I know what you're going to say......."Headroom graphs are unreliable despite being objectively measured - my ears coupled with psychoacoustics and my expectation bias are much more reliable and trusworthy".  Well to those of you who are more objectively inclined, however few, I've compiled some interesting graphs comparing the DT880 / T1 / HD800 showing remarkable similarities in the FQ responses / square wave areas:
 

 
And to show you that there is actually considerable variation of graphs at headroom, I've included an FQ response example, from 3 other open high end headphones, albeit with questionable timbral accuracy and neutrality.
 

 
Your thoughts everyone?  My first reaction is "WOW..maybe I should be happy with my DT880", and it appears as though the smooth treble of the T1 is also apparent in the first graph.  I would expect that things like soundstage / imaging etc aren't represented well by these graphs, but overall tonal balance should be obvious.  Also, notice the rather linear response of the DT880 / T1 / HD800 (the K701, HD600 and HD650 also exhibit this linearity according to Headroom), in contrast to the coloured sound of the R1 / SA5K / W5000.  Hmmm!

Quote:

catharsis said:

......."Headroom graphs are unreliable despite being objectively measured - my ears coupled with psychoacoustics and my expectation bias are much more reliable and trusworthy". 

Click to expand...

Brilliant.
 
But besides that, these graphs have proven both extremely representative of what I've heard, and also seemingly out of left-field. Sometimes they've been so different I wonder if maybe I've received defective units. Other times, I look at them and say "yes, that's exactly how they sound!"
 
In many cases I've found that they 'can' be useful for comparative purposes.
 
shane
 

A frequency response graph will never tell you the whole story.  It doesn't even attempt to.  It's still a pretty useful bit of data though.
 
Some common traps that people fall into, in the subjective impression vs. objective measurement debate is assuming that a few tests are supposed to sum up the entirety of the headphone, and because they don't, then they have no value whatsoever.  There are many things that we do not have objective tests for, but that doesn't make the tests we do have any less useful.
 
The fact is that anything your ears can hear, a sufficiently advanced set of methodologies and equipment can measure.  Tests could be devised to test things like detail, speed, resolution, soundstage, and grain.  Other things like "musical" or "analytical" which are just personal preferences could even be put on a sliding scale.  There are microphones with greater sensitivity and resolution than the human ear.  Through studies, analyzing the output of a transducer, and people's subjective of it, it is possible to discover what objective measurements correspond to what subjective impressions.
 
If there were a bigger market, then it might be worthwhile for someone to develop these kind of tests.  For all I know they already exist but haven't been developed into psuedo commercial products like HeardRooms test rig.  Even then it wouldn't be perfect.  Nothing in science ever is.  Despite the imperfections inhearent in such a system, I will welcome the day when all audio equipment comes with comprehensive specifications and test results.
 
Speed: 7.5/10
Resolution: 7/10
Grainlessness: 6/10
Soundstage: 5/∞
Deviation From Linearity: -1.23
 
It's the next best thing to testing everything in person.

Objective?  On this forum?

I know...I know...it may have been silly of me to attempt any objective perspective on this forum, but I know there's a few of you objective folk out there - this thread's for you!
 
As shane55 has said, at the very least the FQ graphs give you a basis for comparison from one can to another.  If you line up the FQ curves for the HD650, DT880 and K701, they pretty much represent the subjective opinions of their respective headphones here on head-fi.  In that sense, if you actually know how to read measurements, they can be useful to a certain extent.
 
And I REALLY like maverickronin's idea.  I'm all for that level of standardization but I think there's too much money to be made in audiophoolery-land for that type of objectivity.

Quote:

joincoolkidclub said:

Objective?  On this forum?

Click to expand...

Someone's got to stand up for it!

Tyll is working more measurements, guys. I'm really not interested in whining about the methods this site, I believe subjective views should be backed up by objective views but if it's not I'll take the whatever evidence I can get. Objective measurements aren't anything new in audiophilia, Stereophile does them pretty well.
 
Yes, it's really no shock to me the DT880 measures quite close to its HD800/T1 brothers, it's a great headphone and what do you really expect from the $1K headphones? An accurate FR is an accurate FR. But small differences can mean a lot to some people. Plus: (ok, the smokin buds have a terrible treble roll off):
 

 
My question is, we usually think that aspects of a headphone like soundstage or detail as not being captured by frequency response graphs. As I understand it, soundstage comes from cues that let our brains know directionality and distance...  So naturally as it audible it must then be measurable, but how? Not on the FR graphs, I suppose, since speakers are supposed to be flat and they obviously have a soundstage... 

It probably wouldn't be too difficult to measure soundstage by reversing a HRTF or comparing it to a binaural recording of some reference quality speakers.  I think using the reverse HRTF would the best way to go since it will be more easily repeatable and not reliant on as much specific hardware.  I don't know if it would be more accurate, but I'm guessing it will be more consistent.

FR graphs are the linear representation of the measurements of pure tones. Db against Hz. (e.g. at x-frequency it produces y-db.).
 
Soundstage, cool, warm, overall sound signature, etc. are all functions of tones acting in concert (for tonal characteristics), and tones acting in concert combined with the actual physical makeup of the drivers, cups, pads...your own ears and their shape, etc, to create the soundstage. The exact same cans could certainly have a very different soundstage from one set of ear shapes to another. "Wow, these have killer soundstage." "Oh yeah? I think you're nuts. They're all headstage!" Both could be correct. (Ah the trouble with forums and... opinions based on unscientific observation, let alone with all the crazy different setups, rigs, file formats, decoders, and listening conditions out there.)
 
The only thing that you 'might' be able to quantify would again be used to compare from one to another, and only that one aspect understanding the outlying variables.
 
Assuming Headroom (as an example) measures consistently from one sample to another we can say... "hey, I love the way the T1 sounds... therefore I 'should' like any set of cans that have a similar graphic representation." Or... " I love the HD600 but think the highs roll off too quickly. If I find cans with a graphic representation similar but with better 'looking' highs, I might give them a try."
 
I'm tired... g'nite all.
 
shane