[Quaintative]

Quote:

Correct.  You absolutely DO NOT want to equalize out your equal loudness curve.  That curve set is built into human hearing, you don't want to equalize it out, and besides, it's level dependent. 
 
The curve you'll get from this test will be the equal-loudness response of your own hearing with the mask of the response of the headphones/IEMs you used for the test.  The results are pretty meaningless and not useful at all for setting an equalizer.  At best the results present something that is of interest to look at. 
 
The target response curves of the three types of headphones plus IEMs are all different, but none are flat, and certainly none are related to the equal-loudness curves.  Headphone target curves are there to compensate for the difference in the way sound is presented to the hear by headphones vs a free field.  
 
You can experience highly accurate and detailed headphone EQ with the Audyssey "amp" app, and some of the others.  Audyssey also features their dynamic EQ, which compensates for the effect of the change in hearing response at low levels relative to the spectral balance at the level at which music is mixed.  
 
Using an equalizer to compensate for an individual's hearing difference is also not a great idea because every individual has a build-in established "normal" based on every-day hearing.  Attempting to compensate for a deficiency, unless large enough to be considered a hearing loss, will always result in unnatural balance because of its contrast with what's heard in life. It's for this reason that people who have hearing aids prescribed are encouraged to use them all day, every day, or they will always sound "wrong". 

 
Quote:

Consider going to a live performance. If during this performance, tones where reproduced for the audience, the different frequencies will be perceived with different loudness levels (as illustrated by the loudness contours.) Does that mean that one needs to somehow equalize a live performance, or for that matter the entire world, to "correct" for loudness levels? I don't believe so.
 
The job of a high fidelity audio reproduction system is to reproduce a live performance, or a particular set of effects, as close as possible to the original source. This implies that the lumped sound system frequency response should be as flat as possible so as to "get out of the way."
 
If equalizing using the loudness contours, what will happen is that a violin or any other instrument played through the equalized rig will sound substantially different than a "real world" violin or any other instrument.

 

Sure.... but if you make the headphones produce an equal balance curve, then in terms of absolute energy (as measured by energy output) reconstitute the equal loudness curve? In my mind, that would be getting the headphones "out of the way". And I know it's level dependent (as shown in the ISO curves above), I can't control for that yet, but I'm sure I could figure it out. Replaygain would be an interesting first step to get a global range value to work from.... I'll check out the Audyssey application.
 
As far as personalizing curves.... headphones are personal enough so if I personalize them to me.... that's not exactly a bad thing . Besides, sounding "right" or "wrong" is dependent on the coloring of the headphones anyway....
 
And in more practical terms, I don't think that it's physically possible for a headphone to accurately reproduce live frequencies because beyond a certain frequency low point, the actual sound pressure will decay off a headphone faster than it would on a speaker (or live, for that matter).
 
In terms of live performances.... the microphone doesn't impart the equal loudness curve. Our auditory systems do. So the live performance wouldn't be corrected at all.

 

[Quaintative]

Quote:

Another explanation:
 
Let's take any instrument as sound source. We record this with a perfect mic (completely flat frequency response). We replace the instrument with a perfect speaker and play the recording. We record this reproduction with the perfect mic again and get exactly the same as the initial recording.
 
So what's the problem? The problem is your torso, head, outer ear, ear canal ... All of this changes the frequency response of what actually arrives at your eardrums, aka head related transfer function (HRTF). Different types of equalization (diffuse, free, independent-of-direction) try to account for that by reversing the effects.
 
Ideally, you'd measure the instrument or perfect speaker with a perfect mic at your eardrum and compare it to the measurement of your headphones. You'll see a resonance around 3 kHz etc. That is what you need to equalize for, not equal loudness of pure tones.

Sure, what is physically measured is different from what's physically perceived.
 
