[levap]

In the course of work on a thread about talents of various headhones to play metal (http://www.head-fi.org/t/715478/headphones-for-metal-music-ultimate-solution) I consider "Detail level" and "Speed" among key criterias for cans / metal subgenre gestalt response. I tried to combine formal definitions with subjective feeling of cans detail and speed performance and to write down results in a spreadsheet below. Results are relevant for headphones level (top level headphones low speed isn't equal to basic level low speed).
 
Any ideas? Does such a classification make sense? It is useful? Intuitive? Or does it mislead? Or may be someone already made something like this a long time ago?
 

 
 
Head-fi
Detailed - Easy to hear tiny details in the music; articulate. Adequate high frequency response, sharp transient response.
Fast - Good reproduction of rapid transients which increase the sense of realism and "snap".
Speed - A fast system with good pace gives the impression of being right on the money in its timing.
 
Stereophile
detail The subtlest, most delicate parts of the original sound, which are usually the first things lost by imperfect components. See "low-level detail." Compare "haze," "smearing," "veiling."
fast Giving an impression of extremely rapid reaction time, which allows a reproducing system to "keep up with" the signal fed to it. (A "fast woofer" would seem to be an oxymoron, but this usage refers to a woofer tuning that does not boom, make the music sound "slow," obscure musical phrasing, or lead to "one-note bass.") Similar to "taut," but referring to the entire audio-frequency range instead of just the bass.
speed The apparent rapidity with which a reproducing system responds to steep wavefronts and overall musical pace. See "fast," "slow." 

 

[vid]

The main issue is, neither speed nor detail are inherent in headphones - huge variation in individual HRTF guarantees a different sound for different people. So we're talking human phenomena rather than headphones, something that's also reflected in the subjective definitions you gave.
 
It comes down to frequency response if we go by scientific understanding, so this classification might be more at home replacing in the list headphones with response curves - or incorporating this info additionally. Especially given that you don't need a particular headphone to have its frequency response.
 
(I could add that I did an investigation of a similar thing there: http://www.head-fi.org/t/697840/investigating-frequency-response-vs-subjective-report. The graph is down but make a scatterplot of the last two items in the table.)

 

[darkainu]

This list might give an idea for some people, so I can't really say it's useless but also shouldn't be considered as a buying guide. I am also suprised by Mad Dog 3.2's speed.

 

[metal571]

I'd actually put the Q701 as a high speed can, not an average speed one. Even Tyll thought they were very fast.

 

[TMRaven]

I never really liked the notion of 'speed.'  99/100 it's used as an indication of overall frequency balance.  Bright vs dark etc.
 
 
Take the Audeze planars and the Hifiman HE-400 here.  By nature of their driver technology, all planar magnetics are faster at starting and stopping than a dynamic cone driver, just because the mass of the diaphragm is really light, and the excursion is really small, yet their subjective rating for 'speed' here is poor.  Take the LCD3 fazor vs non fazor'd.  The fazors should have nothing to do with the 'speed' of the headphone.  They're their as waveguides, to help imaging only-- yet we see a difference here in 'speed' because of the fazor implementation.
 
When it comes down to it, I never was able to attribute real life performance to 'speed,' I think it's a moot idea.
 
 
Detail retrieval is also something that can be highly influenced by frequency response-- specifically bright headphones with lots of treble.  Nevertheless, it's still a subjective quality that's a little more applicable than 'speed.'

 

[vid]

  I never really liked the notion of 'speed.'  99/100 it's used as an indication of overall frequency balance.  Bright vs dark etc.

 
When someone says headphones x are bright, they mean to say they perceive them to be bright - which isn't to say the headphones actually possess this quality of being bright. And whether they're bright or not and by which outside standard isn't important for the listener; what matters it how the sound is perceived by them.
 
In this sense, if you take a waterfall plot of a flat frequency response, you see speed. If you take a waterfall of a wonky response, you see less speed. As such, I don't see a problem with equating speed with the balance of the frequency response as it's perceived - if it's stated that when the balance is even there's more speed and less speed when otherwise.

