Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Headphone calibration (Read 11347 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Headphone calibration

I hope user saratoga reads and answers this thread. In multiple threads, he has said that by calibrating cheaper headphones they can be matched to a higher quality pair. My question is, how is it done? Is it even possible without actually having accurate measurements of frequency response and then using sofware to derive the curve? Actually, I care more about matching diffuse field equalization or the newer curve developed by the Harman researchers but the question is the same.

Headphone calibration

Reply #1 –
Quote from: embassyhill on
I hope user saratoga reads and answers this thread. In multiple threads, he has said that by calibrating cheaper headphones they can be matched to a higher quality pair. My question is, how is it done? Is it even possible without actually having accurate measurements of frequency response and then using sofware to derive the curve? Actually, I care more about matching diffuse field equalization or the newer curve developed by the Harman researchers but the question is the same.



It seems logical that if a listener can tell that the FR of something is wrong by just listening, they can use proper experimental techniques to make it right with an equalizer by just listening.

The first time they try may be difficult and frustrating, but IME it is a skill that can be learned.

The most common post I see about equalization runs something like: "The first time I tried to equalize my (headphones) (speakers) it did not work perfectly so I quit trying and never went down that road again".

One reasonable way to learn some of the needed skills for equalizing headphones may be to equalize one's speakers/room because you can easily do measurements of the results of that.

Headphone calibration

Reply #2 –
Quote from: embassyhill on
I hope user saratoga reads and answers this thread. In multiple threads, he has said that by calibrating cheaper headphones they can be matched to a higher quality pair.


Good EQ can certainly help improve sound quality.  But in the most recent thread I was asking what people thought of calibration vs. uncalibrated.

Quote from: embassyhill on
My question is, how is it done? Is it even possible without actually having accurate measurements of frequency response and then using sofware to derive the curve? Actually, I care more about matching diffuse field equalization or the newer curve developed by the Harman researchers but the question is the same.


I do something very simple: replaygain a series of tones and then adjust the EQ until they sound approximately correct to my ear.  The process is very subjective, but I found it helps quite a lot.  Other people may have better methods, and of course if you are using IEMs, you could probably build a simulator or some other instrument to do it quantitatively, but thats a bit more work. 

Headphone calibration

Reply #3 –
Thanks for the answer. But now I of course have to ask: what tones? Is a sine generator OK? Also, what do you mean when you use replaygain as a verb? You make the tones have equal loudness?

Headphone calibration

Reply #4 –
Quote
Thanks for the answer. But now I of course have to ask: what tones? Is a sine generator OK?
The only pure single-frequency tone is a sine wave.  If you want to listen to, or test, one frequency at at time, that's your only choice.  You can generate sine wave files with an audio editor such as Audacity.

Quote
Also, what do you mean when you use replaygain as a verb? You make the tones have equal loudness?
I assume he means applying ReplayGain to each tone-file separately to give them all equal loudness.

Headphone calibration

Reply #5 –
Thank you. I already tried sine waves generated with sox but I didn't use ReplayGain. I'll experiment with your method now.

I have an off-topic question. I wonder what headphones/earphones do engineers and audio programmers (like some of you) use?

Headphone calibration

Reply #6 –
Quote from: embassyhill on
Thank you. I already tried sine waves generated with sox but I didn't use ReplayGain. I'll experiment with your method now.

I have an off-topic question. I wonder what headphones/earphones do engineers and audio programmers (like some of you) use?


Not off-topic I think.

I adjust the equalizer until I like how things sound based on listening to a bunch of different recordings.

Headphone calibration

Reply #7 –
Quote from: embassyhill on
Thanks for the answer. But now I of course have to ask: what tones? Is a sine generator OK? Also, what do you mean when you use replaygain as a verb? You make the tones have equal loudness?



I use the one in foobar to make flac files, and then foobar's replaygain to equalize the loudness.  I actually don't know if this is ideal for pure tones, but in practice it seemed to give me a relatively neutral sounding EQ curve, although I did manually tweak it to my personal taste.

Headphone calibration

Reply #8 –
Quote
I have an off-topic question. I wonder what headphones/earphones do engineers and audio programmers (like some of you) use?
I rarely use EQ, and I rarely listen to headphones...  The subwoofer on my home theater setup is adjusted by ear with regular program material to what sounds good to me.    Otherwise, I'm just getting whatever my speakers (or headphones) are giving me, and it sounds good enough to me.

