The most reliable/easiest way to EQ headphones properly to achieve the most ideal sound (for non-professionals)
Feb 15, 2022 at 5:31 PM Post #301 of 316
We've created a new logical fallacy, "reducto ad argumentum". I'm not going to dive into the back and forth because it doesn't look profitable, but I will toss in a question...

See, that's what I don't get. IMO that line is not how speakers are "calibrated", but how they said their averaged group preferred to hear them. To me that's not only unbelievable, but when I applied those correction to my headphones, it sounded bad.

Did you subtract the response curve of your own headphones before you added the Harman curve, or did you just add the Harman curve without any correction?

Headphones don't generally have a flat response. Every manufacturer has their own target curve that they apply to their cans. If you add the Harman curve on top of that, you'll double your response curves and come up with weird results.

Likewise, not every measurement of headphones published on the internet is the same. Some are compensated as Castle points out in 288 and some aren't. So it's possible that the subtraction you need to do before calibrating to Harman might be off.

The third option is just that your personal physiognomy is an outlier from the Harman Curve. Not everyone thought the Harman target sounded best, just most of them did. You could be among the rest of the sample that didn't like the Harman curve.

Personally, I think Harman is a good place to start. It works well as a baseline. Then EQ in small corrections at a time over a long period of time to try to improve it for your own ears. There is no "one size fits all" response curve with headphones like there is for speakers. There's just a "one size fits most". You have to experiment and learn what works best for you.
 
Feb 16, 2022 at 1:54 PM Post #303 of 316
I have no idea what you’re talking about. Harman is just a response curve. It doesn’t involve dynamic compression nor does it introduce compression artifacts. I think you’re just expressing bias here, not information.

This isn’t a football game where we cheer on a team. The goal here is to optimize sound and Harman is one tool for doing that. It’s effective for most people.
 
Last edited:
Feb 16, 2022 at 3:53 PM Post #304 of 316
Maybe my allegory was lost to you. OK, my bad.
Again, you don't address the actual complaint - why are the mids raised and treble squashed? There is not reason why a studio mastered program to be destroyed like that.
 
Feb 16, 2022 at 3:58 PM Post #305 of 316
You can calibrate speakers in a room and it will sound correct for everyone, but headphones are different. The sound doesn't inhabit a space. It's directed straight into your ear canal. Harman may sound like that to you, but the balance may be right for someone who has a different shaped head and inner ear. Harman isn't "bad". In fact it sounds good to most people. But for about a third of people, it isn't right. You fit in that category it seems.
 
Feb 16, 2022 at 4:29 PM Post #306 of 316
Over the ear headphones always sounded like the speakers to me. This is not about IEM coupling.
And even the Harman curves for speakers confirm that - they have the same profile.

This argument that we need to raise the mids +10dB and lower the treble 10dB because "ears are different" is not right.

Harman curve it's just a preference, and a bad one for that. That's because it implies that the mastering engineers make unpleasant sounding records...
 
Feb 16, 2022 at 4:39 PM Post #307 of 316
I don’t think you understand what I’m saying.
 
Last edited:
Feb 16, 2022 at 8:39 PM Post #308 of 316
Maybe my allegory was lost to you. OK, my bad.
Again, you don't address the actual complaint - why are the mids raised and treble squashed? There is not reason why a studio mastered program to be destroyed like that.
giphy.gif

Those graphs are like a language you don't know how to read. At this point I'm convinced it has absolutely nothing to do with Harman, or EQ. You just have no idea how headphones are measured and decided to misinterpret the hell out of some RAW graphs.

Did you never see RAW measurements of headphones you owned?
 
Feb 16, 2022 at 8:50 PM Post #310 of 316
You would understand better if you listened better. Did you follow the instructions in the first post in this thread when you calibrated to the Harman curve? I'm sensing that you don't have the knowledge to be able to actually do that. I think you applied the Harman curve on top of your headphones' built in manufacturer curve. You essentially doubled the Harman bump in the midrange because your cans were already corrected for that. And that huge boost in the upper mids masked all the treble. (I know you won't understand what I'm saying here, but other people might.)

And of course as Castle has been trying to explain to you... Flat response is not reflected on raw measurements as a flat line.
 
Last edited:
Feb 17, 2022 at 4:28 AM Post #311 of 316
I guess some people like to hear only part of the music... oh, well. More power to them!
You mean more power to you. You're the one who seems to have a serious hearing problem because:
Over the ear headphones always sounded like the speakers to me.
Well they don't to the vast majority of people. Humans do not have acoustically transparent skulls, are almost always able to easily differentiate high levels of room reverberation from no reverberation and can easily tell the difference between headphones and speakers.
This argument that we need to raise the mids +10dB and lower the treble 10dB because "ears are different" is not right.
What argument that we need to raise the mids by 10dB and lower the treble by 10dB? No one apart from you has suggested that!
Please show me the RAW graphs that have a -10dB down dip at mids that needs to be corrected by that +10dB hump.
What correction with +10dB hump? You don't seem able to grasp the simple (and obvious) fact that the Harman Target Curve is a TARGET curve, it is NOT a correction curve! As an example, here is the raw graph of the AKG 701 without any EQ correction, which you've stated you own:
1645086874811.jpeg

In this example, you would need to apply a dip to the mids (very roughly about -5dB) in order to match the Harman Target Curve, NOT a +10dB hump. A +10dB hump would take you even further away from the target!
Again, you don't address the actual complaint - why are the mids raised and treble squashed?
Your "complaint" has been addressed numerous times - because that is what room acoustics, the human torso, skull, pinnae and ear canal do!
There is not reason why a studio mastered program to be destroyed like that.
If you don't have a human torso, skull, pinnae and ear canals then it wouldn't be destroyed like that (with speakers) or need to be destroyed like that (with HPs)! But then if you don't have a torso or skull, how are you going to wear your HPs?