I was looking at the headroom frequency curves and on a whim, I checked out their 50Hz square wave response. Every single headphone couldn't recreate a 50Hz square wave. I thought about it, and it made sense. A 50Hz square wave requires pressure to be maintained over 20ms, which is basically impossible to do with a headphone because 1) it isn't designed to sustain that sort of pressure, and 2) the chamber that it's trying to sustain that pressure in is naturally leaky.
 
And I thought about it, and I think that's one big difference between headphones and speakers - speakers are better able to sustain pressure at lower frequencies by the simple fact that they move a lot more air. And I remembered one time when I was experimenting and I was listening to music through my open-air headphones and speakers at the same time. If I balanced them in a certain way, I could feel the music through the speakers and get the detailed treble through the headphones. It was pretty terrific sounding.
 
To add another thought to this:
When we're tuning and building our systems, what is our goal? is it:
1) Reproduce live music
2) Reproduce what the producer intended us to hear?
3) Reproduce what we personally feel the most comfortable with?
4) Reproduce what we think everybody should be comfortable with?
 
I think that each one of these philosophies would result in different build strategies and different manipulations of the sound at output.

 

[jaddie]

Quote:

Sure.... but if you make the headphones produce an equal balance curve, then in terms of absolute energy (as measured by energy output) reconstitute the equal loudness curve? In my mind, that would be getting the headphones "out of the way". 

No, you have to understand what the equal loudness curves are.  They are graphs of hearing response, not targets for headphone response.  In building headphone response, it's not important that any one curve is not flat, because that's a contour that is applied by hearing to everything, is uniform, and the final link in the audio chain.  You don't want to compensate for the fact that they aren't flat.  The important part of the curve families is the difference between them a different levels.  If you were to normalize them for a specific mid-band SPL of say 85dB SPL, and plot only the difference in hearing response as level changes you would have a set of values for each frequency that would correspond to inverse of the correction that would need to be applied if the play level were offset from the original performance or mix level by that specific amount.  If the play level and mix or performance levels were identical there would be no need for correction at any time.  Its only if play and performance levels are offset, as they certainly would be listening to most popular music at background levels, or listening to a movie soundtrack a lower than reference level.  
 
Quote:

And in more practical terms, I don't think that it's physically possible for a headphone to accurately reproduce live frequencies because beyond a certain frequency low point, the actual sound pressure will decay off a headphone faster than it would on a speaker (or live, for that matter).

There's no scientific basis for that statement.  It's entirely possible for a transducer to reproduced any sound pressure level at any point in time along an envelope.  There's no reason for headphones to force an envelope to decay faster than speakers or live.  However, the perspective is different.  If anything, they probably enhance a reverberant space.
Quote:

In terms of live performances.... the microphone doesn't impart the equal loudness curve. Our auditory systems do. So the live performance wouldn't be corrected at all.

This is correct.  Live performances are what they are. The equal loudness correction only becomes necessary when we play recordings at levels other than the original, where we would then be operating along an entirely different equal loudness curve than originally. That's exactly what loudness compensation set out to do, and exactly why it fails when applied as a fixed contour.  The curves only represent measurements, the actual moment to moment response of hearing changes differently at each frequency.  This is why Audyssey Dynamic EQ and Dynamic Volume are so impressive.  The don't apply fixed curves, and the are aware of the exact and specific SPL all frequencies at all times, just like hearing is.  These systems are the only ones that don't over or under compensate.
 
In other words, if you present a sound field from a forward direction in a medium sized acoustic space (with no acoustic issues), natural sound results from flat response plus loudness correction for play level offset from the original.  The forward position of speakers mimics the situation of people listening to a performance in front of them at a distance, and the natural response contours imposed by our heads, chests, and pinna are present. However, when sound is presented at 90 degrees, extremely near-field, the pinna, head and chest effects are all bypassed, and there's a near-field effect present, all of which dictates the non-flat target curve required for natural sound in headphones.  Similarly, IEMs present sound even more near-field, and as a pressure transducer in the hear canal, which again requires a different target curve to sound natural. To all of that we need to add loudness compensation that is correct for the specific frequency at the specific SPL at the correct moment in time.  If all of that is done exactly right, equalization to personal preference would be unnecessary unless we would like to modify the spectral balance choices made in the original recording.  

