Wish I'd found this place sooner
May 14, 2013 at 11:52 AM Post #16 of 30
Just don't focus too much on soundstage size, that's not primarily what headphones are about. If you can pick out details in the recording, including their distinct and believable/rational location somewhere in the soundstage, that's for the most part good enough. You can gamble with a K701 if you'd like but people tend to argue to no end even if they are observing the same thing, let alone when they don't perceive the same things. Others say the soundstage is too wide, there's no clear center, or it enhances the worst of headphones (strong L-C-R with barely anything in-between), some say it's just plain weird, etc. It's too sensitive to source material and decoding hardware, and also the wear on the pads. I've once listened to a K701 (with my old NOS DAC) with the orchestra set in front of and below the band (Epica's "The Classical Conspiracy") like on an orchestra pit at the Kodak Theater, which no other equipment, speakers included, has ever done (the orchestra was usually behind or lost in the same z-axis as the band). I've listened to a demo unit Q701 with the pads screwed up, for some reason the thicker section on one ear cup was towards the front while the other was towards the top, and the percussion was in front of the other instruments. I've listened to a K701 with worn pads on the same amp+DAC (my Meier Cantate.2) and it wasn't that screwed up.
 
May 14, 2013 at 2:47 PM Post #17 of 30
May 24, 2013 at 11:48 AM Post #24 of 30
Quote:
extech said:
I didn't actually mean telling the brand - just if a blind test could uniformly agree on comparative quality.
Click to expand...

 
Throw me a bone here, like more details on the experiment design, comparative quality of which equipment?
 
What I was getting at was that if for example you "blind" test a guy with an HD600 and the K701, you can be literally blind but your skin can tell the difference between an oblong earpad and round earpads with bumps on the headband.
 
May 24, 2013 at 12:04 PM Post #25 of 30
Well, there are essentially two areas that could be measured, you can call them objective and subjective. Rating a headphone objectively could be the typical survey format - categories might be highs, midrange, lows, soundstage, and whatever other categories you aficionados would like to see there. the choices could be from 1 to 5. in the case of highs, e.g. it could go from too weak, to just right, to too strong. Then there could be a subjective section. Those questions would deal with overall impression, music genre, comfort, etc. Results could be measured to see how much of what we read here is subjective and how much objective, in the sense that if there is highly consistent response, we know that the results are objective, but if it's all over the map, they are largely subjective.
 
May 24, 2013 at 12:36 PM Post #26 of 30
Quote:
extech said:
Well, there are essentially two areas that could be measured, you can call them objective and subjective. Rating a headphone objectively could be the typical survey format - categories might be highs, midrange, lows, soundstage, and whatever other categories you aficionados would like to see there. the choices could be from 1 to 5. in the case of highs, e.g. it could go from too weak, to just right, to too strong. Then there could be a subjective section. Those questions would deal with overall impression, music genre, comfort, etc. Results could be measured to see how much of what we read here is subjective and how much objective, in the sense that if there is highly consistent response, we know that the results are objective, but if it's all over the map, they are largely subjective.
Click to expand...

 
In which case they won't be conducted totally blind, nor need to be. Still, the problems will be that since that test format will rely on the judgment of the listener, then it is potentially subjective, although quantified. It's like Scarlett Johansen walking in and some go, "10!", one guy will sing, "...from a 1 to 10, she's a certified 20...but that's just ain't me!" (but if everyone else went, "10!" then it's false), and I'm gonna go, "meh...where's Sushmita Sen?"
 
Since it does focus on on segmented areas of the frequency response for example then what it can avoid is one guy going, "Guarneri homage or Stax treble is worth the $$$$$$!" while another goes, "but I can't hear the beat on those, what's all the $$$$ paying for?!" and instead structures the results so that it will be presented as "Guarneri Homage/Stax - Treble = 5/5, Bass = 2/5" instead of "I prefer the Focal 607V/Sennheiser HD600/Grado RS1 over the Guarneri/Stax, but I have to admit the treble on those is delicious."
 
Objective would be better done using electronics measuring equipment, but of course these aren't without their flaws. Two sources with near-identical flat response and output signal strength can still sound audibly different.
 
For soundstage however maybe there is a way to test it objectively. Pop in CDs, primarily but not limited to, test CDs used in car audio competitions (Focal has one) and have the listener mark a spot on the 3D space around his head where he perceives the staging and imaging test sounds are coming from.
 
May 24, 2013 at 12:44 PM Post #27 of 30
Quote:
protegemaniac said:
 
>In which case they won't be conducted totally blind,
 
Not clear what you mean. "Blind" in this context, simply means no knowledge of what headphone one is listening to.
 
nor need to be. Still, the problems will be that since that test format will rely on the judgment of the listener, then it is potentially subjective, although quantified. It's like Scarlett Johansen walking in and some go, "10!", one guy will sing, "...from a 1 to 10, she's a certified 20...but that's just ain't me!" (but if everyone else went, "10!" then it's false), and I'm gonna go, "meh...where's Sushmita Sen?"
 
As I said, there are objective and subjective sections, and there are results that indicate objective and subjective results. Those words mean two different things in the two different contexts. 
 
Since it does focus on on segmented areas of the frequency response for example then what it can avoid is one guy going, "Guarneri homage or Stax treble is worth the $$$$$$!" while another goes, "but I can't hear the beat on those, what's all the $$$$ paying for?!" and instead structures the results so that it will be presented as "Guarneri Homage/Stax - Treble = 5/5, Bass = 2/5" instead of "I prefer the Focal 607V/Sennheiser HD600/Grado RS1 over the Guarneri/Stax, but I have to admit the treble on those is delicious."
 
Price is not relevant - it would be prejudicial.
 
Objective would be better done using electronics measuring equipment, but of course these aren't without their flaws. Two sources with near-identical flat response and output signal strength can still sound audibly different.
 
Exactly. Measuring equipment results are easy and already known.
 
For soundstage however maybe there is a way to test it objectively. Pop in CDs, primarily but not limited to, test CDs used in car audio competitions (Focal has one) and have the listener mark a spot on the 3D space around his head where he perceives the staging and imaging test sounds are coming from.
Click to expand...

 
May 25, 2013 at 2:29 AM Post #28 of 30
Quote:
extech said:
 
Click to expand...
Not clear what you mean. "Blind" in this context, simply means no knowledge of what headphone one is listening to.
 
Click to expand...

 
OK, I'll repeat my example. Tell me to blind test a headphone, I'm blindfolded, I don't know what it is. I feel round earpads with unequal pressure on my cheeks and behind my ears, and seven bumps on the top of my head. "K70x!" Or I can feel oblong earpads with two wide and soft foam pads on my head. "HD650!" I now have knowledge of what headphone I am listening to.
 
May 25, 2013 at 11:29 AM Post #30 of 30
Quote:
extech said:
That may apply for a small percentage of available models. Not statistically relevant.
Click to expand...

 
Good luck with that. The percentage will be small if you use a wider variety of headphones, but if you take only the most popular ones - the ones that garner the most reviews, or the most polarizing opinions - it becomes significant. The only thing that will be easy to mask will be blind-testing Grado vs Grado.
 
Here's one thing you can add to the experiment to figure out if people can tell what they are : you either have participants not into head-fi (the hobby, not the forum), or if they are, include in the questionnaire their guess as to what headphone they just tested. If enough people can guess at least within the same model range from the correct manufacturer, even with no quantifiable way to account for exactly how much, then it should at least be factored in the analysis of the data.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top