[money4me247]

  When I look at that chart, I see almost identical response out almost to 10kHz +/-3dB or so. Then a considerable variation beyond that. I apply that to what I know about human hearing... i.e.: the most important octaves for balanced response are the ones between 40Hz and 10kHz, 3dB is the just noticeable difference for frequency response variation in music, the last octave from 10kHz-20kHz is the least used octave in music... and it tells me that barring any significant distortion or problems with fit and comfort, these headphones would all sound pretty similar. Odds are the variation due to manufacturing tolerances between two headphones of the same make and model might be as different as the differences between these different makes and models.

hehe! i like your description thank you. I think I would agree with your assessment.
 
I have found that some people here sometimes go crazy about a certain peak, bump, or dip... and I am personally unsure how significant the impact of a slight measured variation between FR curves. I am personally leaning towards that you can probably accurately draw out an expected frequency response from just listening to a pair of headphones, but won't be able to capture the subtle variations. Perhaps those subtle differences does cause the subtle differences in sound between headphones, but I think beyond more big picture view, overanalyzing the slight variations may be more counterproductive. 

 

[bigshot]

"Pulling out subtle variations" is usually subjective audiophile hooey. That's code for "From here on out in my review, I'm going to make everything up."

 

[cjl]

   
When I look at that chart, I see almost identical response out almost to 10kHz +/-3dB or so. Then a considerable variation beyond that. I apply that to what I know about human hearing... i.e.: the most important octaves for balanced response are the ones between 40Hz and 10kHz, 3dB is the just noticeable difference for frequency response variation in music, the last octave from 10kHz-20kHz is the least used octave in music... and it tells me that barring any significant distortion or problems with fit and comfort, these headphones would all sound pretty similar. Odds are the variation due to manufacturing tolerances between two headphones of the same make and model might be as different as the differences between these different makes and models.

I don't see those as being as similar as you do - the LCD-3 is 10-15dB quieter than the T1 at ~8kHz, and that should be pretty audible. I do agree that the T1 and HE500 would sound pretty darn similar based on that graph, but there's a bigger difference between those two and the other two just below 10kHz, and that's a sufficiently low frequency that I would expect it to be a clearly audible difference.

 

[castleofargh]

about the best headphone objectively, if we're talking about listening loud, then I would go for ortho/electro without a doubt. but if we're talking about not too loud listening(what I personally look for), then it gets really complicated. up to a point where you want to listen quietly for hours and it's a nightmare to know what to buy as you have to look for a more V shaped signature, the lowest distortion at low volume levels, and at some point you need isolation. something that tends to ruin the quality of a driver.
I'm not sure we can get one headphone that does behave the best from 60 to 100db.
but if you guys have suggestion for a some top contenders at low volume, I'm mighty interested.
 
 
Quote:

 

 
I was able to use frequency response graphs from a single site that included all of the headphones that I had either auditioned or owned to get a rough idea of the characteristics that I preferred, and those that I did not favor as much.  Then, I read through countless posts and user reviews/experiences to get additional information and feedback on the headphones I was researching.  I had to find people that had similar tastes as mine, so that I could trust their impressions over another with completely different preferences in sound signatures.  The last few headphone purchases that I have made, even without auditioning them beforehand, provided the sound signature that I was fully expecting to hear based on this research.
 
Frequency response graphs can be extremely helpful, if they are consistent with regards to how they are generated.  

I do agree that frequency response graphs are extremely helpful. but really more helpful when you are comparing headphones with drastically different sound signatures or to get a general sense of the sound signature.
 
however, like this example: 

beyond the general sense of the sound signature, I feel like micro-analyzing the slight curve differences is counter-productive. The thing is that these graphs are all smoothed out, so there is probably even more detailed rough spots that you can see if you zoom. I feel like simply doing a blind ABx listening comparison test would be much more productive than trying to extrapolate how a difference in curve at spot X over spot Y affects the sound. I feel like sometimes people do get biased by looking at frequency response curves first and then attribute perceived XYZ sonic changes due to X spot on a graph. Expectation bias can happen from both subjective data and objective data!! That's why I personally prefer blind listening comparisons as the best standard for 'objective' judging. I feel like if it is a well-controlled comparison, it gives that individual a very good and realistic sense of how the headphones perform without any biases.

