Head-Fi.org › Forums › Equipment Forums › Sound Science › Equalizing headphones: What equipment do I need?
New Posts  All Forums:Forum Nav:

Equalizing headphones: What equipment do I need? - Page 3

post #31 of 52
Quote:
Originally Posted by CrystalT View Post

it is my opinion that you should never eq a headphone.

 

And why is that?

post #32 of 52
Quote:
Originally Posted by zspuckl View Post

 

And why is that?

That was a 2 week old post, he might not be around anymore.  Those that don't like EQ usually think of it a corrupting an otherwise pure signal.  In reality every piece of the chain from microphones through speaker and headphones corrupts the signal.  The goal of good EQ is to properly compensate for as much of that as possible.  

post #33 of 52
Quote:
Originally Posted by ComfyGrados View Post

I'm thinking that if I had the equipment, this is what I would do:

 

- Use measurement mic to EQ speaker setup to neutral;

- Place your head where the mic was while wearing in-ear microphones right next to your eardrums, measure the response of the speakers. This becomes your target EQ curve (you'd want to modify it to be perfectly flat from 0-1000 Hz)

- INVERT the target EQ curve, throw it into an EQ, measure the response of the headphones with the in-ear microphones.

- Chain two EQs; first is the inverse target EQ curve, then EQ the second one in the chain so it measures as completely neutral to the in-ear microphones.

- Remove the first EQ from the chain. Your signal should now be neutral according to your own head's HRTF... an approximation of your speaker reference to your head and only your head only.

 

Thoughts?

Flat speakers measured with an in-ear mic at the listening position as target EQ curve: yes, but what do you mean with the "(you'd want to modify it to be perfectly flat from 0-1000 Hz)" part?

 

Yeah, you can use two EQs. But you could also just compare the headphone FR to the target curve as you adjust the EQ.

 

 

Quote:

Originally Posted by jaddie View Post

 

You won't end up with your own HRTF this way, but it doesn't matter, you aren't dealing with HRTF in this case.

He kinda is dealing with his HRTF. To get proper binaural sound he'd just need to measure each ear separately, create two EQ curves per channel (one for each ear). Voila, you've got your own binaural impulse responses.

Or take the simpler route and use a crossfeed.


Edited by xnor - 12/19/12 at 5:39am
post #34 of 52
Quote:
Originally Posted by xnor View Post

Flat speakers measured with an in-ear mic at the listening position as target EQ curve: yes, but what do you mean with the "(you'd want to modify it to be perfectly flat from 0-1000 Hz)" part?

 

Yeah, you can use two EQs. But you could also just compare the headphone FR to the target curve as you adjust the EQ.

 

 

He kinda is dealing with his HRTF. To get proper binaural sound he'd just need to measure each ear separately, create two EQ curves per channel (one for each ear). Voila, you've got your own binaural impulse responses.

Or take the simpler route and use a crossfeed.

And the above is kind of my point.  Flat speakers...hmmm. Flat on axis, sure, but what about off axis? It's also generally recognized that absolutely flat doesn't sound right anyway, there's also a target curve at work.  And the listening position...in an unspecified acoustic space with uncontrolled reflections...now you've got the off axis speaker response reflecting of the surfaces. The reference was a measurement mic outside of the ear, which is omnidirectional, so the result includes at least some reflections integrated into the measurement, even if its a swept sine, because of the limitations of the window.  To do this right, you'd need to make the room a reference also, one that's highly controlled, properly treated, though probably fully anechoic is not necessary.  You'd also have to be very careful about speaker position.

 

Then there's the accuracy of the first EQ.  Just for practicality, there has to be at least some smoothing going on, so the EQ ends up being approximate.  So, the reference is highly approximate, and that's just the reference.

