[Mad Max]

  Exactly.

 
It is actually increased clarity, that's why it seemed like more when there was in fact no audible change in the FR.  I know what you mean, HD650 needs the recable or to have the stock cable modified to get rid of the bloat and improve control and clarity a bit.  Some guys EQ up the low bass because of the poor clarity with the stock cable, then it is no longer necessary after a recable.  A lack of clarity can fool us into there being a drop in volume of whatever frequencies when there is actually no drop.  You can tell easily what the case is with a sine sweep.
On a related note, too much resonance/reverb can mask parts of the FR, resulting in the same "illusion", in my experience.
Dynamat Xtreme on the back of the driver chamber also helps reduce the reverb in it a bit, as well as tighten up the sound a little bit.

 

[vid]

  Possibly it could be great if frequency response diagram + understanding of HRTF told you all about headphones sound signature, but it's just does not work this way. Until you personally try headphones (at least I say for me) you can't say definitely they're "dark" or "bright" etc for you.

 
Leaving out considerations of frequency response makes the exchange of reviews almost random - as in, share a review but don't say which headphone it's for. Having said that, you certainly will know - to a fair extent - how a pair of headphones would sound to you if you know their response at your eardrum as well as your own HRTF.
 
What I guess is happening here is that it's more interesting to have the mystery and the random arguments about whether a given headphone is bright or not - but if the aim is to make a classification to help clear things out, I don't see the need nor logic to exclude frequency response.
 
Also, I'd again point to the thread I originally linked to - the one where I find what's effectively a 100% correlation between energy in the midrange and Head-Fi reviewers' impressions of the accuracy of the headphones. Though the result guarantees nothing, it for sure doesn't speak for the exclusion of the frequency response.

 

[levap]

   
Leaving out considerations of frequency response makes the exchange of reviews almost random - as in, share a review but don't say which headphone it's for. Having said that, you certainly will know - to a fair extent - how a pair of headphones would sound to you if you know their response at your eardrum as well as your own HRTF.
 
What I guess is happening here is that it's more interesting to have the mystery and the random arguments about whether a given headphone is bright or not - but if the aim is to make a classification to help clear things out, I don't see the need nor logic to exclude frequency response.
 
Also, I'd again point to the thread I originally linked to - the one where I find what's effectively a 100% correlation between energy in the midrange and Head-Fi reviewers' impressions of the accuracy of the headphones. Though the result guarantees nothing, it for sure doesn't speak for the exclusion of the frequency response.

I don't leave out frequency response out of consideration, for classification purposes and fast reference I use simplified table below.
And do you know practical way to translate frequency response + personal HRTF into real world headphones experience? 
 

 

[vid]

 
And do you know practical way to translate frequency response + personal HRTF into real world headphones experience? 

 
Practical for whom or in what way?

 

[levap]

   
Practical for whom or in what way?

For everyone in every way )) My question was - how someone can use your advice to "translate frequency response + personal HRT" in practice, supposing it's the answer for ALL the questions? 

 

[vid]

The way you say it makes it seem headphones, frequency responses, and HRTF are separate things.
 
Put it this way: there's no correspondence between name and sound - so you don't find out about the sound by analyzing the name.

 

[levap]

  The way you say it makes it seem headphones, frequency responses, and HRTF are separate things.
 
Put it this way: there's no correspondence between name and sound - so you don't find out about the sound by analyzing the name.

Couple of posts above you said "Having said that, you certainly will know - to a fair extent - how a pair of headphones would sound to you if you know their response at your eardrum as well as your own HRTF." So my question is - in practical way how to understand "how a pair of headphones would sound to you"? Do you know practical way to fill the gap from quantitative (frequency responce etc) to qualitative evaluation ("too bright", "too dark", "small soundstage" etc)?
May be it would be more easy for you to give some real word examples?

 

[vid]

What can I say - see the thread I've linked you to about four times. It gives you one of the methods to translate data to feelings. But I don't think the problem is about lacking a method - more to do with not wanting to accept that frequency response would have anything to do with subjective musings. As it is, though, there's more proof that it does than not, and a classification aiming to be useful, as I've said, certainly wouldn't pass that by.

 

[levap]

I see only one link http://www.head-fi.org/t/697840/investigating-frequency-response-vs-subjective-report
And there is no answer on my questions above. Thread is about correlation of "% of reviews with keyword "accurate" with "2+3+4 kHz, total dB vs 1 kHz level" for 5 headphones. Conclusion is, that some objective measurable value correlates with subjective reviews. 
 
My question was about practical manual for translation of frequency responce + HRTF into qualitative evaluation of headphones. And I ask again. In practical way how to understand "how a pair of headphones would sound to you"? Do you know practical way to fill the gap from quantitative (frequency responce etc) to qualitative evaluation ("too bright", "too dark", "small soundstage" etc)?
 
You say, that "Leaving out considerations of frequency response makes the exchange of reviews almost random", but you don't give any practical advice how to use it.