reading a graph is just like anything else, if the guy doesn't know how to read a graph and doesn't have enough references to actual sound(having a few of the measured headphones, and having actual experience in identifying one frequency), it will be useless. you can't blame measurements for not being objective simply because most audiophiles are ignorant and too lazy to read a scale or the name of the compensation curve applied. or because some think they're talking about trebles when they hear 2khz and talk about how low a headphone goes because it has a massive 125hz boost. blame the user not the tool.
the main problem with audiophiles isn't subjectivity vs objectivity(some don't even understand what that means). it's that they are mostly ignorant about everything related to audio. like you say, some will fight over 1db more at 8khz, but how many know that one T1 vs another T1 might very well have more than 1DB variation? how many know that placing them one way or another will change the signature again? or old pads vs new ones...
why someone claim graphs are useless, why someone claims that high res sounds better, why someone keeps looking for NOS DACs at the end of 2014, why people buy 1000$ cables for an IEM...
it all comes down to how ignorant and how mislead people can be. when we know what we are reading, namely one graph of one headphone of a given series, we know what we can make of it and what should not be taken at face value.
and because even reading a sentence gets distorted and someone out there will feel offended for no reason, I feel like I have to explain that being ignorant isn't the same as being an idiot. a bird crashing on a window isn't an idiot, it just doesn't have all the data.
 
I sure like to audition the headphones, but meets and festivals aren't that great as a tool too me. usually there is too much noise, not enough time and we usually can't afford to go back and forth between 2 or 3 headphones. so to me, meetings are where I can look for huge differences, and measurements are where I can fine tune ^_^. kind of the opposite of what you suggest. ofc owning the headphones is another story, but I don't have tens of them :'(.

 

[bigshot]

  I don't see those as being as similar as you do - the LCD-3 is 10-15dB quieter than the T1 at ~8kHz, and that should be pretty audible. I do agree that the T1 and HE500 would sound pretty darn similar based on that graph, but there's a bigger difference between those two and the other two just below 10kHz, and that's a sufficiently low frequency that I would expect it to be a clearly audible difference.

Try recreating it with EQ. See what you get. I bet it won't amount to a whole heck of a lot. By the time it gets up to the 8kHz and above range, the only thing that would be affected would be cymbals. You might be able to hear a slight difference in cymbal crashes if masking didn't cover it up anyway, but it wouldn't matter a whole lot to the overall response, because that is above the point where human ears perceive sound as a musical note and it is only occurring in brief cymbal hits, not continuous tones. 8kHz and up is pretty doggone high. People tend to overestimate the importance of that top octave to the response of transducers. If you can get fairly flat out to the top octave, you have accomplished 95% of your job.
 
10-15dB at 3kHZ OH YEAH. Big difference. 10-15dB at 200Hz YES INDEED. But above 8kHz, you're starting to get into the ozone.

 

[money4me247]

@castleofargh, I'm definitely not saying graphs are useless, but yea, I do kinda feel like the fact that there are so many other variables (manufacturer variation in model, positioning, seal, pads, microphone placement, types of filtering used, type of weighting used, resolution of the graph you see vs the raw) that can influence the measurements that they are better served as guidelines rather than like a blueprint of exactly how the headphones are going to sound.
 
@liamstrain, the reason I stated earlier that there isn't any objective standard in the hobby (though slight exaggeration) is kind of being demonstrated here with how different people can have different interpretations of the objective data point: the frequency response curves. There are technically objective data points, but people have so much varying personal interpretations, I don't think there is really any hard universal objective standard. I would say that frequency response curves is probably the closest to an objective measuring stick as we have, but that does not seem to account for every aspect of the headphone's sound. There is also some objective value of comparing to a reference point (comparing one pair of headphones to another established sound of headphones in relative terms -> aka the HD600 has less bass than the M100s) as those claims can be duplicated and confirmed, but then you have lots of opinions that don't match, which is weird. Then factor in people's listening preferences, musical preferences, sound signature preferences... everything becomes really subjective really quickly. I think it is important to note that not everyone is looking for neutral headphones which makes things even more dependent on subjective "what do i like the best" type impressions.

 

[bigshot]

It's easy to find out for sure what the difference is. EQ the two settings and compare them for yourself. 10dB above 8kHz isn't the same as 10dB anywhere lower than that.

 

[money4me247]

  It's easy to find out for sure what the difference is. EQ the two settings and compare them for yourself. 10dB above 8kHz isn't the same as 10dB anywhere lower than that.

thank you for the information. i will definitely try that in the future to see what's up.

