gear IRL is the only real way to do this.

But.. as said.. lots of calculations, FR plots, accurate data of which one knows how it has been obtained and a PC with level monitoring, DR ratings of a recording etc. will get you in the ballpark for peak SPL and probably average SPL. The latter is recording dependent as well.

Don't overthink... when you can listen to a level for an hour and getting the idea you can get a little louder you are generally more than 'safe'.

 

[F1308]

Calculating exact SPL may not be exact but when all aspects are taken into consideration you can get a pretty good idea.
When one is a calculus nerd it may be fun to do all the calculations.

I believe @Robbo99999 once attempted to estimate how loud one was listening by using sensitivity numbers, a DAC with a known output voltage and amp with 1x gain.

Now and then play something loud (impressively loud) for the duration of a song.
In general, when you get the urge of turning the volume down you are probably playing too loud.

The danger lies is the 'automatic gain control' in your hearing. When one has visited a loud (pop/rock) concert one knows it is loud yet stays there for hours.
Once you left the venue you might have noticed traffic is not 'loud' and the world has gotten a bit 'quieter' around you.
Likewise, late in the evening or early in the morning you can hear a clock tick but not during the day.

Therein lies the problem with 'sensible levels' based on the urge to dial down.
In the morning or late in the evening (while having a quiet evening) you automatically do not play as loud as during the day.

Still... in general when you get the urge to dial down the volume (after 1 or 2 songs) you are playing above 80-85 dB SPL average.

Note that a sensitive headphone 120dB/V or higher can reach non-sensible levels even from a phone/tablet/PC or player.
Sometimes when I am next to some young bloke and I can clearly hear the IEMs I wonder if they even get the urge to dial down the volume.


Ultimately measuring the output voltage under actual load (not easy to do, requires some knowledge) or measuring it with suitable calibrated gear IRL is the only real way to do this.

But.. as said.. lots of calculations, FR plots, accurate data of which one knows how it has been obtained and a PC with level monitoring, DR ratings of a recording etc. will get you in the ballpark for peak SPL and probably average SPL. The latter is recording dependent as well.

Don't overthink... when you can listen to a level for an hour and getting the idea you can get a little louder you are generally more than 'safe'.

Sure, but again, we are talking about "my issue" and that it took about 1/4 of volume travel, not micro-movements...

 

[Robbo99999]

Calculating exact SPL may not be exact but when all aspects are taken into consideration you can get a pretty good idea.
When one is a calculus nerd it may be fun to do all the calculations.

I believe @Robbo99999 once attempted to estimate how loud one was listening by using sensitivity numbers, a DAC with a known output voltage and amp with 1x gain.

Now and then play something loud (impressively loud) for the duration of a song.
In general, when you get the urge of turning the volume down you are probably playing too loud.

The danger lies is the 'automatic gain control' in your hearing. When one has visited a loud (pop/rock) concert one knows it is loud yet stays there for hours.
Once you left the venue you might have noticed traffic is not 'loud' and the world has gotten a bit 'quieter' around you.
Likewise, late in the evening or early in the morning you can hear a clock tick but not during the day.

Therein lies the problem with 'sensible levels' based on the urge to dial down.
In the morning or late in the evening (while having a quiet evening) you automatically do not play as loud as during the day.

Still... in general when you get the urge to dial down the volume (after 1 or 2 songs) you are playing above 80-85 dB SPL average.

Note that a sensitive headphone 120dB/V or higher can reach non-sensible levels even from a phone/tablet/PC or player.
Sometimes when I am next to some young bloke and I can clearly hear the IEMs I wonder if they even get the urge to dial down the volume.


Ultimately measuring the output voltage under actual load (not easy to do, requires some knowledge) or measuring it with suitable calibrated gear IRL is the only real way to do this.

But.. as said.. lots of calculations, FR plots, accurate data of which one knows how it has been obtained and a PC with level monitoring, DR ratings of a recording etc. will get you in the ballpark for peak SPL and probably average SPL. The latter is recording dependent as well.

Don't overthink... when you can listen to a level for an hour and getting the idea you can get a little louder you are generally more than 'safe'.

Yep, I worked it out here at this post:

*NEW* AKG K702 has been measured by Oratory!