 

HRTF is by definition a three dimensional data set.  Even if you only do one vector, you have to pick which one. For a forward speaker, it might be zero degrees.  But headphones are positioned pretty much at 90 degrees for an on ear designs (IEMs bypass the pinna).  Measuring each ear separately further invalidates the original reference, because to get the full HRTF requires that the ears be in different physical positions so the time domain data will be there. The specific EQ curves for the separate ears won't be very different, though, as most ears are symmetrical.  My point is, a single vector response measurement may include the FR effect of the full ear, but since it's only one vector, hardly qualifies as HRTF.  But again, it doesn't matter in this case.  Full HRTF isn't important in profiling an equalizer for headphones. But a good reference is, and that's where I see the most problems in this.  The entire measurement process is not trivial.  It's a project for someone with some pretty significant resources.


Edited by jaddie - 12/19/12 at 9:22am
post #35 of 52
Quote:
Originally Posted by jaddie View Post

And the above is kind of my point.  Flat speakers...hmmm. Flat on axis, sure, but what about off axis? It's also generally recognized that absolutely flat doesn't sound right anyway, there's also a target curve at work.

Your whole post there was a lot of great reasons to not bother EQing. I've heard all of them a hundred times at HeadFi from other people who don't want to bother to EQ. The problem is, I've actually EQed my system and not much of what you say there really applies.

It's like every issue in audiophilia... The tendency is to make things so technical and so complicated with non real world absolutes that one just glazes over.

Just to address a few of your points...

Flat speaker systems do sound the best. If you like bass, it's got plenty. If you like treble, plenty of that too. The reason is, when you eliminate the problem of frequency masking, ALL frequencies sound better.

You don't have to achieve a flat response for every point in the room. You balance for the main central listening points and the rest of the room falls into line along the natural acoustics of the room. Just like if a band was playing in a room.

It isn't hard or difficult to equalize. Automatic EQ on receivers, published headphone response curves, etc are all good starting points. From there you just listen to a wide variety of recordings engineered to a flat response and make small incremental corrections.

Equalization can make a bigger improvement on your existing equipment than buying new equipment can. It's basically free. It just takes time and attention to detail.
Edited by bigshot - 12/19/12 at 10:09am
post #36 of 52
Quote:
Originally Posted by bigshot View Post

...Automatic EQ on receivers, published headphone response curves, etc are all good starting points. 

Equalization can make a bigger improvement on your existing equipment than buying new equipment can. It's basically free. It just takes time and attention to detail.

Automatic EQ on AVRs are great...and the target curves are not flat.  

 

+1 to the last line.

post #37 of 52
Quote:
Originally Posted by jaddie View Post

And the above is kind of my point.  Flat speakers...hmmm. Flat on axis, sure, but what about off axis? It's also generally recognized that absolutely flat doesn't sound right anyway, there's also a target curve at work.

Of course the speakers should point at the head (on-axis). The mic in the ear will measure anything but flat, that's the target curve, the HRTF. It's like a mix of the free/diffuse field EQ, but highly individual. That's the reason to do these measurements!

 

Quote:
And the listening position...in an unspecified acoustic space with uncontrolled reflections...now you've got the off axis speaker response reflecting of the surfaces. The reference was a measurement mic outside of the ear, which is omnidirectional, so the result includes at least some reflections integrated into the measurement, even if its a swept sine, because of the limitations of the window.  To do this right, you'd need to make the room a reference also, one that's highly controlled, properly treated, though probably fully anechoic is not necessary.  You'd also have to be very careful about speaker position.

I agree with reflections being a problem. Probably not a huge one, but it's definitely something to look into. You can always cut off reflections from the impulse response. You could even use an impulse instead of a sine sweep for the measurement. The interesting areas are upper mids to highs anyway.

 

Quote:
Then there's the accuracy of the first EQ.  Just for practicality, there has to be at least some smoothing going on, so the EQ ends up being approximate.  So, the reference is highly approximate, and that's just the reference.

Yes, some smoothing won't hurt.

