[NotARobot]

What exactly are we inferring when we describe sound signature as "neutral"?
 
I think of headphones/speakers with a strong bass or "V-shaped sound" to be incompatible with "reference". They supposedly exaggerate parts of the frequency band. I also associate "neutral sound" with "representational faithfulness". But is this really the case? Bass, by its nature, is very powerful. At orchestra concerts, when they hit the bass drum, the entire hall reverberates and the bass section often stands out over the other instruments. High frequencies also stand out naturally. I actually think that my Ultrasone Pro 900, commonly described as having a "V-shaped sound", sounds extremely realistic, especially with piano and orchestral music.
 
If a V-shaped sound actually provides the faithful reproduction (in my opinion, of course), then does so-called "neutral" equipment actually subdue the bass/treble frequencies? Or is a V-sound actually part of the definition of neutrality?

 

[spruce music]

Neutral should sound exactly like the signal fed into it.  A v-shaped response by definition isn't neutral.  It may be preferred, and preferences are okay, but it is not neutral. 
 
Harman is doing research into what a neutral sounding headphone will be.  Like speakers a truly flat response isn't perceived as neutral.  An even response is however, and it isn't going to be V-shaped.  Not with good recordings as the source.

 

[castleofargh]

a neutral frequency response means it follows a reference frequency response in a given frequency range. for our amateur audio use, the reference sound of the album is decided at the mastering and unless you can hear it, you have no real mean to know how the album should sound. thinking we do is usually just wishful thinking. so judging how some gear sounds based on such unknown reference... that doesn't work very well.
if you change the reference for "how I feel the album should sound", then only you can now tell if it's neutral or not for you. it's a new reference.
 
 electrically neutral in the audible range for a DAC or an amp is the expected meaning of neutral for such devices. so if there is no precision, I will think the person is talking about electrical neutral. but ideally it would be better to say the exact range, the accuracy (+/-1db or something like that), if the measure is weighted or not... the better the reference, the better we know what neutral really means.
for a headphone, there is no unique reference because people have different bodies and ears. so you absolutely need to tell what you refer to when you call something neutral(or V shaped for that matter).
it can be "to my hear", it can be neutral in a diffuse field compensation, neutral for harman target of preferred headphone sound, electrically neutral, neutral for my grandma because she likes the color....   without reference it means nothing.
 
now you mention V shaped(I assume you mean V shaped compared to something like innerfidelity's compensated graphs?), one of the possibilities for such preference could be https://en.wikipedia.org/wiki/Equal-loudness_contour . if you listen quietly, your ear starts sucking at picking up low and high frequencies, so we often find the need to boost both ends for such a reason when we listen quietly. are you that kind of listener? but maybe you just enjoy music more when you have good rumble and clear trebles? maybe the usual diffuse field reference is just not good for your ears because you don't have an average ear(I know I don't)? maybe it has to do with the type of music you listen to the most? rap music is rarely mastered the same way classical music is.
in any case it's all good. don't let others tell you what is right for your ears, and in return don't try to convince other that your preference is some grand ideal of fidelity and righteousness like some self proclaimed elite audiophiles do. if the reference is you, you're the boss. ^_^

 

[RRod]

  a neutral frequency response means it follows a reference frequency response in a given frequency range. for our amateur audio use, the reference sound of the album is decided at the mastering and unless you can hear it, you have no real mean to know how the album should sound. thinking we do is usually just wishful thinking. so judging how some gear sounds based on such unknown reference... that doesn't work very well.
if you change the reference for "how I feel the album should sound", then only you can now tell if it's neutral or not for you. it's a new reference.

 
Pretty much this. If I put you in several different mastering studios and measured impulse responses at the ear canal, you'd probably get a decent bit of variation. Until there is a consensus there, there's little hope of saying what a "neutral" headphone signature should be. So people tend to take a negative definition "these headphones a NOT neutral because X".

 

[spruce music]

http://www.innerfidelity.com/content/harman-researchers-make-important-headway-understanding-headphone-response#EWOI9lhXrupe77eO.97
 
Interesting read on the matter of headphones target response.

 

[Voyageur]

Thanks for sharing the link, very comprehensive article.