Hi, like the title says, Oratory measured my K702 I sent him, I've got two pairs and I sent him my old pair from year 2015 (made in China, not Austria). Yes, so Oratory published the result today: https://www.dropbox.com/s/033kxyf8a0o3iv6/AKG K702.pdf?dl=0...

www.audiosciencereview.com

(in case anybody else wants to do something similar)

 

[xnor]

Ironically, I have added lots of things to the spreadsheet today including output impedance. Just squashed the last bugs.

@F1308 @solderdude explained why output impedance can make a huge difference.
If you set the amp Zout in the spreadsheet then it should give you correct numbers now.

The last big thing left to do is looking into the accuracy of sensitivity/impedance in combination with EQ presets that target the Harman target curves.

 

[xnor]

The danger lies is the 'automatic gain control' in your hearing. When one has visited a loud (pop/rock) concert one knows it is loud yet stays there for hours.
Once you left the venue you might have noticed traffic is not 'loud' and the world has gotten a bit 'quieter' around you.
Likewise, late in the evening or early in the morning you can hear a clock tick but not during the day.

Yep, our hearing also tries to protect itself by triggering the stapedius reflex, which reduces the vibrational energy transmitted to the cochlea.
It can be triggered in a very wide range from 70 to 100 dB SPL.

Years ago some research was published showing that the reflex can be triggered at low SPL with in-ears that do not have ports that let static pressure escape.
Anecdotally, I've always hated in-ears for that reason.

Also, there's a high-fidelity argument for lower listening levels: while SNR will be lower it is also true that our hearing distorts more and more with increasing SPL.

 

[F1308]

Ironically, I have added lots of things to the spreadsheet today including output impedance. Just squashed the last bugs.

@F1308 @solderdude explained why output impedance can make a huge difference.
If you set the amp Zout in the spreadsheet then it should give you correct numbers now.

The last big thing left to do is looking into the accuracy of sensitivity/impedance in combination with EQ presets that target the Harman target curves.

Great.
When finished, get data for a Roland Fantom 8 and the IEMs I mentioned and see if the answers match those settings at 11 and at 1...
Then the computations had nailed it.

 

[xnor]

When finished, get data for a Roland Fantom 8 and the IEMs I mentioned and see if the answers match those settings at 11 and at 1...

Since there are zero specs on the headphone jack, this is impossible to figure out other than measuring it on the device.
If the volume steps were constant then we could roughly estimate the output impedance given your two data points, but this is probably not possible either because the volume control step from 0 to 1 is very likely not the same as from 10 to 11.
The integrated amp could also have load impedance sensing, switching gain with headphones...

 

[Soundescape]

The last big thing left to do is looking into the accuracy of sensitivity/impedance in combination with EQ presets that target the Harman target curves

Yes as I've tried the last spreadsheet version and just transposing the negative EQ gain in the headroom it seems to me too much for the output values that I see.

 

[F1308]

Since there are zero specs on the headphone jack, this is impossible to figure out other than measuring it on the device.
If the volume steps were constant then we could roughly estimate the output impedance given your two data points, but this is probably not possible either because the volume control step from 0 to 1 is very likely not the same as from 10 to 11.
The integrated amp could also have load impedance sensing, switching gain with headphones...

Cheeses...

 

[xnor]

I've added a new V2 sheet. Loudness equalization is a lot easier to configure now: enter the general loudness target and you're done.
That's because the music parameters now also include BS-1770 loudness, as calculated in #17.

I've set the loudness target to -20 LUFS such that both types of music would be properly loudness-equalized without any (potentially clipping) boosts.
This will obviously not work if you use a service that uses a higher target and listen to both types of music. The jump in loudness will be shown in red in the music parameters.

All the fields now also have notes with explanations. And I've renamed "headroom" to "attenuation", because that's what it is.

 

[Soundescape]

Often it is a little bit hard to understand how much negative gain to use in the EQ and what is the effect of the IR normalization and their combined effect on loudness:

Normalize convolution presets to prevent clipping · Issue #388 · jaakkopasanen/AutoEq

I assumed that the convolution filters would be tuned to prevent digital clipping; however, with no preamp or any other audio processing, I notice digital clipping with certain audio/video files. T...

github.com

Convolver auto-gain · Issue #1309 · wwmm/easyeffects

EasyEffects Version 6.1.5 What package are you using? Fedora Distribution Fedora 35 Describe the bug When using the convolver, an unpredictable amount of gain is applied to the IR kernel, and there...

github.com

So It could be hard to indentify a good value for the attenuation cell in the spreadsheet.