 

[vid]

Not sure what the problem is. As the thread you've come to notice says, less energy in the midrage = higher chance of people rating the headphones as "accurate". You can extrapolate from that and say, "less energy in the xxxxx = higher chance of people rating the headphones as yyyyy". Of course, you'll first put in the work to figure out what x and y are, and then it becomes a question of how to represent that to people in an understandable way, who can then relate their previous experiences to the data you provide them to draw conclusions. To that you'll of course repeat your question with no action, but I'd say there's no way at present for a person reading the classification to know what causes this or that headphone to rank above the other, they just see the effect, and the cause underlying the effect may or may not be applicable to them or may or may not have been misrepresented in the classification, which the reader can't appreciate from this as a list of effects.

 

[levap]

  Not sure what the problem is. As the thread you've come to notice says, less energy in the midrage = higher chance of people rating the headphones as "accurate". You can extrapolate from that and say, "less energy in the xxxxx = higher chance of people rating the headphones as yyyyy". Of course, you'll first put in the work to figure out what x and y are, and then it becomes a question of how to represent that to people in an understandable way, who can then relate their previous experiences to the data you provide them to draw conclusions. To that you'll of course repeat your question with no action, but I'd say there's no way at present for a person reading the classification to know what causes this or that headphone to rank above the other, they just see the effect, and the cause underlying the effect may or may not be applicable to them or may or may not have been misrepresented in the classification, which the reader can't appreciate from this as a list of effects.

I am sorry. You pretend to be scientific, but you're extremely anti-scientific actually. You are joking or you are troling.
 1. You have observation on object with no definition - "accurate". I do not know, what those people mean when they describe something "accurate", . (Meanwhile you have definitions for "speed" and "detailed" terms from me)
 2. You say "less energy in the midrage = higher chance of people rating the headphones as "accurate"". Do you really think, that it's enough to call headphones "accurate"? Recessed midrange = "accurate" headphones???
 3. You say "You can extrapolate from that and say, "less energy in the xxxxx = higher chance of people rating the headphones as yyyyy"". If it's really so simple, please give me a link to xxxxx - yyyyy relations.
 
Anyway. I record my own opinion on headphones in systematic way. For myself. If it's possible to make my impressions (more universal and so more useful for public, I can adjust my marks. But I need some practical advice.

 

[vid]

1. I see. The aim of the paper was to evaluate the applicability of a corpus-based methodology to the study of headphone psychoacoustics - whether there's need to arrange for in situ subject participation as often done or whether pre-existing data not originally meant for this kind of work would suffice. Whether frequency response correlates with subjective impressions wasn't the research question - there's enough evidence to suggest it does to a large extent - but rather, whether similar results would surface from applying the corpus methodology as do with arranged listening experiments. Turns out similar results came up, in the sense that subjective impressions were strongly correlated with aspects of the frequency response. As I say, the results guarantee nothing, but they're certainly more convincing than the plain opinion that "it just isn't so". As to what accurate means - was part of the design of the study not to define it for the respondents, yet it seems possible they agreed on it anyway. Not surprising, given that people don't use language at random; identify the community where the term is used and find out what it might mean in that community, starting from HF's 'official' glossary of terms.
 
2. A reduction in response means recessed in regards to the baseline only, there being but one baseline and that's the measurement ear on which the response data were produced. Recessed for that ear, or for any human ear, will mean more neutral to another ear - which I touched on when I suggested the shape of the correlation to reflect the distributions of HRTF variation in the sample.
 
3. Like I said, you'll need to do the work. Start by looking up previous research, then maybe design a study of your own.

 

[levap]

  1. I see. The aim of the paper was to evaluate the applicability of a corpus-based methodology to the study of headphone psychoacoustics - whether there's need to arrange for in situ subject participation as often done or whether pre-existing data not originally meant for this kind of work would suffice. Whether frequency response correlates with subjective impressions wasn't the research question - there's enough evidence to suggest it does to a large extent - but rather, whether similar results would surface from applying the corpus methodology as do with arranged listening experiments. Turns out similar results came up, in the sense that subjective impressions were strongly correlated with aspects of the frequency response. As I say, the results guarantee nothing, but they're certainly more convincing than the plain opinion that "it just isn't so". As to what accurate means - was part of the design of the study not to define it for the respondents, yet it seems possible they agreed on it anyway. Not surprising, given that people don't use language at random; identify the community where the term is used and find out what it might mean in that community, starting from HF's 'official' glossary of terms.
 
2. A reduction in response means recessed in regards to the baseline only, there being but one baseline and that's the measurement ear on which the response data were produced. Recessed for that ear, or for any human ear, will mean more neutral to another ear - which I touched on when I suggested the shape of the correlation to reflect the distributions of HRTF variation in the sample.
 
3. Like I said, you'll need to do the work. Start by looking up previous research, then maybe design a study of your own.