 

[appsmarsterx]

  There is absolutely no objective standard test for anything in this hobby. Everything is subjective... even stuff that sounds objective is usually either taken out of context or overstating the importance of certain measurements that do not have that big of an effect in real life. You will learn soon enough that most people here form their opinions first, and then cherry-pick information to support their view. If you have come for objective-talk, this website is probably not the best place for you.
 
There is the frequency response curve serves as a rough guideline on how to expect headphones to sound overall. There is harmonic distortion curves, square wave response, and cumulative spectral decay (waterfall) plots/impulse response. I am not sure how valuable any of these objective measurements are. I am actually in the pro-measurements/objectivity camp, but it doesn't seem like any of those measurements really means much in real life applications except the frequency response curve. Everything else seems kinda just bragging rights that it measured well. 
 
However, I think there is actually a lot of value in an ear test. If you have ears (no need for special ears), and just do a double blind ABx comparison between two headphones that you have done no research about in terms of price/reviewed performance... I think whatever the difference you hear would probably be pretty accurate (as after you remove expectation bias, our ears are quite good at picking random differences).
 
In fact, I honestly think that the lack of objective criteria is one of the reasons some people are drawn to this hobby. They like the fact that you can be a "golden ear/audiophile" and sound like a really legit expert without really saying/knowing anything beyond owning something expensive. The fact that there isn't a comprehensive sticky about common audiophile myths & what current scientific testing has found (cable difference, burn-in, decay, the real-world impact of amps/dacs) and the fact that head-fi is actually ad-sponsored... I really think that the main purpose of this website is really just to promote purchasing 'audiophile' gear and for people who enjoy the gear to socialize. Questioning sketchy manufacturer practices like charging an additional >$100+ for a small tweak for essentially the same product doesn't really get anywhere around here. Questioning priceerformance ratio usually just piss off owners of expensive stuff. Anytime you try to get into any objective type debate on the general forums, any random person's subjective impressions is always "more legit" due to the head-fi mentality that statements about headphones are only valid after hearing them. 
 
Just what I've found from my experiences here. Didn't mean to sound negative about this place, but I think you have to see this website for what it is: bunch of headphone geeks that are using it as a platform to express their opinions and have people listen to them as audiophile experts.

 
yes I think you are right, this enthusiast niche might not for me. I'm not super rich or super crazy to spend 1000s of $$$ for something like super expensive cables, and these so called super high end amps, that doesn't makes much sense.  I'm thinking about quitting this place for a while but for some reason I'm keep coming back. I'm currently happy with my mid-fi setup, probably will stick with it forever.. lol
 
I think once Tyll from innerfidelity did a burnin experiment with a Q701 but the results were

inconclusive. there is a good article on ​

goldenears, addressing the cables and the sound quality myth. it shows how capacitance/resistance in cables can only slightly changes the high and low frequencies, so some people might think buying high end cables can magically increase sound quality.
http://en.goldenears.net/1301
 
lack of objective criteria might be the reason why celebrity endorsed brands dominating the market share in this niche. probably they ll keep dominating forever..

 

[limpidglitch]

 
Try recreating it with EQ. See what you get. I bet it won't amount to a whole heck of a lot. By the time it gets up to the 8kHz and above range, the only thing that would be affected would be cymbals. You might be able to hear a slight difference in cymbal crashes if masking didn't cover it up anyway, but it wouldn't matter a whole lot to the overall response, because that is above the point where human ears perceive sound as a musical note and it is only occurring in brief cymbal hits, not continuous tones. 8kHz and up is pretty doggone high. People tend to overestimate the importance of that top octave to the response of transducers. If you can get fairly flat out to the top octave, you have accomplished 95% of your job.
 
10-15dB at 3kHZ OH YEAH. Big difference. 10-15dB at 200Hz YES INDEED. But above 8kHz, you're starting to get into the ozone.

 
Practical question here.
When doing such a comparison, how should volume be normalized, if at all?

 

[SilverEars]

   
yes I think you are right, this enthusiast niche might not for me. I'm not super rich or super crazy to spend 1000s of $$$ for something like super expensive cables, and these so called super high end amps, that doesn't makes much sense.  I'm thinking about quitting this place for a while but for some reason I'm keep coming back. I'm currently happy with my mid-fi setup, probably will stick with it forever.. lol
 
I think once Tyll from innerfidelity did a burnin experiment with a Q701 but the results were

inconclusive. there is a good article on ​

goldenears, addressing the cables and the sound quality myth. it shows how capacitance/resistance in cables can only slightly changes the high and low frequencies, so some people might think buying high end cables can magically increase sound quality.
http://en.goldenears.net/1301
 
lack of objective criteria might be the reason why celebrity endorsed brands dominating the market share in this niche. probably they ll keep dominating forever..