 

[TMRaven]

It really does make no sense.  Grados are very sloppy sounding headphones-- the'yre full of ringing issues and uncontrolled bass, yet they're given a high for 'fast' just because of their treble and upper mid emphasis.  A cable changing the speed of the 650 from low to average-- yet detail remains the same?  Since when did a cable make a headphone driver have faster articulation or faster decay?  An HE-400 having slower 'speed' than an M50?
 
There's a lot of confusing things when it comes to this whole 'speed' notion.  What I'm trying to get at is that there are subjective qualities that govern the term, but the term itself is nonsensical. 

 

[levap]

  The main issue is, neither speed nor detail are inherent in headphones - huge variation in individual HRTF guarantees a different sound for different people. So we're talking human phenomena rather than headphones, something that's also reflected in the subjective definitions you gave.
 

So your statement is - there is no slow or fast, no detailed or muddy headphones, no bright or dark, bassy or bass-light - it's your personal phenomena? The same could be (and for some extent is) right for our eyes colours perception. So I find your statement too strong, There are inherent headphones characteristics and it's approved by independent observations. Fortunately ))
 
 

  This list might give an idea for some people, so I can't really say it's useless but also shouldn't be considered as a buying guide. I am also suprised by Mad Dog 3.2's speed.

Othodynamics are tough target for "speed" description. In discussion on doctorhead.ru friend of mine compared LCD-2 to bear - meaty, big, fat but able to move extremely fast when it's needed, so I find othodynamics "slowness" as metaphoric therm to define their character. If you try LCD (or Mad Dogs), you obviously understand me 
 
 

I'd actually put the Q701 as a high speed can, not an average speed one. Even Tyll thought they were very fast.

Okay, my mistake...
 
 

  I never really liked the notion of 'speed.'  99/100 it's used as an indication of overall frequency balance.  Bright vs dark etc.
 
 
Detail retrieval is also something that can be highly influenced by frequency response-- specifically bright headphones with lots of treble.  Nevertheless, it's still a subjective quality that's a little more applicable than 'speed.'

You're oversimplifying here, please read definitions I quoted above.
 
 

   
There's a lot of confusing things when it comes to this whole 'speed' notion.  What I'm trying to get at is that there are subjective qualities that govern the term, but the term itself is nonsensical. 

Stereophile
speed The apparent rapidity with which a reproducing system responds to steep wavefronts and overall musical pace. See "fast," "slow."
slow Sound reproduction which gives the impression that the system is lagging behind the electrical signals being fed to it. See "fast," "speed," "tracking."
fast Giving an impression of extremely rapid reaction time, which allows a reproducing system to "keep up with" the signal fed to it. (A "fast woofer" would seem to be an oxymoron, but this usage refers to a woofer tuning that does not boom, make the music sound "slow," obscure musical phrasing, or lead to "one-note bass.") Similar to "taut," but referring to the entire audio-frequency range instead of just the bass.

 

[Claritas]

I think Raven's analysis is in the right direction, even if I partly disagree. The subjective impression of both speed and detail does seem to correlate with a brighter-sounding FR (whether because of treble emphasis or lack of bass), which might explain why some Grados sound fast.

 

[vid]

  So your statement is - there is no slow or fast, no detailed or muddy headphones, no bright or dark, bassy or bass-light - it's your personal phenomena? The same could be (and for some extent is) right for our eyes colours perception. So I find your statement too strong, There are inherent headphones characteristics and it's approved by independent observations. Fortunately ))

 
What I was getting at is that there's a pair of headphones, and there's a frequency response. And there's another frequency response, but the same headphones. And a third response, etc. This is what you hear, one of those responses. Which it is depends on what kind of physical ears you have; with some ears, you have +20 dB in the treble, and with others, -10 dB in the mids, and so on. It's not inherent in the headphones to sound bright or dark or neutral or slow, until you make a headphone with eg. -40 dB in the mids, which you won't find an ear for that finds it neutral, but this rarely happens.
 
When you sample for this kind of thing, I'd say you want to keep HRTF in mind. I suggested in the thread I linked to that by identifying the spectral range responsible for the phenomenon you're looking at the term for, altering the frequency response in that range, and counting the percentage of listeners that find each version of the response to fulfill the criteria of the term, you find the distribution of HRTF variation at the spectral range responsible for what the term refers to. You'll also find the distribution fairly broad, as mentioned above, so if you sample the reaction of one person to a given headphone, you've not sampled the headphone but one potential reaction on a continuum (cf. TMRaven's suspicions about the effects of a cable change).

