j-curve
100+ Head-Fier
- Joined
- Apr 15, 2002
- Posts
- 489
- Likes
- 11
By way of a quick introduction, I'm going to post some graphs in this thread which look like headphone response plots. With a bit of luck people will be inspired to contribute their interpretations of the graphs and their opinions on whether the graphs have any value, based on what they've heard when actually listening to any of the headphones.
Since I don't want to alienate and bore people with a page of drivel about how I came up with these, all that stuff can be found in the following thread, DIY Headphone Response Plots - Methodology, for those who are interested. Please post your criticisms of the methodology there, not here, since it would be nice not to get bogged down in the technicalities here, where people might have something interesting to say about the headphones. Thanks!
Interpretation of the Graphs
The horizontal scale is the frequency divided by 10. So where it says "16", that's 160 Hz. Each graph goes from 10 Hz to 20 kHz. [The final marker on the right hand side is 16 kHz].
The vertical scale is in decibels and is meant to show the response of the headphone relative to the HD580. In other words, the measurement system was tweaked so that the HD580 measured flat. The idea is that peaks signify frequencies where a headphone sounds louder than the HD580, and dips signify the opposite, i.e. quieter than the HD580.
The red trace is the right channel, the white trace is the left.
Enjoy, and let us know how what you see stacks up with what you hear!
Edited Jan 15
Note: The 10-40Hz region should not necessarily be interpreted as an indicator of bass extension of the headphone (although ideally it would be). If the curve is flattish down there it could be a warning that traffic noise or some other external low frequency noise has contributed to the whole bass spectrum. By contrast, a dropping curve down below 40Hz indicates that either:-
i) there was no significant background noise, or
ii) the headphone has good isolation!
In either case, you can trust that the bass response came from the headphone, not vehicles or air-conditioning systems.
Edited Jan 25
Although the HD580 reference was adjusted to plot essentially flat down to 10Hz, subsequent testing of HD580's and HD600's in quieter environments has shown that the HD580 actually drops at the low end, about 4dB per octave from 40Hz down, relative to the chosen reference. What this means is that a headphone which drops off at any slower rate or lower frequency than this probably has better bass extension than the HD580. Two caveats apply - one in relation to background noise, which is discussed at length in the methodology thread, the other being that earpad seal (or lack thereof) can have a major impact on low end response, and these tests are for the most part one-offs. Earpad seal is also discussed in the methodology thread.
The good news is that since updating many of the earlier graphs, I now have greater confidence that there is meaning below 40Hz. YMMV.
Since I don't want to alienate and bore people with a page of drivel about how I came up with these, all that stuff can be found in the following thread, DIY Headphone Response Plots - Methodology, for those who are interested. Please post your criticisms of the methodology there, not here, since it would be nice not to get bogged down in the technicalities here, where people might have something interesting to say about the headphones. Thanks!
Interpretation of the Graphs
The horizontal scale is the frequency divided by 10. So where it says "16", that's 160 Hz. Each graph goes from 10 Hz to 20 kHz. [The final marker on the right hand side is 16 kHz].
The vertical scale is in decibels and is meant to show the response of the headphone relative to the HD580. In other words, the measurement system was tweaked so that the HD580 measured flat. The idea is that peaks signify frequencies where a headphone sounds louder than the HD580, and dips signify the opposite, i.e. quieter than the HD580.
The red trace is the right channel, the white trace is the left.
Enjoy, and let us know how what you see stacks up with what you hear!
Edited Jan 15
Note: The 10-40Hz region should not necessarily be interpreted as an indicator of bass extension of the headphone (although ideally it would be). If the curve is flattish down there it could be a warning that traffic noise or some other external low frequency noise has contributed to the whole bass spectrum. By contrast, a dropping curve down below 40Hz indicates that either:-
i) there was no significant background noise, or
ii) the headphone has good isolation!
In either case, you can trust that the bass response came from the headphone, not vehicles or air-conditioning systems.
Edited Jan 25
Although the HD580 reference was adjusted to plot essentially flat down to 10Hz, subsequent testing of HD580's and HD600's in quieter environments has shown that the HD580 actually drops at the low end, about 4dB per octave from 40Hz down, relative to the chosen reference. What this means is that a headphone which drops off at any slower rate or lower frequency than this probably has better bass extension than the HD580. Two caveats apply - one in relation to background noise, which is discussed at length in the methodology thread, the other being that earpad seal (or lack thereof) can have a major impact on low end response, and these tests are for the most part one-offs. Earpad seal is also discussed in the methodology thread.
The good news is that since updating many of the earlier graphs, I now have greater confidence that there is meaning below 40Hz. YMMV.