 

[Soundescape]

I think that some services, like Spotify, are going to apply both positive and negative gain compensation. You can also chose the normalization level (-11, -14, -19 LUFS):

Loudness normalization - Spotify

artists.spotify.com

 

[Soundescape]

For example, if you listen to the modern pop/rock track on Tidal, Amazon, YouTube or Spotify, there's an additional digital attenuation of 6 dB. That's -6 dB.
The classical track on the other hands needs to be boosted by +5.6 dB.

Wait.. what? How can a player boost the classical track that has sample peaks near 0 dBFS by nearly ~6 dB? It' cant, except if you accepted an extreme amount of clipping or engaged dynamic range compression.
That's the tradeoff that these services made: the higher target level results in less attenuation but it also means that silent tracks cannot be properly loudness-equalized.

I don't think that we have this kind of effect on Spotify:

Positive gain is applied to softer masters so the loudness level is -14 dB LUFS. We consider the headroom of the track, and leave 1 dB headroom for lossy encodings to preserve audio quality.
Example: If a track loudness level is -20 dB LUFS, and its True Peak maximum is -5 dB FS, we only lift the track up to -16 dB LUFS.

 

[xnor]

I don't think that we have this kind of effect on Spotify:

Thanks for the hint. I had just copied the -14 dB LUFS value without doing any further research.

Turns out we are both right because the loud option engages DRC:

• Loud: -11dB LUFS
  Note: We set this level regardless of maximum True Peak. We apply a limiter to prevent distortion and clipping in soft dynamic tracks. The limiter’s set to engage at -1 dB (sample values), with a 5 ms attack time and a 100 ms decay time.
• Normal: -14dB LUFS
• Quiet: -19dB LUFS

A limiter is a special case of a DRC.

Also, there is this peculiarity that you quoted:

Negative gain is applied to louder masters so the loudness level is -14 dB LUFS. This lowers the volume in comparison to the master - no additional distortion occurs.

Positive gain is applied to softer masters so the loudness level is -14 dB LUFS. We consider the headroom of the track, and leave 1 dB headroom for lossy encodings to preserve audio quality.
Example: If a track loudness level is -20 dB LUFS, and its True Peak maximum is -5 dB FS, we only lift the track up to -16 dB LUFS.

Source: https://artists.spotify.com/help/article/loudness-normalization

So in the worst case a very silent master with 0 dBFS peak will get attenuated by another dB to create some headroom for the lossy codecs used in streaming, I guess.
I think it's time for Spotify lossless.


Anyway, the spreadsheet currently doesn't support the "loud" setting because the limiter changes the music parameters. This is a minor point that I'll look into some day.
I'll update the notes.

 

[xnor]

V3 adds almost 700 headphones using the data (sensitivity, impedance) from DIY-Audio-Heaven with permission from @solderdude! Thanks again.

 

[GaryH]

Yep, our hearing also tries to protect itself by triggering the stapedius reflex, which reduces the vibrational energy transmitted to the cochlea.
It can be triggered in a very wide range from 70 to 100 dB SPL.

Years ago some research was published showing that the reflex can be triggered at low SPL with in-ears that do not have ports that let static pressure escape.

Interesting, do you have a link to that paper?

 

[Robbo99999]

Interesting, do you have a link to that paper?

Are you more interested in that last part? "Years ago some research was published showing that the reflex can be triggered at low SPL with in-ears that do not have ports that let static pressure escape."

What do you think the real world implications of that would be? Perhaps it could affect perceived frequency response at even low SPL? If that's the case then I suppose your experience wouldn't equal that of the dummy head measurements.

 

[xnor]

Interesting, do you have a link to that paper?

