Fidelity/Detail/Resolution on reviews

[llmercll]

I EQ'd my gates according to oratory's filters and did some more listening. I really enjoyed the near perfect frequency response and lack of noise. Regardless, the sound remained less detailed, and with a worse soundstage when compared with my Sundaras. It felt like there was a veil blocking out what almost sounds like the instruments being right in front of me with the Sundaras. They're overall less fun to listen to because of this.

I tried your recommendation CiX and interestingly I notice dips around 7Khz on both my gates and sundaras. Not sure if it's me or the headphone, but I notice the gates measured a dip in 7khz in a few reviews. The other frequencies seems very level.

 

[Madlop26]

Fidelity?, detail?, resolution?..... do we have any clear definitions?, how can we discuss something which has no consensus? futile; IMO as far as data can support, all are synonym to frequency response. Think about it, if a headphone is capable to reproduce all frequencies humans can hear at adequate level and without meaningful distortion, common sense says there is no reason for any of them to be different.

 

[CiX]

If you need 26 points of EQ to make the the headphones sound right to your ears ... there is something wrong at a certain point.

The reason my EQ ends up with so many points is that this is an entry-level planar headphone, which typically has certain frequency ranges that are hard to control, unlike Dan Clark Audio's headphones, which solve most of the inherent problems of planar drivers at the design stage through acoustic structures.

Since I need to use this headphone for mixing work, I wanted to resolve most of these issues up front — hence an EQ like this.
And some of these points may well be redundant, or simply misjudgments on my part.

The interaction between an over-ear headphone and the ear inside the pad cavity is neither as direct as listening to speakers, nor as simple as with IEMs. So given that my goal here is to change the details of the sound rather than preserve this headphone's original character, I don't think an EQ of this magnitude is excessive, unless the Q values are too high.

That said, an EQ method like this still can't solve everything perfectly.
Factors such as the temperature inside the ear cups, the thickness and compression of the pads, and how well they seal against the head will cause the resonances above 6kHz and the content below 200Hz to drift slightly in both frequency and level.

 

[heraldo_jones]

The reason my EQ ends up with so many points is that this is an entry-level planar headphone, which typically has certain frequency ranges that are hard to control, unlike Dan Clark Audio's headphones, which solve most of the inherent problems of planar drivers at the design stage through acoustic structures.

Since I need to use this headphone for mixing work, I wanted to resolve most of these issues up front — hence an EQ like this.
And some of these points may well be redundant, or simply misjudgments on my part.

The interaction between an over-ear headphone and the ear inside the pad cavity is neither as direct as listening to speakers, nor as simple as with IEMs. So given that my goal here is to change the details of the sound rather than preserve this headphone's original character, I don't think an EQ of this magnitude is excessive, unless the Q values are too high.

That said, an EQ method like this still can't solve everything perfectly.
Factors such as the temperature inside the ear cups, the thickness and compression of the pads, and how well they seal against the head will cause the resonances above 6kHz and the content below 200Hz to drift slightly in both frequency and level.

I'm afraid you haven't really understood the concept of why and how headphones should be EQ'd, and, more importantly, how to properly interpret the graphs you see on SQUIG.

Using 26 EQ points in minimum phase with Q values that narrow is going to introduce more phase-related problems than it solves. On top of that, you've EQ'd out the measurement rig's fixture resonance/cancellation around the 9–10 kHz region. And that's not even getting into the fact that your headphones aren't going to be exactly the same unit as the one measured by the measurbator.

No offense, but I honestly think the EQ work you've done is a complete mess. And the worst part is that it may actually sound good to you, which makes it even harder to identify the problems by ear.

If I were you, I'd spend a bit more time researching how headphone EQ should actually be done and how to interpret measurement data correctly.

I'm out.

 

[solderdude]

I EQ'd my gates according to oratory's filters and did some more listening. I really enjoyed the near perfect frequency response and lack of noise, and they are certainly a good pair of IEMs for $20. Regardless, the sound remained somewhat less detailed, and with a worse soundstage when compared with my Sundaras. It felt like there was a veil blocking out what almost sounds like the instruments being right in front of me with the Sundaras. They're overall less fun to listen to because of this.

I tried your recommendation CiX and interestingly I notice dips around 7Khz on both my gates and sundaras. Not sure if it's me or the headphone, but I notice the gates measured a dip in 7khz in a few reviews. The other frequencies seems very level.

Sundara has a dip close to 7kHz (and other frequencies too) A 10dB dip means it sounds about half as loud.

 

[CiX]

I'm afraid you haven't really understood the concept of why and how headphones should be EQ'd, and, more importantly, how to properly interpret the graphs you see on SQUIG.

Using 26 EQ points in minimum phase with Q values that narrow is going to introduce more phase-related problems than it solves. On top of that, you've EQ'd out the measurement rig's fixture resonance/cancellation around the 9–10 kHz region. And that's not even getting into the fact that your headphones aren't going to be exactly the same unit as the one measured by the measurbator.

No offense, but I honestly think the EQ work you've done is a complete mess. And the worst part is that it may actually sound good to you, which makes it even harder to identify the problems by ear.

If I were you, I'd spend a bit more time researching how headphone EQ should actually be done and how to interpret measurement data correctly.

I'm out.

Since planar drivers introduce even more variables into the HpTF, the measurement graphs on SQUIG — while useful as a reference — aren't something I pay that much attention to, especially above 4kHz.

My current setup is the FiiO JT7 with FT1C pads.
My corrections above 200Hz were made almost entirely without referencing SQUIG, and below 200Hz I also tuned by ear.

The boost at 10kHz isn't purely a correction for the headphone itself: even when listening to speakers, 10kHz has always sounded slightly recessed to me,
so this compensation also accounts for the dip in my own hearing at that frequency, not an attempt to fill in a null.
(Nulls are much narrower, and their frequency drift with fit, and fill in a null will cause more issue.)

I'm well aware that high-Q corrections can produce ringing — especially when the resonance drifts and the correction no longer lines up. As for how high a Q value or how much gain counts as "too much," ultimately that has to be verified by listening for audible decay tails.
And after several multi-hour listening sessions, I haven't found this to be an issue.

That said, the points you raised are a fair reminder: my original post was missing a discussion of the side effects this kind of EQ approach can bring.