G
 
Mar 6, 2022 at 3:48 PM Post #312 of 316
Okay it's easy use a sonynw-zx500 to break it into SQ. This gives it the reprucussionable acoustic properties needed for a driver to go loud once it's ready, you won't need an amplifier. I recommended a SONY DAP. The best current one? Probably the SONYNWZX507.

NOW - onto your second dilemma, the 2.2khz curve is for audio voice. SO - lower that all the way down. And level everything else appropriately; starting with 10,000Hs+ so that the highs are clean without white noise being noticeable and annoying.

NEXT control midbass - then bass. Then guitar, then dial in that chorus if you can't hear any sound on the audio regarding voices. REMEMBER step 1.

After that.... just tune to the instruments your trying to isolate and promote without over extending / under surging them.
 
Mar 11, 2022 at 2:13 AM Post #313 of 316
Okay it's easy use a sonynw-zx500 to break it into SQ. This gives it the reprucussionable acoustic properties needed for a driver to go loud once it's ready, you won't need an amplifier. ….

Sorry, I have no idea what you are trying to say. It almost looks like some translation software has messed your post up.

G
 
Jul 12, 2022 at 10:34 PM Post #314 of 316
If it could help someone else, I just want to recommend software I have found helpful in equalizing my sets to neutral.

I have been using the NCH Tone Generator to great success (some kind of freemium, Windows, Mac, iOS, Android), or as a completely free alternative the Tolvan Data Tone Generator (Windows).

(Edit: I now only use the Tolvan Data Tone Generator. Its coarse and fine sliders are very useful, as is the quick entering of frequencies in the box below the sliders.)

The NCH Tone generator can output grey/pink/white noise which you can use to first set your volume to your comfortable/max listening SPL. Then you should proceed with its variety of options of generating a sine wave and changing its frequency, the following being the ones I use alternately and with a variety of settings:
  • log sweep - can set start/end frequencies and sweep length in ms (e.g. sweep slowly or quickly, from 20hz-20khz, or from 20hz-1khz, or problem areas, etc.)
  • continuous tone - can use page up/down or numpad +/- to step in intervals of semitones, or by any amount of hz you set (e.g. reducing the hz interval when deciding where to center a cut)
  • you can even play multiple sines of different hz simultaneously (which I have only tried briefly, could be useful to set the balance equally between the different freqs)
I do this on top of a base auto EQ to get me quickly in the ballpark, from 20hz-5khz to a Harman-ish target (e.g. I use an equal loudness compensated Harman target to 65 SPL). My individual resonance peak lies above 5khz around 7khz, so that is where I limit the auto EQ. You can auto EQ higher; you can use any Harman-ish target; this step is flexible because in the end you adjust the frequency balance to what sounds flat to you (or your preference).

Don't feel pressured to neutralize your set in one sitting. Like a rolling pin over dough, gradually over time, with every nudge (and overlaid band if your EQ can afford it) of EQ you can turn the FR into whatever you like.
 
Last edited:
Jul 14, 2022 at 1:28 AM Post #315 of 316
If it could help someone else, I just want to recommend software I have found helpful in equalizing my sets to neutral.

I have been using the NCH Tone Generator to great success (some kind of freemium, Windows, Mac, iOS, Android), or as a completely free alternative the Tolvan Data Tone Generator (Windows).

The NCH Tone generator can output grey/pink/white noise which you can use to first set your volume to your comfortable/max listening SPL. Then you should proceed with its variety of options of generating a sine wave and changing its frequency, the following being the ones I use alternately and with a variety of settings:
  • log sweep - can set start/end frequencies and sweep length in ms (e.g. sweep slowly or quickly, from 20hz-20khz, or from 20hz-1khz, or problem areas, etc.)
  • continuous tone - can use page up/down or numpad +/- to step in intervals of semitones, or by any amount of hz you set (e.g. reducing the hz interval when deciding where to center a cut)
  • you can even play multiple sines of different hz simultaneously (which I have only tried briefly, could be useful to set the balance equally between the different freqs)
I do this on top of a base auto EQ to get me quickly in the ballpark, from 20hz-5khz to a Harman-ish target (e.g. I use an equal loudness compensated Harman target to 65 SPL). My individual resonance peak lies above 5khz around 7khz, so that is where I limit the auto EQ. You can auto EQ higher; you can use any Harman-ish target; this step is flexible because in the end you adjust the frequency balance to what sounds flat to you (or your preference).

Don't feel pressured to neutralize your set in one sitting. Like a rolling pin over dough, gradually over time, with every nudge (and overlaid band if your EQ can afford it) of EQ you can turn the FR into whatever you like.
That's pretty interesting, I'll have to try that generator out.
I've been using SineGen for a few years now for sine sweeps, in order to pinpoint peaks on my headphones and IEMs. https://sinegen.en.lo4d.com/windows
I basically do it all manually, sweep throughout the whole audible frequency range and check for any obvious elevations or dips. I usually ignore the dips unless they are very severe.
This is especially useful for fixing the timbre on certain headphones such as the Focal Clear, which has narrow peaks in the 6kHz and 10.5 kHz range causing that "metallic" timbre. I know using narrow Q value filters is sometimes frowned upon for whatever reason (I don't really know the scientific explanation why, something to do with affecting the phase which I just can't hear), but it works very well for me. In particular, I have the Dunu Zen Pro which has a few sharp narrow peaks from 7-12 kHz, most are about 6 dB higher than the rest of the frequency range. Using narrow EQ filters smooths this IEM out very nicely, allowing the excellent technical performance of the Dunu Eclipse style driver to shine through.
 

Users who are viewing this thread

Back
Top