I seem to usually have the bass in my Honda turned-up to +2 most of the time (whatever +2 means), and depending on what I'm listening to I may give the treble a little boost.  The bass from the factory car stereo is a little wimpy but it gets distorted and/or something vibrates if I turn it up any more.

I've got a van with a tri-amped system and the only EQ is the 3 frequency-band adjustments.  Again just done by ear to what sounds good to me.

Headphone calibration

Reply #9 –
Quote
Thank you. I already tried sine waves generated with sox but I didn't use ReplayGain
SOX is fine.  But since our ears don't have flat response, you can't use equal amplitude tones to check frequency response by ear.    The idea of ReplayGain would be to get equal perceived volume at different frequencies.

I imagine this takes some listening-practice....  I'll bet if you asked 10 different people to set tones at 100Hz, 1kHz, and 10kHz to equal loudness, you'd get 10 different results and differences of more than 6dB.  And of course, different people do have different hearing so not everyone's hearing will match the Fletcher-Munson curves.

I'd say the ear is a poor measurement device and it's especially poor for measuring frequency response.  But, there's nothing wrong with listening to music and adjusting the EQ 'till it sounds good to you.  That's the way music production is done...  The mixing or mastering engineer makes adjustments by ear (with good monitors in a good room).  Mixing/mastering engineers usually have one or two known-good reference recordings that they are very familiar with to "keep their ears calibrated", but it's all about how it sounds to the (trained & experienced) human ear.

Headphone calibration

Reply #10 –
Based on the facts stated here about EQ and all, is it justifiable to buy an IEM let's say worth greater $300?

Headphone calibration

Reply #11 –
Quote from: SlickSteiner on
Based on the facts stated here about EQ and all, is it justifiable to buy an IEM let's say worth greater $300?


I don't think anyone said that all IEMs can be calibrated to be the same? Cheaper IEMs usually have a single driver, more expensive ones have more drivers and crossovers (?) so I wouldn't expect that they could all be calibrated to sound the same.

Headphone calibration

Reply #12 –
Quote from: probedb on
Quote from: SlickSteiner on
Based on the facts stated here about EQ and all, is it justifiable to buy an IEM let's say worth greater $300?


I don't think anyone said that all IEMs can be calibrated to be the same? Cheaper IEMs usually have a single driver, more expensive ones have more drivers and crossovers (?) so I wouldn't expect that they could all be calibrated to sound the same.


Interesting that you mention multi-driver headphones. Besides his well known loudspeaker exploits for a number of years Dr. Earl Geddes was the leading technical manager for Knowles Research, one of the largest and best known manufacturers of high performance earphone components in the world. 

If memory serves Earl has been pretty dogmatic in my presence about the lack of technical justification for multiple drivers in earphones. I've never seen him using any other kind of earphone than the single driver balanced armature kind.

I wonder how many multiple driver earphones are used in hearing aids?  I'm not really familiar with this market, but my online searching has so far come up empty.

Headphone calibration

Reply #14 –
Quote from: Arnold B. Krueger on
I wonder how many multiple driver earphones are used in hearing aids?  I'm not really familiar with this market, but my online searching has so far come up empty.


Hearing aids are made to enable people to hear speech, motor horns, and the like. The frequency range is limited.
The most important audio cables are the ones in the brain

Headphone calibration

Reply #15 –
Quote from: Arnold B. Krueger on
Interesting that you mention multi-driver headphones. Besides his well known loudspeaker exploits for a number of years Dr. Earl Geddes was the leading technical manager for Knowles Research, one of the largest and best known manufacturers of high performance earphone components in the world. 

If memory serves Earl has been pretty dogmatic in my presence about the lack of technical justification for multiple drivers in earphones. I've never seen him using any other kind of earphone than the single driver balanced armature kind.

I wonder how many multiple driver earphones are used in hearing aids?  I'm not really familiar with this market, but my online searching has so far come up empty.


I don't believe I was making any claims about the reasons to use them or whether they are better or not, just that the more expensive ones tend to use multiple drivers. Nor did I mention hearing aids.

Quote from: saratoga link=msg=0 date=
In terms of frequency response they can.


Even if one set of IEMs is incapable of reproducing a certain frequency band/range accurately or even at all or with the required levels as the rest of the range?

Headphone calibration

Reply #16 –
Which commercial iems are incapable of producing audible frequencies? Which frequencies?