 

[zombicube]

Someday I'd like to get my ears tested to determine which frequencies I have "lost" over the years, and by how much, so that I could boost those by the right amount (with an EQ) and artificially restore my young ears. Because I guarantee you that even with a perfect player and perfect headphones or speakers, a neutral sound would not sound neutral to me--would be interesting to have the data to "cancel out" the physiological changes/damage so a "perfect" system would sound perfect to me. Of course, I can always just play with the EQ until the music sounds best, but that would not be nearly as nerdy as what I have in mind.

 

[castleofargh]

  Someday I'd like to get my ears tested to determine which frequencies I have "lost" over the years, and by how much, so that I could boost those by the right amount (with an EQ) and artificially restore my young ears. Because I guarantee you that even with a perfect player and perfect headphones or speakers, a neutral sound would not sound neutral to me--would be interesting to have the data to "cancel out" the physiological changes/damage so a "perfect" system would sound perfect to me. Of course, I can always just play with the EQ until the music sounds best, but that would not be nearly as nerdy as what I have in mind.

but that's the counter intuitive thing about human hearing, most likely your brain has already adapted to your hearing loss. you spend your days with sounds that you have heard all you life, so your brain adapts to make them sound ass close as possible to what "they should sound like" from the perspective of your brain. it's a little similar to using tainted glasses. after an hour or so, your brain has compensated most of the tainted color and it's when you take the glasses off that for a moment the world seems to have lost it's usual calibration. if the glasses are green, then the world without the glasses should have too much magenta if I remember my color chart correctly. the brain constantly adjusts to what is the new normal, and your actual hearing is your own new normal. trying to compensate for it wouldn't be the answer unless you can compensate all day long like with hearing aids, and not just when you want to listen to music.
so again it's counter intuitive, but a good almost neutral speaker in a treated room should still be the most realistic and neutral experience for all of us when playing an album.
 
for headphones it's more complicated because you need to compensate for HRTF, because all sounds in real life are compensated that way by your body, except headphone sounds. so the brain has a hell of a time trying to make sense of any headphone(and that's added to the wrong stereo if the album was mastered for speakers).

 

[upstateguy]

  <snip>
 
......  I can always just play with the EQ until the music sounds best......

 
What's wrong with that?

 

[goodvibes]

Ima getting flamed (ducks) but neutral extends beyond measured response. Every decent amp measures flat but they don't all sound the same with load. Similarly measuring transducers don't always sound similar. If a certain frequency isn't damped as quickly or happens to interact with what it's driving or what's driving it, it will call attention to itself and sound unlinear. Why I have issue with steady state tests as an absolute of goodness.

 

[castleofargh]

  Ima getting flamed (ducks) but neutral extends beyond measured response. Every decent amp measures flat but they don't all sound the same with load. Similarly measuring transducers don't always sound similar. If a certain frequency isn't damped as quickly or happens to interact with what it's driving or what's driving it, it will call attention to itself and sound unlinear. Why I have issue with steady state tests as an absolute of goodness.

 
 if a device is electrically flat unloaded and not a mess with distortions, then it's "neutral" in that sense unloaded. that measure never pretended to say that it would be neutral with a headphone plugged in. that's something that is poorly concluded by people who think all they need to read a graph is to look at the shape of the line.  I'm sure the all market hopes for people to erroneously interpret it that way, but it's not the unloaded measurement's fault. measure one condition, get a result for that one condition. that was always the deal ^_^.
same thing for anything measured without the rest of the chain, measuring an incomplete electrical circuit, means measuring a different circuit.
 
for headphones, again everything is more complicated because they can have way too much distortions, can react to, or impact the amp's signal, and can have all kinds of specs without any clear range we can expect them to stick to. so the way headphones have been measured may not clearly represent the conditions of them on your ears on your system. but if we measured a flat frequency response with mics in your ears from 2 headphones, then I suspect there wouldn't be that much differences between them anymore as FR is still a main factor for what I hear. but if it still is clearly different then we can measure that there is effectively some massive and audible ringing or distortions in at least one of the headphones in the conditions of our use. my point is, it's not magic. if we can measure the right things, we get relevant information.
 
I understand your post, I really do and you're not saying anything wrong. I also complain about how we need different and more measurements all the time. and have some mandatory standards for the way manufacturers write nomenclatures so that it's not just another marketing tool hiding just enough not to be sued, but misleading on purpose anyway. so I really get your point, I want to insist on this. but I also fear that some people might read your post as an anti measurement argument.
so for them, we measure something in a given set of conditions, and the readings don't necessarily stay true out of those conditions. it's our fault for making the wrong measures or expecting them to mean something they never did. but measurements are good at what they do.