Use tone generator to smoothing out treble significantly improves tonality, with some mystery unresolved

[moosso]

(Sorry that I'm not a native english speaker so some of the sentences may not be precise.)

Last couple days I'm trying to use tone generator improve my EQ profile, the result is beyond my expectation. Previously when I try EQ preset from AutoEQ I feel my gear become more "harman" like but there always something wrong in somewhere, oratory1990's one sometimes work well and sometimes still feel weird, so I usually did EQ with REW myself that only add peak filter to bass and mid but not treble (only use high shelf on there), in addition I use a sweep wav to check the sound is smooth enough after EQ. I though that's the best strategy until I try to use tone generator, the tone generator shows me some narrow peaks and dips in the treble range, after fix them I feel the tonality improved significantly, the sound balance is similar but everything is more clear and dimensional.

For example my DT770 (80Ohm with Dekoni Velour pad) have peaks at 4.7k 6.7k 8.8k and a dip at 7.8k, then I fix them with this EQ profile:

Filter  1: ON  PK       Fc   20.00 Hz  Gain   3.00 dB  Q  0.900
Filter  2: ON  PK       Fc   70.00 Hz  Gain   2.00 dB  Q  0.900
Filter  3: ON  PK       Fc   105.0 Hz  Gain   4.90 dB  Q  3.000
Filter  4: ON  PK       Fc   130.0 Hz  Gain  -4.90 dB  Q  2.200
Filter  5: ON  PK       Fc   215.0 Hz  Gain   6.50 dB  Q  2.000
Filter  6: ON  PK       Fc   955.0 Hz  Gain  -1.50 dB  Q  4.500
Filter  7: ON  PK       Fc    2550 Hz  Gain  -1.20 dB  Q  2.800
Filter  8: ON  PK       Fc    3350 Hz  Gain   5.20 dB  Q  4.000
Filter  9: ON  PK       Fc    4700 Hz  Gain  -2.00 dB  Q  7.000
Filter 10: ON  PK       Fc    6700 Hz  Gain  -2.00 dB  Q  7.000
Filter 11: ON  PK       Fc    7800 Hz  Gain   3.50 dB  Q  7.000
Filter 12: ON  PK       Fc    8800 Hz  Gain  -2.00 dB  Q  7.000

the last 4 filter is for the peaks and the dip, after apply it the tone generator shows me a more smooth treble, I can confirm it brings my DT770 to the next level when I listening to the music. However, this profile will most likely not work for you, because the location of peaks and dip maybe different, and Dekoni Velour pad changes frequence response (less bass and cause a dip around 8k, I remember solderdude's review metion that).

Later I try this to my other closed back SRH1540, the tone generator shows me peaks at 4.7k 6.7k and a dip at 8.0k (this dip also metioned in solderdude's review, it's caused by pad like 770), after fix them SRH1540 gives me IEM levels clear sound that I thought I will never hear on headphone at this price point. I have not try it to open backs so next I want to see if I could bring the same improvement to my HD560s.

Still, there some mysteries unresolved:

• I hear a 4.7k peak on both DT770 and SRH1540, while most measurement did not show it or show a dip there, is this just my ears and it's not actually as loud as I perceive? or it's the "resonance" peoples said?
• The dip around 8k was caused by pad, Dekoni Velour and Shure Alcantara pad have this issue while DT770 stock pad and SRH940 pad haven't, usually pad only changes bass so I don't know why some pad will cause a dip at treble range.
• The fix to the dip and the peaks can't be more than 4db, if I use a larger gain I will hear a more smooth treble with tone generator, but the music start sound boring, I have to allow small peaks and dips there...
• Some of amirm's review points out locations of treble that can't be EQed beucase a group delay peak there, so it maybe risky to use this method while I don't have the equipment to measure group delay, still I don't know what will happen when a filter is applied at a group delay peak location

Also, I may get a stealth at the end of the month if DCA have black friday sales like last year (it will definitely hurt my bank account), then I could know how far I behind from it.

 

[adama99]

Measurement rigs provide a reproducible test platform for comparing one headphone against another and/or against some sort of defined reference target. However, the shape of your pinnae (outer ear anatomy) and the sensitivity of your stereocilia (hairs in your inner ear that detect various sound frequencies) are unique to you - so what you perceive will not always match what a given measurement rig recorded.

Pads alter specific aspects of frequency response due to different materials in use, the thickness/diameter of the pad, and whether it adds additional layers of material between the driver and your ear. As an example, you can see in the below chart that three different pads on the same headphone show substantial differences in treble characteristics.

.

(Graph courtesy of @solderdude at DIY Audio Heaven)

The goal of using EQ with headphones is not to make the frequency response completely flat - as you discovered, doing so quickly makes any musical content boring. Rather, the goal is to eliminate glaring problems that either make listening unpleasant (or painful) or that negatively affect the balance and quality of the music. When adjusting EQ to your personal taste, use music you're very familiar with to listen for and correct specific problems with the overall sound rather than sweeping with a tone generator.

 

[hex168]

Sigfried Linkwitz wrote about EQing headphones using a sweep from a signal generator, so you are in good company:

Reference earphones

"We used an audio signal generator that was smoothly tunable by hand to sweep the frequency range. It was not too difficult to hear where the headphone response had peaks, though we were not able to quantify their magnitudes. The next step was to build an equalizer that would notch the peaks and we continued adjusting it until we were satisfied with the smoothness of auditory response."

 

[moosso]

The goal of using EQ with headphones is not to make the frequency response completely flat - as you discovered, doing so quickly makes any musical content boring. Rather, the goal is to eliminate glaring problems that either make listening unpleasant (or painful) or that negatively affect the balance and quality of the music. When adjusting EQ to your personal taste, use music you're very familiar with to listen for and correct specific problems with the overall sound rather than sweeping with a tone generator.

So pad did change treble. And you're right, the tone generator is very useful to finding the exact location of peaks and dips, but I have to use the music I listen about hundred times to adjust the gain.

Sigfried Linkwitz wrote about EQing headphones using a sweep from a signal generator, so you are in good company:

Reference earphones

"We used an audio signal generator that was smoothly tunable by hand to sweep the frequency range. It was not too difficult to hear where the headphone response had peaks, though we were not able to quantify their magnitudes. The next step was to build an equalizer that would notch the peaks and we continued adjusting it until we were satisfied with the smoothness of auditory response."

That link is impressive, I remember diy audio heaven also provide guides for design analog filter that can be used in any environment.