Anyone else find they *don't* prefer the Harman curve for EQ?

[ChickenChaser]

Having reached my personal endgame - by which I mean I'm not dissatisfied with anything in my current setup in any way, and if I do dabble in anything else from here on it's just out of curiosity and not out of the "need to get better sound" I had for maybe a year and a half - here are my thoughts on and experiences with EQ. I've been using all this with a Lyr 2 (LSST) - Modi 2, and switching to an Atom actually sounds like an upgrade to me. I'm working on replacing the Modi 2 with a KTB.

I find the Rosson RAD-0's tuning perfect for enjoyment without any EQ. Occasionally I add an EQ with a slight bass shelf, boost the lower mids while lowering the upper mids and reduce the ~7-8khz peak. All literally less than 1db in either direction. If anything I'm just stuck in the habit of playing around with EQ, but with the RAD-0 I never end up going past plus or minus 0.4-0.8dB.

The Verite Closed sounds great to me with no EQ, but it also needs a stronger version of the above formula for me to blast the volume loud with extreme metal, mostly because of the elevated upper-mids. So the only EQ I ever put on it makes it less like the Harman curve, as far as I know.

In general, I like having these different flavors to rotate between. I really don't think I would ever be optimally happy with just one signature for life. There's a sweet spot just after I've gotten used to a new signature but before I'm really "accustomed" to it where the amount of engagement I have with the music I'm listening to reaches a peak. Having a few cans to rotate between is my actual "endgame," and I'm really, really happy rotating between these headphones and really could see myself doing it forever with no more big changes.

The SR1a is the only thing left that intrigues me. I can't imagine the Empyrean or Abyss sounding better than the RAD-0 for "pure pleasure" listening. Apparently you can get Abyss-sounding bass on the SR1a anyway. And that $350 driver replacement, for something so expensive, is very appealing to me. As are the Rosson and ZMF's lifetime warranties.

The ER3XR sounds far better to me than the Harman in-ear target. I was actually shocked how much I hated it when I turned the EQ on. I definitely feel like I can hear some distortion, though. According to the Klippel listening test, with open back headphones and a mildly noisy environment without straining I can consistently pick up distortion at -26dB. Make of either claim what you will. It's still good enough to be an "endgame" IEM for me, anyway.

I still own the first HD800 I picked up as my first step into "the hobby" and put the SDR mod in. The full Oratory EQ settings are far too extreme for my tastes, I usually put about 50% of his values on while slightly raising the "mudrange" around 200HZ instead of cutting it like Oratory does. I also have settings I use at different times put to 25% and 75% of his values. When not at the 50% setting, I use the 25% more often than the 75%. I feel like it has something really uniquely special even after EQ'd like this.

I picked up an HD6XX from Massdrop a couple years ago, and it honestly sounded like a muddled mess. I tried getting used to it, I hated it too much to even bother. I did like the HD600 I heard, but that was before I got hold of an HD800 and it ruined me on anything I heard from then onwards below its (used) price range.

I loved the LCD4. Overall it may have been my favorite headphone I've heard yet, but I didn't love it enough to justify the price owning it out of warranty. Occasionally it felt slightly too dark, and I had the HD800 for those moments, but I actually preferred it without the Reveal plug-in regardless.

In other words: I've found the Harman curve useful for making the HD800 one of the most enjoyable headphones I've ever heard, but that's about it. And even there, even with extreme metal like Meshuggah or Car Bomb, I don't prefer it set all the way to Harman values.

 

[majingotan]

The SR1a is the only thing left that intrigues me.

After hearing the SR1a quite a few times with various amps, from McIntosh MHA150, Jotunheim R and Raal's very own SRH1a amp, I never felt the need to EQ them as it's tuned close to a room treated speaker response. My only con is that I prefer not using the interface and just go direct to Jotunheim R since the MHA 150 can clip at loud volumes while the Jotunheim R and the SRH1a handles loud volumes with authority

IMO, if I ever go back to headphone world, I'd go straight to the SR1a or E-stats like the Dan Audio Voce (I prefer this to the Stax SR-009)

 

[wwenze]

Is the Harman curve one for enjoyment or accuracy? (To compensate for physical effects)

What happens if someone did a sine sweep test with headphones? Will the result be the same as the Harman curve?

 

[solderdude]

The Harman curve is for preference and is loudness and 'training' and even music/recording dependent.
Preference, loudness and training differs from person to person.

One would have to be really experienced to do get an accurate FR response from a sinesweep.
It would be better to use Griesingers method and/or use band limited noise.

There is no one curve fits all but in general some boosted lows and 'flattish' response (no high peaks or dips) is preferred by most people.
Manufacturers sell the most to most people not to the minority.

 

[BillG]

In other words: I've found the Harman curve useful for making the HD800 one of the most enjoyable headphones I've ever heard, but that's about it.

I just use it as a guideline for IEMs that vary quite a bit from the target, and only a rough one at that. I'll take Oratory's settings and adjust them somewhat to taste by ear.

For the two that I own that measure in proximity to Harman, I leave EQ alone as they sound quite good running virgin...

 

[wwenze]

One would have to be really experienced to do get an accurate FR response from a sinesweep.

Hopefully that hurdle can be eventually overcome as it would give us more accurate hearing measurements.