I've seen FR sweep of cables.  Pretty much a flat line, purely resistive.  Capacitance and inducatance is negligible. A metal wire doesn't have capacitive or inductive characteristics at audio frequencies therefore it's flat, a resistor.  I haven't seen a sweep of cable other than a flat line.  This means it acts as a resistive element and only resistance matters to the signal.  Only cable that I have seen have significant(3.9ohm could be significant for very low impedance phones) resistance is Estron cables(because of high resisance, and why would you want high resistance?):
 
http://cymbacavum.com/2014/09/24/estron-linum-balanced-2-5-trrs-think-thin-then-think-thinner/
 
Manufacturer impedance specifications are as follows:

Vocal: 3.9Ω
Music: 1.9Ω
BaX: 1.4Ω

For optimal technical performance with very sensitive multi-BA applications, the BaX model with lower impedance is recommended. Linum has posted some measurements of all three, using the same source gear and IEM, as performed by Cosmic Ears. Keep in mind that these measurements might not be applicable to all earphones.

 

 

[Steve Eddy]

Suffice it to say, I disagree entirely. 

I don't know, Liam. I think the only thing that has any real meaning to the listener is the subjective experience. You can't really experience a set of numbers or measurements. You can only experience the gestalt. A favorite quote of mine:

The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed—Robert Pirsig

se

 

[Rajikaru]

I don't know, Liam. I think the only thing that has any real meaning to the listener is the subjective experience. You can't really experience a set of numbers or measurements. You can only experience the gestalt. A favorite quote of mine:

The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed—Robert Pirsig

se

I believe in Gestaltism as well, when it comes to cars and motorcycles, the experience of which you can touch, hear, see, smell, and probably even taste (on a motorcycle). It is inherently easier to determine what's snake oil and what isn't with motor vehicles. Easier to measure (i.e. 0-60, lap times etc.), along with numerous safety and measuring standards etc. that vehicles have to go through before entering the market. We get to enjoy the gestalt mental dialogue when driving for the pleasure of it afterwards. 
 
I cannot say it's the same case for audio though, which is why there should at least be a set of reasonable assumptions backed by solid science and measurements to operate on. A separator between what we know for sure (info backed by solid science, measurements, and listening tests) vs. unproven claims. 

 

[esldude]

 
I believe in Gestaltism as well, when it comes to cars and motorcycles, the experience of which you can touch, hear, see, smell, and probably even taste (on a motorcycle). It is inherently easier to determine what's snake oil and what isn't with motor vehicles. Easier to measure (i.e. 0-60, lap times etc.), along with numerous safety and measuring standards etc. that vehicles have to go through before entering the market. We get to enjoy the gestalt mental dialogue when driving for the pleasure of it afterwards. 
 
I cannot say it's the same case for audio though, which is why there should at least be a set of reasonable assumptions backed by solid science and measurements to operate on. A separator between what we know for sure (info backed by solid science, measurements, and listening tests) vs. unproven claims. 

The point about the Pirsig quote is in the last sentence.  Either the machine or your mind is changed.  Even if the machines in audio are audibly equivalent a host of other factors alter the mind.  Altering the mind is what is needed.  But until one reaches that point in trusted the audible equivalence one likely can't change their mind.

 

[Rajikaru]

 
The point about the Pirsig quote is in the last sentence.  Either the machine or your mind is changed.  Even if the machines in audio are audibly equivalent a host of other factors alter the mind.  Altering the mind is what is needed.  But until one reaches that point in trusted the audible equivalence one likely can't change their mind.

 

 
When we talk about 'altering/changing the mind', the question is - Who is doing it?
 
The listener? The marketer? Head-fi group think? The Sith? 
 
I'm highly resistant to Jedi mind tricks (or at least I hope I am). And if anyone is altering my mind, I would hope that person would be me.
 
Seriously though, I know what you mean. There is a high level of subjectivity in this hobby. Objective measurements are important up to a certain point, after which subjective preferences come into play. 
 
BTW, I've read your articles on nulling differences between cables, and wonder why no one does the same for Amplifiers and DACs.