 

Quote:

HRTF is by definition a three dimensional data set.  Even if you only do one vector, you have to pick which one. For a forward speaker, it might be zero degrees.  But headphones are positioned pretty much at 90 degrees for an on ear designs (IEMs bypass the pinna).  Measuring each ear separately further invalidates the original reference, because to get the full HRTF requires that the ears be in different physical positions so the time domain data will be there. The specific EQ curves for the separate ears won't be very different, though, as most ears are symmetrical.  My point is, a single vector response measurement may include the FR effect of the full ear, but since it's only one vector, hardly qualifies as HRTF.  But again, it doesn't matter in this case.  Full HRTF isn't important in profiling an equalizer for headphones. But a good reference is, and that's where I see the most problems in this.  The entire measurement process is not trivial.  It's a project for someone with some pretty significant resources.

There's no such thing as "full HRTF". A HRTF is defined using a single point in space. You only need the +-30° HRTF, since that's how stereo speakers are set up (assuming you do not move your head, otherwise you need something like the Smyth Realiser). The concerns about headphone positioning and bypassing the pinna are non-issues - that's the reason to do these measurements and equalization in the first place.

 

I don't think you need significant resources at all. See this.


Edited by xnor - 12/19/12 at 10:12am
post #38 of 52
Quote:
Originally Posted by xnor View Post

 

There's no such thing as "full HRTF". A HRTF is defined using a single point in space. You only need the +-30° HRTF, since that's how stereo speakers are set up (assuming you do not move your head, otherwise you need something like the Smyth Realiser). The concerns about headphone positioning and bypassing the pinna are non-issues - that's the reason to do these measurements and equalization in the first place.

If an in-ear response measurement is taken with a source at 30 degrees, then the headphone is placed on-ear at 90 degrees but with the eq curve of 30 degrees, my guess is there will be a confusion be a disparity between the localization cue derived from the curve and the cues derived from the inter-aural time delay, or rather lack of it, from the headphones.  Perhaps its not a problem, or at least a minor one.  It also seems that a single point HRTF would result in a target curve that's right for that vector only, and wouldn't account for other positions in the stereo field. Couldn't find this quickly, but I know I've seen essentially polar plots of ears, and perhaps it's a flaw in memory, but I recall there is a significant difference between between response at zero, 30, and 90 degree angles.  Probably in a paper about localization.

 

I'm not convinced that the pinna isn't a factor in on-ear headphone response, but it's clearly bypassed with IEMs, so that would mean different target curves for IEM vs on-ear. Mostly at the upper end of the spectrum, of course.  No sound in real life originates from within the ear canal, so the results of that transducer position would have to be accounted for.

post #39 of 52
You guys make it so complicated. It really isn't that hard.
post #40 of 52
Quote:
Originally Posted by jaddie View Post

If an in-ear response measurement is taken with a source at 30 degrees, then the headphone is placed on-ear at 90 degrees but with the eq curve of 30 degrees, my guess is there will be a confusion be a disparity between the localization cue derived from the curve and the cues derived from the inter-aural time delay, or rather lack of it, from the headphones.  Perhaps its not a problem, or at least a minor one.

This is just about frequency response. The 30° measurement is the target curve. The headphone is measured and equalized to match that response (the way your ear "hears" a flat speaker).

Regarding localization, as I said before, you can go crazy with binaural impulse responses or just take a crossfeed.

 

 

Quote:
It also seems that a single point HRTF would result in a target curve that's right for that vector only, and wouldn't account for other positions in the stereo field. Couldn't find this quickly, but I know I've seen essentially polar plots of ears, and perhaps it's a flaw in memory, but I recall there is a significant difference between between response at zero, 30, and 90 degree angles.  Probably in a paper about localization.

Again, you measure the speakers AND the headphones and equalize the frequency response accordingly.

 

Different positions in the stereo field are simulated by interaural time and level differences (see localization above), but the sound from the left and right speaker arrives from the same angle(s).