Currently I'm using decreasing major scale (C major to be specific) from 250Hz downwards to apply room correction via hearing. I managed to bring the variation down to 4dB and I could still hear imperfections but I got bored of mucking around with the parametric EQ. The hard part is finding the correct Q and the corresponding center frequency iteratively, made worse when real life peaks don't follow a smooth curve aren't even symmetric.

It is still way better than nothing and I strongly recommend it to anyone who doesn't have room correction hardware.

Ideas for future improvements include using arpeggio and octaves to ensure equal loudness across bigger frequency ranges, and graphical equalizer at the note frequencies with high Q to adjust the notes individually to generate the response to be corrected via parametric EQ.

 

[Fluffy]

Having reached my personal endgame - by which I mean I'm not dissatisfied with anything in my current setup in any way, and if I do dabble in anything else from here on it's just out of curiosity and not out of the "need to get better sound" I had for maybe a year and a half - here are my thoughts on and experiences with EQ. I've been using all this with a Lyr 2 (LSST) - Modi 2, and switching to an Atom actually sounds like an upgrade to me. I'm working on replacing the Modi 2 with a KTB.

I find the Rosson RAD-0's tuning perfect for enjoyment without any EQ. Occasionally I add an EQ with a slight bass shelf, boost the lower mids while lowering the upper mids and reduce the ~7-8khz peak. All literally less than 1db in either direction. If anything I'm just stuck in the habit of playing around with EQ, but with the RAD-0 I never end up going past plus or minus 0.4-0.8dB.

The Verite Closed sounds great to me with no EQ, but it also needs a stronger version of the above formula for me to blast the volume loud with extreme metal, mostly because of the elevated upper-mids. So the only EQ I ever put on it makes it less like the Harman curve, as far as I know.

In general, I like having these different flavors to rotate between. I really don't think I would ever be optimally happy with just one signature for life. There's a sweet spot just after I've gotten used to a new signature but before I'm really "accustomed" to it where the amount of engagement I have with the music I'm listening to reaches a peak. Having a few cans to rotate between is my actual "endgame," and I'm really, really happy rotating between these headphones and really could see myself doing it forever with no more big changes.

The SR1a is the only thing left that intrigues me. I can't imagine the Empyrean or Abyss sounding better than the RAD-0 for "pure pleasure" listening. Apparently you can get Abyss-sounding bass on the SR1a anyway. And that $350 driver replacement, for something so expensive, is very appealing to me. As are the Rosson and ZMF's lifetime warranties.

The ER3XR sounds far better to me than the Harman in-ear target. I was actually shocked how much I hated it when I turned the EQ on. I definitely feel like I can hear some distortion, though. According to the Klippel listening test, with open back headphones and a mildly noisy environment without straining I can consistently pick up distortion at -26dB. Make of either claim what you will. It's still good enough to be an "endgame" IEM for me, anyway.

I still own the first HD800 I picked up as my first step into "the hobby" and put the SDR mod in. The full Oratory EQ settings are far too extreme for my tastes, I usually put about 50% of his values on while slightly raising the "mudrange" around 200HZ instead of cutting it like Oratory does. I also have settings I use at different times put to 25% and 75% of his values. When not at the 50% setting, I use the 25% more often than the 75%. I feel like it has something really uniquely special even after EQ'd like this.

I picked up an HD6XX from Massdrop a couple years ago, and it honestly sounded like a muddled mess. I tried getting used to it, I hated it too much to even bother. I did like the HD600 I heard, but that was before I got hold of an HD800 and it ruined me on anything I heard from then onwards below its (used) price range.

I loved the LCD4. Overall it may have been my favorite headphone I've heard yet, but I didn't love it enough to justify the price owning it out of warranty. Occasionally it felt slightly too dark, and I had the HD800 for those moments, but I actually preferred it without the Reveal plug-in regardless.

In other words: I've found the Harman curve useful for making the HD800 one of the most enjoyable headphones I've ever heard, but that's about it. And even there, even with extreme metal like Meshuggah or Car Bomb, I don't prefer it set all the way to Harman values.

I have different gear (Focal Clear and Audeze LCD2C), but I agree 100% with your approach. Having different cans for different uses and moods is the best type of endgame set up. I don't think any headphone could be just EQ'ed to the Harman response and be expected to sound it's best – the whole thing is super individual and no two ears are the same. playing with EQ is the final step in personalization, after choosing phones and electronics. and sometimes it's the lengthiest one (I'm still fiddling with my presets for the focals more than a year after I bought them).

Regarding the question in your title, I think the Harman target curve is terrible. Every headphone I tried to apply it to sounded unbearable with it. I don't understand why it's like that, but to me the target curve has insanely elevated treble response. So it's basically the opposite of what I usually do to my headphones with EQ, which is lowering treble and raising the bass. Only headphone that came with enough bass out of the box is the Meze 99 classics.

 

[solderdude]

Hopefully that hurdle can be eventually overcome as it would give us more accurate hearing measurements.

Currently I'm using decreasing major scale (C major to be specific) from 250Hz downwards to apply room correction via hearing. I managed to bring the variation down to 4dB and I could still hear imperfections but I got bored of mucking around with the parametric EQ. The hard part is finding the correct Q and the corresponding center frequency iteratively, made worse when real life peaks don't follow a smooth curve aren't even symmetric.

For that I would think a measurement mic might be handy instead of the ears.