 

[levap]

   
What I was getting at is that there's a pair of headphones, and there's a frequency response. And there's another frequency response, but the same headphones. And a third response, etc. This is what you hear, one of those responses. Which it is depends on what kind of physical ears you have; with some ears, you have +20 dB in the treble, and with others, -10 dB in the mids, and so on. It's not inherent in the headphones to sound bright or dark or neutral or slow, until you make a headphone with eg. -40 dB in the mids, which you won't find an ear for that finds it neutral, but this rarely happens.
 
When you sample for this kind of thing, I'd say you want to keep HRTF in mind. I suggested in the thread I linked to that by identifying the spectral range responsible for the phenomenon you're looking at the term for, altering the frequency response in that range, and counting the percentage of listeners that find each version of the response to fulfill the criteria of the term, you find the distribution of HRTF variation at the spectral range responsible for what the term refers to. You'll also find the distribution fairly broad, as mentioned above, so if you sample the reaction of one person to a given headphone, you've not sampled the headphone but one potential reaction on a continuum (cf. TMRaven's suspicions about the effects of a cable change).

I do understand what you say and absolutely agree with it. I would add, that "a frequency response" and "another frequency response" can take place in one listeners head i.e. in the morning and in the evening on in different mind and body conditions. So I take 2 weeks in general for headpnones review for my metal thread. At least I can say, that FOR ME these headphones sound THIS way. And I do understand, that for another person it could sound somewhat different. And it deeply complicates communication process, but makes it so fun )
 
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. 
 
But, fortunately, (again at least in my observations) there are lots of terms like "bright", "dark", "warm", that make sense for me when you read headphones review and you can predict your own reaction well.
 
Seems like "speed" is another beast and at least in straightforward way as I done in spreadsheet above it confuses more, than provides reliable information. And to approve (or reject) this statement I created this thread )
 

PS 650 cable chahge had very clear and repeatable effect - removal of "veil" plus bass and treble rise. Removal of "veil" made 650 "faster" and highs rise actually did not added new details, resolution was the same. That's all subjective anyway )

 

[Asr]

The table is missing the one headphone I consider the "speed" king of headphones, the Audio-Technica AD2000. Anyone who listens to thrash metal and hasn't heard the AD2000 owes to himself (or herself) to check it out.

 

[levap]

  The table is missing the one headphone I consider the "speed" king of headphones, the Audio-Technica AD2000. Anyone who listens to thrash metal and hasn't heard the AD2000 owes to himself (or herself) to check it out.

My personal current "King of speed" - Abyss AB-1266 - stunnigly powerful and agile. They ride like lightning, start and stop instantly.

 

[Mad Max]

  It really does make no sense.  Grados are very sloppy sounding headphones-- the'yre full of ringing issues and uncontrolled bass, yet they're given a high for 'fast' just because of their treble and upper mid emphasis.  A cable changing the speed of the 650 from low to average-- yet detail remains the same?  Since when did a cable make a headphone driver have faster articulation or faster decay?  An HE-400 having slower 'speed' than an M50?
 
There's a lot of confusing things when it comes to this whole 'speed' notion.  What I'm trying to get at is that there are subjective qualities that govern the term, but the term itself is nonsensical. 

 
The "fast" quality usually described about Grados has to do with the attack speed of reproduced sounds.  They sound faster than most others, you can EQ down the treble and upper mids emphasis and that won't change its "speed".  Try it.
But I'll agree that it is a nonsense term.
Additionally, I'm a "believer" of cables and I sure as heck don't believe in "speed" being affected by cables in the least.
 
Also, "low" speed for HE-400 is absolutely ludicrous.  They are not "slow", regardless of whether you pay attention to the attack/release of sound or FR balance.
 
 
 
 

  ... PS 650 cable chahge had very clear and repeatable effect - removal of "veil" plus bass and treble rise. Removal of "veil" made 650 "faster" and highs rise actually did not added new details, resolution was the same. That's all subjective anyway )

 
"Increased" bass and treble you mean?

 

[levap]

   
"Increased" bass and treble you mean?

Exactly.