I think it's this one: https://www.aes.org/e-lib/browse.cfm?elib=15786

Abstract:

When a sound producing device such as insert earphones or a hearing aid is sealed in the ear canal, the fact that only a tiny segment of the sound wave can exist in this small volume at any given instant, produces an oscillation of the static pressure in the ear canal. This effect can greatly boosts the SPL in the ear canal, especially at low frequencies, a phenomena which we call Trapped Volume Insertion Gain (TVIG). In this study the TVIG has been found by numerical modeling as well as direct measurements using a Zwislocki coupler and the ear of a human subject, to be as much as 50dB greater than sound pressures typically generated while listening to sounds in an open environment. Even at moderate listening volumes, the TVIG can increase the low frequency SPL in the ear canal to levels where they produce excursions of the tympanic membrane that are 100 to 1000 times greater than in normal open-ear hearing. Additionally, the high SPL at low frequencies in the trapped volume of the ear canal, can easily exceed the threshold necessary to trigger the Stapedius reflex, a stiffing response of the middle ear, which reduces its sensitivity, and may lead to audio fatigue. The addition of a compliant membrane covered vent in the sound tube of an insert ear tip was found to reduce the TVIG by up to 20 dB, such that the Stapedius reflex would likely not be triggered.

(emphasis by me)

Basically, the tighter the seal and the smaller the enclosed volume of air the stronger this effect gets. Worst case is probably deep-insertion in-ears that completely seal the ear canal.

 

[GaryH]

Are you more interested in that last part? "Years ago some research was published showing that the reflex can be triggered at low SPL with in-ears that do not have ports that let static pressure escape."

What do you think the real world implications of that would be? Perhaps it could affect perceived frequency response at even low SPL? If that's the case then I suppose your experience wouldn't equal that of the dummy head measurements.

Yeah, it might be one of the reasons why Harman found the preferred bass for in-ears to be higher than for headphones, because the stapedius reflex mostly reduces transmission at lower frequencies, so listeners may have compensated for this by turning up the bass control. The paper linked above does also say though:

The stiffening of the oscular chain and the tympanic membrane brought about by the stapedius reflex reduce the sensitivity of the hearing system to other frequencies (especially in the midrange)

(Maybe this is a contributor to the flattening of the equal loudness contours in the midrange with increasing SPL.)

Also of note is the prevalence of the reflex was actually found to only be ~75% in this study, but more common at ~85% for younger listeners ages 18-30. This may (partly) explain the variation in preferred bass Harman found with in-ears:

And also the higher preferred bass by younger listeners in their over-ear tests (which may still have triggered the stapedius reflex for some, as people's trigger SPL threshold ranges from ~70-100 dB, and Harman's listening level is within this range):

Interestingly, the paper @xnor linked also mentions infrasound (which I've previously posited could be contributing to perceptions of bass impact):

Sound at frequencies lower than 20 Hz is known as infrasound.[33,34] Most people cannot hear these very low frequencies, but may feel them as vibrations. Although the experimental results presented above were only measured down to 20 Hz, the top end of this range, it is clear that insert headphones of the type tested will be able to produce frequencies in the infrasound range. Additionally, the trapped volume effect of the ear canal will boost these infrasound frequencies via oscillating static pressure effects. The infrasound content of recorded music and other audio material is not typically reported, but measurements reveal the presence of significant spectral content below 10 Hz. Normal, open air sound equipment like home and car stereo systems typically cannot produce much output volume below about 50 Hz, and thus the low frequency content of recordings is not commonly heard. However, insert headphone, can produce these low frequencies, and these frequencies are dramatically boosted by static pressure oscillations in closed volumes, such as the ear canal.

This boosted infrasound would more readily trigger the stapedius reflex. Initially you'd think this would mean less perceived bass impact, but it might be more complicated than that. The reflex has a letency of ~10 milliseconds, but maximum tension of the intra-aural muscles involved may not be reached for 100 milliseconds or more, too late for sudden, intense sounds (e.g. percussive sounds like a kick-drum). Also, this tension reduces by about 50% after a few seconds. This could result in repeated initiation of the stapedius reflex when listening to music, with the sound during periods after sudden loud (e.g. percussive) parts sounding quieter, but not those sudden loud parts themselves, due to the latency of the reflex, and its relaxation. This could then be an explanation for an increased perception of percussive bass impact, and could even explain @Resolve 's (and others') perception that some headphones (which it seems mostly have well-sealed front volumes) have greater 'macrocontrast/macrodynamics' than others with a similar response above 20 Hz - because this repeated triggering of the stapedius reflex would literally change the dynamic range of the ear as you're listening to music (from the paper xnor linked again):

Above [the reflex] threshold tightening of the stapedius muscle compresses dynamic range.

Unless sub-20 Hz responses of such headphones are measured, we won't be able to investigate this potential influence of infrasound and static pressure on perceived sound further.