 

[xnor]

@Quaintative: I don't think accurately reproducing squares waves is important.
 
As for those goals, why not just accurately reproduce what is on the CD?

 

[jaddie]

Quote:

Sure, what is physically measured is different from what's physically perceived.

...because of the effects of the outer ear, inner ear, head, chest, and very significantly, angle of incidence.  However, within limits, measurements do agree quite well.
 
Quote:

I was looking at the headroom frequency curves and on a whim, I checked out their 50Hz square wave response. Every single headphone couldn't recreate a 50Hz square wave. I thought about it, and it made sense. A 50Hz square wave requires pressure to be maintained over 20ms, which is basically impossible to do with a headphone because 1) it isn't designed to sustain that sort of pressure, and 2) the chamber that it's trying to sustain that pressure in is naturally leaky.

Square wave response measurements are VERY difficult to correlate with what is heard.  I've presented to this forum before several examples of identically sounding square waves that look completely different from each other. What you say regarding sustaining pressure is true, but the fact that a square wave has tilt to it does not match up well with how it sounds.  Square waves are interesting diagnostic tools, but very poor for qualitative judgement.
Quote

And I thought about it, and I think that's one big difference between headphones and speakers - speakers are better able to sustain pressure at lower frequencies by the simple fact that they move a lot more air. And I remembered one time when I was experimenting and I was listening to music through my open-air headphones and speakers at the same time. If I balanced them in a certain way, I could feel the music through the speakers and get the detailed treble through the headphones. It was pretty terrific sounding.

Actually, headphones have more ability to sustain pressure at low frequencies than speakers, particularly well-sealed IEMs.  But that's not what's responsible for what you heard.  Speakers do move more air, and that pressure wave hits your entire body and the clothes that cover it, and the seat you sit on.  The impact of a bass wave on your body is a significant part of the experience, and headphones alone can't provide that.
 
Quote

   
To add another thought to this:
When we're tuning and building our systems, what is our goal? is it:
1) Reproduce live music
2) Reproduce what the producer intended us to hear?
3) Reproduce what we personally feel the most comfortable with?
4) Reproduce what we think everybody should be comfortable with?
 
I think that each one of these philosophies would result in different build strategies and different manipulations of the sound at output.

 
1 is pretty much impossible in a practical sense.  Reproduced music is, at best, an acceptable representation of a live event.  The primary limits are a fixed number of channels and transducers, and a very different acoustic space.  If you had a similar or identical space as the original performance and had a speaker for each instrument, you'd have a shot.  Otherwise, it's all a simulation requiring the suspension of disbelief.  Floyd Toole treats this quite well in the first few chapters of "Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms".
 
2 is not only achievable, but is the primary goal of THX.  With film sound in particular, mixing spaces are very well standardized and controlled.  Replicating them in a consumer's environment is completely possible.  THX standardization hits the required SPL and response targets for the room size, there are industry standard calibration levels, and soundtracks end up being quite transportable.  Once all of that is under control, correcting for a level offset is a minor issue.  Music is a bit more wild in terms of standards, particularly mix levels, but there is an industry average that isn't all that different from film work.  The channel count is different of course, but it's not impossible to replicate a recording studio control room sound field in the home.  Not cheap, but not impossible.
 
Once you hit 2, 3 and 4 end up being very minor issues.  It turns out, if you present the general population with neutral uncolored sound that covers the audio spectrum well at sufficient volume, they all like it.  The preference biases show up when we move away from the neutral and uncolored, then mix in acoustic issues.  

 

[xnor]

@Quaintativeunno if you skipped #14, but I want to add to that that what arrives at the eardrum is not translated 1:1 into what you hear.
 