1. "There is no scientific relaton of term "accurate" (in absense of term) and our findings"
2. "There is no exact value to measure "accurate"
3. "Please, go research further our vague findings"
 
That's really sad for hi-fi industry if we have such answers for simple looking questions in 2014.

 

[Head Injury]

  1. "There is no scientific relaton of term "accurate" (in absense of term) and our findings"
2. "There is no exact value to measure "accurate"

I don't think this is what he's saying.
 
Accurary or "neutrality" is easily definable as reproducing what's on the recording exactly with no deviation in frequency response or distortion; if the system adds nothing to the signal, it's accurate to the signal.
 
For headphones this is made more complicated by our ears, which alter and distort the sound produced based on their shape. We can say that a baseline neutral system is a measurably neutral speaker system in a neutral room some given distance in front of us. The sound produced by this system will be accurate to the source file, the sound will travel through the air and interact with our outer ear and ear canal the same way a natural sound source (like the original band) will. It will be, in theory, as if the band were in front of you playing the recording themselves, ignoring all the dirty mastering tricks.
 
Headphones skip the way the sound travels through the air, and they skip most of the outer ear to pump sound more or less right into our ear canal. A neutral headphone, then, will have a frequency response that exactly compensates for the things it removes. If you've seen any headphone measurements, the most obvious thing this results in is a boost from about 1 kHz peaking at 3 kHz, then back down again. This is the part of the effect our outer ear has to enhance the mid-range frequencies responsible for communication.
 
But, because each of our ears are different, we're used to hearing the world in slightly different ways and our brains are wired to compensate in different ways. This is our HRTF. The only way we can know our own HRTF, our own personal neutrality, is to measure it. One way to do this is with sine sweeps, playing all frequencies and making note of where certain frequencies are too loud or quiet. Then compare that with graphs of headphone measurements, and use EQ to correct it or find a headphone that better matches it. No two people are going to have exactly the same "accurate" sound, even ignoring the subjective preferences that always muddy up the definition.
 
Anyway, on the thread topic itself, I won't comment on everything I think is wrong with the chart you made, but I will share the one part that made me dismiss the whole thing: A new cable turned the HD650 from Dark to Light

 

[levap]

  I don't think this is what he's saying.

He is not saying this but it sounds like this in a result (( And I'm sorry if my words sound like bully. What bothers me - we don't have objective, clear and straightforward way for sound gear classification in human-impressions terms.
 
 

  Accurary or "neutrality" is easily definable as reproducing what's on the recording exactly with no deviation in frequency response or distortion; if the system adds nothing to the signal, it's accurate to the signal.
 
For headphones this is made more complicated by our ears, which alter and distort the sound produced based on their shape. We can say that a baseline neutral system is a measurably neutral speaker system in a neutral room some given distance in front of us. The sound produced by this system will be accurate to the source file, the sound will travel through the air and interact with our outer ear and ear canal the same way a natural sound source (like the original band) will. It will be, in theory, as if the band were in front of you playing the recording themselves, ignoring all the dirty mastering tricks.
 
Headphones skip the way the sound travels through the air, and they skip most of the outer ear to pump sound more or less right into our ear canal. A neutral headphone, then, will have a frequency response that exactly compensates for the things it removes. If you've seen any headphone measurements, the most obvious thing this results in is a boost from about 1 kHz peaking at 3 kHz, then back down again. This is the part of the effect our outer ear has to enhance the mid-range frequencies responsible for communication.
 
But, because each of our ears are different, we're used to hearing the world in slightly different ways and our brains are wired to compensate in different ways. This is our HRTF. The only way we can know our own HRTF, our own personal neutrality, is to measure it. One way to do this is with sine sweeps, playing all frequencies and making note of where certain frequencies are too loud or quiet. Then compare that with graphs of headphone measurements, and use EQ to correct it or find a headphone that better matches it. No two people are going to have exactly the same "accurate" sound, even ignoring the subjective preferences that always muddy up the definition.
 
Anyway, on the thread topic itself, I won't comment on everything I think is wrong with the chart you made, but I will share the one part that made me dismiss the whole thing: A new cable turned the HD650 from Dark to Light 

 

 
Let's look on "accurate" and "neutral" matters from another side. My favorite cans are generally neutralish. But I think, that a little bit of imperfection or character (anyway we have to call it "coloration") - that makes me really like them. And I have some anti-neutralish favorites. And I think it's great, that here in a headphones world (opposed to speakers world) we have a luxury to have a wide pallete of sound signatures at hand, available to use in a real world home.
 
In this perspective, by the way, to call headphones "slow" is not insult by default. It could mean oppurtunity for impressive not fast genres performance.
 

 
Anyway, on the thread topic itself, I won't comment on everything I think is wrong with the chart you made, but I will share the one part that made me dismiss the whole thing: A new cable turned the HD650 from Dark to Light 

I'm not a big headphones cables fan (there are only 3 of 60 in my spreadsheet). But headphones cables (unlike interconnects, at least in my experience) can considerably change headphones sound signature. So, how would you call 650 with removed veil and U-shaped balance?