 

 

Quote:
I'm not convinced that the pinna isn't a factor in on-ear headphone response, but it's clearly bypassed with IEMs, so that would mean different target curves for IEM vs on-ear. Mostly at the upper end of the spectrum, of course.  No sound in real life originates from within the ear canal, so the results of that transducer position would have to be accounted for.

The target curve is always the same. What changes is the measurement of the head/earphone and therefore EQ curve. For IEMs you'd need a tiny in-ear mic though.

 

 

 

Quote:
Originally Posted by bigshot View Post

You guys make it so complicated. It really isn't that hard.

The process described by ComfyGrados is probably more elaborate than just equalizing by hand and ear, but it's not complicated. I'm just trying to clear up some misunderstandings.


Edited by xnor - 12/19/12 at 12:08pm
post #41 of 52
Adjusting the tone on your stereo shouldn't be as complicated as putting a space station into orbit. Practical advice based on what you've done yourself with your own system is always more helpful than theory. Did you really go through all these steps yourself?
post #42 of 52
Quote:
Originally Posted by bigshot View Post

Adjusting the tone on your stereo shouldn't be as complicated as putting a space station into orbit. Practical advice based on what you've done yourself with your own system is always more helpful than theory. Did you really go through all these steps yourself?

It's all a matter of what you're happy with.  For some, adjusting tone controls is just fine.  Others want more precision, hence the complexity.  So far this part of the thread has all be theoretical.  Nobody in the thread has claimed to have actually done any of this.  I believe I made mention of the fact that it has been done, though, and more on that later.  We live in a world where the tools for achieving precision are far more available than ever before.  So what we're talking about in essence is the feasibility of getting some actual measurements from which to develop precision EQ curves.  Can't see anything wrong with that.

post #43 of 52
Quote:
Originally Posted by xnor View Post

 

The target curve is always the same. What changes is the measurement of the head/earphone and therefore EQ curve. For IEMs you'd need a tiny in-ear mic though.

 

As a practical matter, it's doubtful that any of us will come up with a way to place a tiny measurement mic next to our eardrum, stuff IEMs in our ears and get them to seal around the mic wires, and make any useful measurements.  Hence, the different target for IEM, which would basically artificially add the effect of the pinna and outer part of the ear canal.  I'm not saying its big, but if you're going to do it, it's a factor.

 

We can go around about the angle influencing the target all day, in the end, whatever is done has to be a step closer to getting headphone EQ right.

post #44 of 52
Quote:
Originally Posted by jaddie View Post

As a practical matter, it's doubtful that any of us will come up with a way to place a tiny measurement mic next to our eardrum, stuff IEMs in our ears and get them to seal around the mic wires, and make any useful measurements.  Hence, the different target for IEM, which would basically artificially add the effect of the pinna and outer part of the ear canal.  I'm not saying its big, but if you're going to do it, it's a factor.

 

We can go around about the angle influencing the target all day, in the end, whatever is done has to be a step closer to getting headphone EQ right.

Well I agree that doing this with IEMs is near impossible without professional help. Now I also get what you mean with different targets. Ear canal entrance measurements for over-ear headphones and maybe on-ear ones vs. near eardrum measurements for IEMs.

 

Of course you can go the much more practical route: fire up your favorite parametric EQ and compare the IEM's sound to that of your calibrated speakers. That's what I did with my headphones (on- & over-ear) too.

post #45 of 52
Quote:
Originally Posted by jaddie View Post

So far this part of the thread has all be theoretical.  Nobody in the thread has claimed to have actually done any of this.

I've done what I'm talking about and it has made a huge improvement in the sound of my system. The proper use of DSPs makes a big difference too. If someone here is interested in trading tips on how to EQ their speakers, I'm all in. But I really don't have anything to add to armchair theories.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Equalizing headphones: What equipment do I need?