Let's take an instrument that produces two pure tones 85 Hz and 1 kHz, each at exactly 80 dB SPL. Let's ignore the pinna. ear canal, air ...  for a moment and assume these tones arrive at exactly the same level at your eardrum. Despite that, what you will hear is that the 1 kHz tone is a lot louder. According to the equal loudness contours about 20 dB louder. That's the way this instrument sounds to us, end of story. Any amount of boosting bass using an EQ will make it sound different.

 

[ultrabike]

Quote:

Sure.... but if you make the headphones produce an equal balance curve, then in terms of absolute energy (as measured by energy output) reconstitute the equal loudness curve? In my mind, that would be getting the headphones "out of the way". And I know it's level dependent (as shown in the ISO curves above), I can't control for that yet, but I'm sure I could figure it out. Replaygain would be an interesting first step to get a global range value to work from.... I'll check out the Audyssey application.

 
I don't think that is how it works. Suppose you record a live performance of a piano. When you hear the live performance, there is no equal loudness compensation anywhere. If you play your recording and apply an equal loudness curve through your speakers or headphones, it will distort the original signal. Equalization is best used to compensate for issues in our audio reproduction rig, not to compensate for the way we hear sounds in real life.
 
 

And in more practical terms, I don't think that it's physically possible for a headphone to accurately reproduce live frequencies because beyond a certain frequency low point, the actual sound pressure will decay off a headphone faster than it would on a speaker (or live, for that matter).

 
 
Most decent headphones will be able to reproduce low frequencies better than most speakers with ease. I believe this is partially due to the fact that headphones need to move less air than speakers in order to attain a particular SPL around or inside our ears. The headphone cup (or an IEM tip that seals the canal) provides a certain acoustic impedance that facilitates this. A headphone needs to move air between the driver and the ear, while a speaker needs to drive the whole room that houses it.
 
Many affordable headphones reach 30 Hz with ease (consider the FR of the HD558 and HD202 below). Most speakers need a sub-woofer to do so.
 
 

 
 

 
Note: Mods, these FR plots are my own (not from another forum.)

 
In terms of live performances.... the microphone doesn't impart the equal loudness curve. Our auditory systems do. So the live performance wouldn't be corrected at all.

 
Your auditory system will impart an equal loudness curve the same way to both a live performance and a recording, so no need to compensate for it.
 
Quote:

Sure, what is physically measured is different from what's physically perceived.
 
I was looking at the headroom frequency curves and on a whim, I checked out their 50Hz square wave response. Every single headphone couldn't recreate a 50Hz square wave. I thought about it, and it made sense. A 50Hz square wave requires pressure to be maintained over 20ms, which is basically impossible to do with a headphone because 1) it isn't designed to sustain that sort of pressure, and 2) the chamber that it's trying to sustain that pressure in is naturally leaky.
 
And I thought about it, and I think that's one big difference between headphones and speakers - speakers are better able to sustain pressure at lower frequencies by the simple fact that they move a lot more air. And I remembered one time when I was experimenting and I was listening to music through my open-air headphones and speakers at the same time. If I balanced them in a certain way, I could feel the music through the speakers and get the detailed treble through the headphones. It was pretty terrific sounding.
 
To add another thought to this:
When we're tuning and building our systems, what is our goal? is it:
1) Reproduce live music
2) Reproduce what the producer intended us to hear?
3) Reproduce what we personally feel the most comfortable with?
4) Reproduce what we think everybody should be comfortable with?
 
I think that each one of these philosophies would result in different build strategies and different manipulations of the sound at output.

 
50 Hz square waves have no frequency content below 50 Hz. To verify bass it is best to check the FR. Square waves look odd many times because the phase is messed up or because the odd harmonics of the 50 Hz waveform are distorted by the headphone. A typical speaker might do much worse.
 
EDIT: Fixed a quote that seemed to be missing a frame, sorry about that...