Headphone calibration

Reply #17 –
Quote
Based on the facts stated here about EQ and all, is it justifiable to buy an IEM let's say worth greater $300?
In terms of "sound quality", $300 is probably too high...  Just my opinion, and I don't own any IEMs (other than the ear buds that came with my iPod that Iv'e never used)

I actually don't know what custom ear molds cost.    A good pair of IEMs that are good sounding, rugged & reliable, good looking, with a good warranty, and custom ear molds might cost $300 or more.

But, does Bill Gates need to "justify" spending more than that for listening to his iPod...  Errr... for listening to his Zune.    Or does Mick Jagger or Lady Gaga have to "justify" using expensive IEMs on stage?

Quote
Hearing aids are made to enable people to hear speech, motor horns, and the like. The frequency range is limited.
Modern hearing aids are custom "equalized" to (at least partially) compensate for high-frequency loss which is usually worse than mid-frequency loss.  An audiogram shows your ears "frequency response" and the audiologist will adjust your hearing aids appropriately.

Many modern digital hearing aids have a "music mode", "loud environment mode", as well as the normal mode, and sometimes more options.  Once you go digital, there's a lot you can do!

Quote
If memory serves Earl has been pretty dogmatic in my presence about the lack of technical justification for multiple drivers in earphones. I've never seen him using any other kind of earphone than the single driver balanced armature kind.
Interesting!  I understand the physics behind multi-driver speakers, but I've been skeptical about multi-driver IEMs.  But, reputable manufacturers do make them, so I thought maybe there's reason (other than marketing). 

And I'm not aware of any 2-way or 3-way headphones, although I think I've seen some advertized in the past.  And again, my "physics intuition" is telling me it shouldn't be necessary.

Headphone calibration

Reply #18 –
Quote from: saratoga on
Which commercial iems are incapable of producing audible frequencies? Which frequencies?


I said "what if" didn't I?

So you're now saying all IEMs can produce all audible frequencies equally well?

Mine have an upper limit of 18KHz.

Is it me or is this forum just descending into arguments in every thread rather than discussions?

Headphone calibration

Reply #19 –
Quote from: probedb on
Quote from: saratoga on
Which commercial iems are incapable of producing audible frequencies? Which frequencies?


I said "what if" didn't I?

So you're now saying all IEMs can produce all audible frequencies equally well?


It depends on what you think "equally well" means.  As we were discussing, you can equalize frequency response, but obviously you can't do anything for nonlinear effects. 

Quote from: probedb on
Mine have an upper limit of 18KHz.


I doubt this is true.  How did you measure it?

Headphone calibration

Reply #20 –
Some in-ears have serious roll-off somewhere between 10 and 18 kHz. You'd have to look on a case by case basis. InnerFidelity and HeadRoom have many measurements.
"I hear it when I see it."

Headphone calibration

Reply #21 –
Quote from: saratoga on
It depends on what you think "equally well" means.  As we were discussing, you can equalize frequency response, but obviously you can't do anything for nonlinear effects. 

I doubt this is true.  How did you measure it?


You'll moan at me for quoting spec sheets now, but that's what it says. Now we can't even quote specs unless we have all the required measurement equipment to do it ourselves? At any rate, I thought all manufacturers overstated their specs so surely if they want to sell they'd have put 20KHz or higher? 18KHz wouldn't appeal to an "audiophile" due to their ability to hear angelic harps in heaven so no sale there.

There's various charts around the net... http://cdn.head-fi.org/6/6a/6af03198_02.FR...3X_Etymotic.png

Headphone calibration

Reply #22 –
Quote from: Thad E Ginathom on
Quote from: Arnold B. Krueger on
I wonder how many multiple driver earphones are used in hearing aids?  I'm not really familiar with this market, but my online searching has so far come up empty.


Hearing aids are made to enable people to hear speech, motor horns, and the like. The frequency range is limited.


True - according to a review of some published professional literature in the field the standard operational frequency range of hearing aids is 250 Hz to 4 KHz.

However, looking at published FR tests of single driver headphones, I don't see that people are having trouble getting response way outside of that range.

The real question is whether the two drivers are actually any kind of requirement for true hifi performance. BTW how many 2-driver headphones are on the market? I think the answer is close to zero.

Headphone calibration

Reply #23 –
Balanced armature? There are a couple of them. Also 3 drivers and more ...
"I hear it when I see it."

Headphone calibration

Reply #24 –
Quote from: saratoga on
...you can equalize frequency response, but obviously you can't do anything for nonlinear effects.


Obvious that you can't do anything about nonlinear distortion?

Most nonlinear distortion in audio gear is already addressed via inverse feedback (even sometimes with speakers), but when that is inconvenient nonlinear feedforward nonlinear compensation will work.