Audeze LCD-X Review (2021 Edition Headphone)

[Ken Tajalli]

It's not madness, the Oratory EQ's are not madness. Using a frequency response measurement to derive an exact EQ is as good as the measurement is, factors like unit to unit variation coming into play and some on-head variation, but it's not madness to do an exact EQ from a measurement. Often it's a great starting point, but it can be an end point too, depends how lucky you get and if you like the Harman Curve. Generally those EQ's will improve a headphone for most people, it's a result of the research. That doesn't mean you have to like the Harman Curve though, perhaps you gain enough EQ experience with different headphones to know what kind of a target curve you prefer - it might be a slight or somewhat significant variation on Harman, or it might just be Harman, then you can use measurements to do exact EQ to your favourite target curve and then that would be your best starting point or even an end point.

We agree to disagree.
I can not change the EQ-loving people, they can not convince others either.
But to claim, that it is a Science ! that is rich.
You can measure the very same headphone on different days and get significantly different results. As you mentioned, there is also unit to unit deviations. And there is the Harman curve that was arrived at as a favourite among a select group. On top of that there is individuals hearing loss or preferences.
All of that, throws any exact science out of the window.
I accept the FR curves can be guideline or a starting point, but seriously, EXACTLY?

 

[Bernard23]

We agree to disagree.
I can not change the EQ-loving people, they can not convince others either.
But to claim, that it is a Science ! that is rich.
You can measure the very same headphone on different days and get significantly different results. As you mentioned, there is also unit to unit deviations. And there is the Harman curve that was arrived at as a favourite among a select group. On top of that there is individuals hearing loss or preferences.
All of that, throws any exact science out of the window.
I accept the FR curves can be guideline or a starting point, but seriously, EXACTLY?

You can either understand the scientific process, or not. If you have read the paper written by Sean Olive then you might have a much better understanding of Harman targets, and how they were derived.

 

[Bernard23]

It's not madness, the Oratory EQ's are not madness. Using a frequency response measurement to derive an exact EQ is as good as the measurement is, factors like unit to unit variation coming into play and some on-head variation, but it's not madness to do an exact EQ from a measurement. Often it's a great starting point, but it can be an end point too, depends how lucky you get and if you like the Harman Curve. Generally those EQ's will improve a headphone for most people, it's a result of the research. That doesn't mean you have to like the Harman Curve though, perhaps you gain enough EQ experience with different headphones to know what kind of a target curve you prefer - it might be a slight or somewhat significant variation on Harman, or it might just be Harman, then you can use measurements to do exact EQ to your favourite target curve and then that would be your best starting point or even an end point.

The Harman curve has to be used with the understanding of HRTF, which is the same as room correction for speakers in a room. Each user application is different so there is no magic bullet (though RC is easier in that respect)

 

[Ken Tajalli]

You can either understand the scientific process, or not. If you have read the paper written by Sean Olive then you might have a much better understanding of Harman targets, and how they were derived.

So if I disagree it is because I am ignorant? there could be nothing else?

 

[Bernard23]

So if I disagree it is because I am ignorant? there could be nothing else?

That's a possibility, yes. That's how it works unfortunately.

 

[Ken Tajalli]

 

[Robbo99999]

We agree to disagree.
I can not change the EQ-loving people, they can not convince others either.
But to claim, that it is a Science ! that is rich.
You can measure the very same headphone on different days and get significantly different results. As you mentioned, there is also unit to unit deviations. And there is the Harman curve that was arrived at as a favourite among a select group. On top of that there is individuals hearing loss or preferences.
All of that, throws any exact science out of the window.
I accept the FR curves can be guideline or a starting point, but seriously, EXACTLY?