 

[Quaintative]

Read everything so far, thinking....

 

[Quaintative]

I think I'll try this and see what happens:
http://www.head-fi.org/t/413900/how-to-equalize-your-headphones-a-tutorial
 
For me, the goal/purpose of thinking about replicating an equal-loudness curve on my headphones, is that it will allow me to better evaluate the characteristics of the headphone in the "relative" absence of frequency coloration. That way, I'll be better able to characterize the characteristics of the headphone (grain, speed, sustain and impact), without the artificial boosting that frequency coloration can give.

 

[xnor]

Yeah but equal loudness is colored in a both absolute and relative way. If there was a simple and accurate way to EQ headphones by using equal loudness contours everyone would do it.
 
I could think of a process like this:
- choose an equal loudness contour (SPL)
- generate a couple of sine waves at the equal loudness curve specific levels
- play those sine waves through your headphones at the chosen SPL
- enable an equalizer and adjust until all tones sound equally loud, save as preset
- add something like the diffuse field equalization to your EQ preset, save as final preset
- use final preset to listen to music
 
There are a couple of problems in the details, but basically you'd end up with a more or less perceived flat frequency response.

 

[Quaintative]

Yeah, no matter what, getting rid of/standardizing "color" will be a little screwy because well... frequency coloration is three dimensional, while EQ only fixes two dimensions.

 

[Skooter]

Some interesting "loudness curve" supplemental information, particularly relating to the Fletcher Munson Effect, is linked below.
 
http://www.hometheatershack.com/forums/electronic-processing-equalization-devices/12116-dynamic-eq-recommendations.html#post108351
 
For low volume listening on an ordinary hi-fi system with speakers, boosting low frequencies is a very effective method of increasing listening pleasure and compensating for our "ears."
 
The famous "loudness" button in old integrated amps tried to do this for an "average" listener.
 
But a modern digital eq allows one to adjust boost frequencies/levels according to his ears and system, and some systems reduce the boost as volumes increase. An excellent exercise if you have a few hours and want to learn a bit about audio or listen to music at lower volumes.
 
Unfortunately digital eq has a way of killing sound . . .but that is a different topic.

 

[jaddie]

Quote:

For low volume listening on an ordinary hi-fi system with speakers, boosting low frequencies is a very effective method of increasing listening pleasure and compensating for our "ears."
 
The famous "loudness" button in old integrated amps tried to do this for an "average" listener.

The "Loudness" control never worked right because for any loudness correction to be correct, it has to be related to a specific SPL difference between play and record level. The loudness control based on Fletcher-Munson curves was never right, usually too much, because it was not based on any specific SPL, and a few other reasons.
 
Quote:

But a modern digital eq allows one to adjust boost frequencies/levels according to his ears and system, and some systems reduce the boost as volumes increase. An excellent exercise if you have a few hours and want to learn a bit about audio or listen to music at lower volumes.
 
Unfortunately digital eq has a way of killing sound . . .but that is a different topic.

If you have an iOS device, you might try the Audyssey "amp" app.  If your headphones are in their list, they've been measured, and not only an excellent and accurate EQ curve provided, but accurate Dynamic EQ (that's their calibrated loudness compensation based on actual SPL at your ear) is provided. And you still have a bit of your own EQ to apply in the app too. Pretty cheap to try.
 
Your generalization about digital EQ "killing sound" is understood, but there are good and bad methods.  The good ones don't kill sound.

 

[jaddie]

Here, they describe the entire loudness issue better than I:
http://www.audyssey.com/audio-technology/dynamic-eq

 

[xnor]

Quote:

Yeah, no matter what, getting rid of/standardizing "color" will be a little screwy because well... frequency coloration is three dimensional, while EQ only fixes two dimensions.

Do you mean magnitude and phase response of the frequency response and distortion? Distortion definitely can be a problem when testing the equal loudness of very low frequency tones, which have to be played at very high SPL.