Measuring headphone neutrality
post-10655967
Thread Starter
Post #1 of 7

KlarkKentThe3rd

100+ Head-Fier
Joined
Apr 28, 2014
Messages
216
Reaction score
19
Joined
Apr 28, 2014
Posts
216
Likes
19
I am not exactly sure how old this thread is, but I'll bump it anyway.
 
There is something I, as a n00b, do not understand. How reliable are those graphs anyway? I mean, really? How are they made, and who makes them?
 
 
Also, I have figured out two ways to measure neutrality:
 
1) Relative neutrality. Pick a well recorded and well mastered track that has the full spectrum of frequencies. Listen to it through your test (°_°)phones. If some of the instruments (at the lower and higher ends) are difficult to pick out, your pair is not relatively neutral.
 
2) Almost absolute neutrality. Make a high quality binaural recording of your (°_°)phones. Then, listen to the source file and your recording back to back, through a good pair (can be the same one). If the altered!original and altered!altered files sound almost indistinguishable, then your test pair is pretty much as neutral as it gets.
 
     Share This Post       
post-10656013
Post #2 of 7

AzraelDarkangel

500+ Head-Fier
Joined
Dec 28, 2006
Messages
580
Reaction score
20
Joined
Dec 28, 2006
Posts
580
Likes
20
On goldenears graph and waterfall, the HD600 is the flatest I've ever seen:
http://en.goldenears.net/8072
 
     Share This Post       
post-10656121
Post #3 of 7
Joined
Jan 4, 2008
Messages
17,134
Reaction score
9,984
Location
Fukuoka, Japan
Joined
Jan 4, 2008
Location
Fukuoka, Japan
Posts
17,134
Likes
9,984
Website
www.hear.reviews
I split your thread necro bump off into a new thread for you instead.
 
     Share This Post       
post-10669516
Post #4 of 7

ieee754

100+ Head-Fier
Joined
Nov 11, 2013
Messages
167
Reaction score
14
Joined
Nov 11, 2013
Posts
167
Likes
14
On goldenears graph and waterfall, the HD600 is the flatest I've ever seen:
http://en.goldenears.net/8072
I think that HiFiMAN headphones containing '5' and/or '6' in their names are the flattest measuring headphones around (especially in the subbass region).
 
http://www.innerfidelity.com/images/SennheiserHD600.pdf
http://www.innerfidelity.com/images/HiFiMANHE500.pdf
http://www.innerfidelity.com/images/HiFiMANHE6.pdf
 
And probably HE-560, but there are no measurements yet. Also HE-6 have the cleanest waterfall graph I have ever seen, but I can't post it here.
 
     Share This Post       
post-10669532
Post #5 of 7

ieee754

100+ Head-Fier
Joined
Nov 11, 2013
Messages
167
Reaction score
14
Joined
Nov 11, 2013
Posts
167
Likes
14
  I am not exactly sure how old this thread is, but I'll bump it anyway.
 
There is something I, as a n00b, do not understand. How reliable are those graphs anyway? I mean, really? How are they made, and who makes them?
 
 
Also, I have figured out two ways to measure neutrality:
 
1) Relative neutrality. Pick a well recorded and well mastered track that has the full spectrum of frequencies. Listen to it through your test (°_°)phones. If some of the instruments (at the lower and higher ends) are difficult to pick out, your pair is not relatively neutral.
 
2) Almost absolute neutrality. Make a high quality binaural recording of your (°_°)phones. Then, listen to the source file and your recording back to back, through a good pair (can be the same one). If the altered!original and altered!altered files sound almost indistinguishable, then your test pair is pretty much as neutral as it gets.
I believe that graphs are extremely reliable and that FR graph can tell you 90% of the headphone sound quality, if you can read them of course. However when you compare two headphone graphs you must remember that they should be measured in identical environment, using the same compensation curve, etc.
 
Your ideas for subjective measurements sound pretty good, I think that's the way most people evaluate headphones using ears, either by how they feel it (most subjective way) or by comparing them to the reference pair.
 
     Share This Post       
post-10669938
Post #6 of 7

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Messages
9,085
Reaction score
4,423
Joined
Jul 2, 2011
Posts
9,085
Likes
4,423
 
  I am not exactly sure how old this thread is, but I'll bump it anyway.
 
There is something I, as a n00b, do not understand. How reliable are those graphs anyway? I mean, really? How are they made, and who makes them?
 
 
Also, I have figured out two ways to measure neutrality:
 
1) Relative neutrality. Pick a well recorded and well mastered track that has the full spectrum of frequencies. Listen to it through your test (°_°)phones. If some of the instruments (at the lower and higher ends) are difficult to pick out, your pair is not relatively neutral.
 
2) Almost absolute neutrality. Make a high quality binaural recording of your (°_°)phones. Then, listen to the source file and your recording back to back, through a good pair (can be the same one). If the altered!original and altered!altered files sound almost indistinguishable, then your test pair is pretty much as neutral as it gets.
I believe that graphs are extremely reliable and that FR graph can tell you 90% of the headphone sound quality, if you can read them of course. However when you compare two headphone graphs you must remember that they should be measured in identical environment, using the same compensation curve, etc.
+1
graphs should be used to compare 2headphones, one you own, one you want. and measurements should indeed come from the same source else it really doesn't mean anything.
I got used to Tyll's raw measurements, maybe I would check at the compensated values with some olive&welti eye candy because that compensation is a lot more "in phase" with what I really hear. but yeah anyway I'm now used to work with innerfidelity's raw data so as long as Tyll doesn't change something in his equipment, it's an amazingly reliable source of data for me.
golden ear compensated graphs make me cry a little, they use a detail exterminator, the results are so smooth that they would hide mountains and fault lines on a world map.
 
     Share This Post       
post-10672435
Post #7 of 7

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Messages
21,582
Reaction score
3,621
Location
Hollywood USA
Joined
Nov 16, 2004
Location
Hollywood USA
Posts
21,582
Likes
3,621
Website
www.facebook.com
The thing is, those smoothed out graphs are probably a lot more representative of what your ears actually hear. I'd actually rather see graphs corrected for human hearing, with Fletcher Munson and a bit of smoothing to reflect the actual sensitivity of the ear.
 
     Share This Post       

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top