Even though we accept that there unknowns in the system like unit to unit variation & on head variation then it still doesn't mean that it's not worth EQ'ing accurately to your best starting point from the measurement because that's the best information you've got at the time, so as a starting point there would be no point in randomly deviating your EQ from the best information available by not EQ'ing accurately to it. If you do EQ accurately to it, then that's either your starting point or if it sounds good then it's your end point. I don't see any merit in not EQ'ing accurately to a measurement in the first place. The level of precision that Oratory uses to EQ to the target curve is sensible, he doesn't EQ up really narrow sharp dips and he uses "only" 10 bands of EQ. His initial measurement is also smoothed in some respect inasmuch that it's an average of multiple measurements per unit & also of multiple units of headphones, so that is already taking out what would otherwise be unnecessary or detrimental EQ.

(Just to one of your other points you mention in there, my headphones don't measure differently on different days when I measure them on my miniDSP EARS - if you've got an accurate measurement protocol then the only difference will be pad wear.)

View attachment 390676

This is already taken into account in the process I described.

 

[Bernard23]

View attachment 390676

If you take the time to read Olives paper, you'd reach the same conclusion.
Science is a process, not an outcome. The outcomes of science are rarely black or white, there is inherent variation in almost everything we measure. Understanding that uncertainty is the key to engineering things that work. All that Olive does is calculate the uncertainties around HRTF and develop some average targets, which will work perfectly for a proportion of people, but not everyone because of the natural variation in the shapes of our ears and heads. As, far as I know, there isn't one recognised standard measurement protocol based on a standard head RQI for example. I could be wrong, but since different instruments give very different results then this is likely.
Now of course if your own HRTF lies at the edge of this distribution then a Harman curve might sound very odd, hence most people use it as a starting point and then modify to taste. Or just leave it to the whims of the engineers who designed and built them and hope you find one that works for you.

 

[Ken Tajalli]

Even though we accept that there unknowns in the system like unit to unit variation & on head variation then it still doesn't mean that it's not worth EQ'ing accurately to your best starting point from the measurement because that's the best information you've got at the time, so as a starting point there would be no point in randomly deviating your EQ from the best information available by not EQ'ing accurately to it. If you do EQ accurately to it, then that's either your starting point or if it sounds good then it's your end point. I don't see any merit in not EQ'ing accurately to a measurement in the first place. The level of precision that Oratory uses to EQ to the target curve is sensible, he doesn't EQ up really narrow sharp dips and he uses "only" 10 bands of EQ. His initial measurement is also smoothed in some respect inasmuch that it's an average of multiple measurements per unit & also of multiple units of headphones, so that is already taking out what would otherwise be unnecessary or detrimental EQ.

I had already said:

My beef has never been to dismiss FR measurements, my only beef has been with the notion of "EQ'ing accurately" or "exact EQing".
To use the available curves "As a guideline" , a starting point, perhaps yes.
Now if others wish to put words in my mouth, so be it.
I still stick to my guns that headphone measurements are approximate & variable. To use them to derive an EXACT EQ profile is just plain silly, I have other reasons too.
But I do know, there are even cults out there that use 20 band, even more PEQ, to iron out every bleep they see on a curve.
They can have their fun, just don't call it science!
Even if they have a copy of Dr. Olive's books, they should put them up on eBay.

 

[Robbo99999]

I had already said:

View attachment 390732

My beef has never been to dismiss FR measurements, my only beef has been with the notion of "EQ'ing accurately" or "exact EQing".
To use the available curves "As a guideline" , a starting point, perhaps yes.
Now if others wish to put words in my mouth, so be it.
I still stick to my guns that headphone measurements are approximate & variable. To use them to derive an EXACT EQ profile is just plain silly, I have other reasons too.
But I do know, there are even cults out there that use 20 band, even more PEQ, to iron out every bleep they see on a curve.
They can have their fun, just don't call it science!
Even if they have a copy of Dr. Olive's books, they should put them up on eBay.

You're just saying what you initially said without listening to my replies. Oh well.

 

[Ken Tajalli]

It is because , I don't know where you stand. Let's take your reply:

Even though we accept that there unknowns in the system like unit to unit variation & on head variation then it still doesn't mean that it's not worth EQ'ing accurately to your best starting point from the measurement because that's the best information you've got at the time, so as a starting point there would be no point in randomly deviating your EQ from the best information available by not EQ'ing accurately to it.

Here you are putting words in my mouth. At what point did I claim or state that we should randomly deviate, or randomly do anything? Indeed, I have said quite the opposite, haven't I?
"Use the FR curve as a guideline", those were my words, then and now.
And also tell me, on what logic or reasoning, can you justify, that on one hand you accept:
- there are unknowns
- there are unit to unit deviations and head variations
And on the other hand, despite those, you want to do a PRECISE (am I shouting the operative word loud enough?) EQ based on that? am I getting you wrong?
And then on the third hand (if you had one) you claim, it is only to start off with!
You are all over the place with your arguments.

If you do EQ accurately to it, then that's either your starting point or if it sounds good then it's your end point. I don't see any merit in not EQ'ing accurately to a measurement in the first place. The level of precision that Oratory uses to EQ to the target curve is sensible, he doesn't EQ up really narrow sharp dips and he uses "only" 10 bands of EQ. His initial measurement is also smoothed in some respect inasmuch that it's an average of multiple measurements per unit & also of multiple units of headphones, so that is already taking out what would otherwise be unnecessary or detrimental EQ.

I don't see any logic to do it EXACTLY , PRECISELY OR ACCURATELY! We know, the FR curves are approximate, so a loose EQ, using the available measurements, is a logical starting point, to be fine tuned by ear! the listener's ear, not Oratory's ear, not miniDSP ears, EQ is mostly subjective. One of the biggest variables is the listener's hearing deficiencies.

You're just saying what you initially said without listening to my replies. Oh well.

Forget about the 10kHz and above discrepencies. Below that, I can see +/- (god knows)dB variations, that is the same headphone just tested again and again.

Headphones v1.5 - Graph Tool

www.rtings.com

Here is another:

 

[markstein]

Oratory's eq sounds too bright to me. I am using Amir's eq and adjust the bass level to taste (usually less bass than Amir prefers).

 

[Bernard23]

I've modified all of his profiles for my phones, other than LCD-X where I use the Roon Audeze profile. They're just a starting point, mostly to eliminate some of the design inadequacies.

 

[Robbo99999]

It is because , I don't know where you stand. Let's take your reply:

Here you are putting words in my mouth. At what point did I claim or state that we should randomly deviate, or randomly do anything? Indeed, I have said quite the opposite, haven't I?
"Use the FR curve as a guideline", those were my words, then and now.
And also tell me, on what logic or reasoning, can you justify, that on one hand you accept:
- there are unknowns
- there are unit to unit deviations and head variations
And on the other hand, despite those, you want to do a PRECISE (am I shouting the operative word loud enough?) EQ based on that? am I getting you wrong?
And then on the third hand (if you had one) you claim, it is only to start off with!
You are all over the place with your arguments.

I don't see any logic to do it EXACTLY , PRECISELY OR ACCURATELY! We know, the FR curves are approximate, so a loose EQ, using the available measurements, is a logical starting point, to be fine tuned by ear! the listener's ear, not Oratory's ear, not miniDSP ears, EQ is mostly subjective. One of the biggest variables is the listener's hearing deficiencies.

Forget about the 10kHz and above discrepencies. Below that, I can see +/- (god knows)dB variations, that is the same headphone just tested again and again.

View attachment 390828

Headphones v1.5 - Graph Tool

www.rtings.com

Here is another:

View attachment 390836

There's nothing wrong with EQ'ing exactly from a measurement if that's your best data available, as that's your best starting point. As long as you don't EQ up sharp dips, and best case you're using something like Oratory information where his published measurement is an average of multiple reseats of multiple units of headphones. I don't know how better to describe it - basically if you're averaging reseats as well as data between units the unit to unit variation and reseat quirks are being averaged out so the unreliable quirks don't get EQ'd as part of the final EQ because it's already been smoothed out in the measurement. And by choosing an average measurement to EQ from then there's a higher statistical chance of your headphone being closer to that measurement (characteristics of bell curve normal distribution if I remember rightly), which makes the EQ more valid for your unit. EQ'ing accurately to the target from the average measurement (which is what Oratory does) then just ensures you're as close as possible to the target. If it's not EQ'd accurately to the target then you're introducing small or larger variations away from the target unnecessarily. And because he only uses 10 bands of EQ then it's not like for every headphone he can EQ it with total precision to the target curve over the whole frequency range, and particularly not if there are some sharp dips that can't be EQ'd up, but otherwise there's no downsides & only positives if EQ'ing accurately to target from an average measurement - it just gets you to your best starting point & possibly already at your end point. So even though there are unknowable variations regarding unit to unit variation and on head variation then still it's in your best interest to EQ accurately to a target curve, especially if it's good information of the likes of an averaged measurement from multiple units. I don't know how better to explain it.

 

[goat76]

Just adding my input --- my own experience so far also does not support the EQ that Oratory has up. I wouldn't use his measurements for the LCD-X, rather I'd use sine sweeps and comparison to other headphones, speakers, and IEMs that you altready know sound perceptually "flat" or "correct" to your ears, maybe referencing Oratory's measurements for a general idea of where some issue areas might be and where the center frequencies for some peaks could be set. I don't know why this is; usually Oratory's EQs are pretty useful for starting points, but in this case his profile causes more issues than it solves imo.

I also found the Oratory's EQ settings to be way too bright for my LCD-X. But with just a few changes to some of the filters, I now find it pretty neutral sounding. Original Oratory EQ settings versus the changes I made in red. It's nothing drastic but I find it enough with the fairly small changes I made.
Try it out and see if you like it.

 

[Rayman30]

My preferred EQ I use for my LCD-X (2021)

 

[Crash]

Hey everyone! I recently bought a pair of these after listening and comparing for three hours on a T&A DAC200 (or maybe it was the HA200) and now I'm looking to recreate that experience on a smaller budget. Any good AIO solutions under 1k that will get me that deep jaw-dropping bottomless chasm of sound that I know these are capable of? Would something small like the Element III really deliver? Any real audible difference with discrete vs chip based amplifiers?

 

[Rayman30]

Hey everyone! I recently bought a pair of these after listening and comparing for three hours on a T&A DAC200 (or maybe it was the HA200) and now I'm looking to recreate that experience on a smaller budget. Any good AIO solutions under 1k that will get me that deep jaw-dropping bottomless chasm of sound that I know these are capable of? Would something small like the Element III really deliver? Any real audible difference with discrete vs chip based amplifiers?

The Element III is fantastic, it has nice fit and finish and a digital potentiometer that has 2 gain stages that it automatically selects between. It has plenty of power and will drive the LCD-X easily, I have the Element II and it has no issues with my LCD-X. As far as sound quality, it won't be different enough to matter from the other setup you tested on.

 

[Nkam]

Either a headphone sounds right across all music and reproduces the music correctly or it doesn’t.

This whole not using EQ for headphones is a bunch of hoopla for people to buy more headphones.

Just because it’s more difficult to measure headphones because of ear anatomy and other variables doesn’t mean there is no ‘ proper ‘ freq response with maybe some slight deviations.

For example the LCDX without EQ sounds like a$$. Plain and simple.
It’s a difficult headphone to tune without doing some sweeps from 1khz and up.
Oratory is a good starting point but I agree it’s too bright and shouty.

Id post my EQ settings for it, but because of ear anatomy it’s gonna be different for others

 

[the8o88y]

Hi everyone! I've had my LCD-X for over 10 years, and sent them in for the 2021 drivers upgrade when they became available. I very recently started dabbling with EQ, and am glad I did (using MathAudio with Foobar/wasapi.)

Without any measuring due to my inexperience, I sampled several EQ's I saw on this thread, including Resolve's from Headphones.com, and have really been enjoying the simple 5 band variation from Amir's original post here. I became curious enough to e-mail Audeze about my upgraded freq. test chart, as well as buying and installing the upgraded mesh/dampening, and could really use your help.

I know every chart is at least slightly different, and I'm not completely sure that, while my headphones sound better than they ever have with the settings I'm using, they are the optimal settings for my phones. Would anyone be able to show me eq settings that would properly correspond with my attached graph?