Audeze LCD-XC 21 & effects of small venting

[Ken Tajalli]

A while back, I asked on LCD-XC review page, for feedback on subjective effects on bass output, by removing one screw from the cups.
I got talked down to by some members, as to the pitfalls of such an experiment, and how without measurements, it would be a waste of time. And the wasting time argument carried on for a good few posts! (I know, I know, it takes two to tango, guilty as charged )
I did not get even one member reporting back.
Subjectively, I believe users have acknowledged, that this headphone lacks deep bass. Amir has said it, others have said it, and I agree with them.
So why not just EQ?
On ASR, and I believe on our sister forum the Headfi, there is at least one member, who continually posts new EQ figures, with preamp cuts of up to 18-20dB!
When you need to EQ bass up by a large amount, one must apply preamp cuts, or you get into digital clipping. You can soon run out of gain! -18dB is alot!
My +4dB gain in bass (see later) equates to 2.5X in amplifier power.

So I invested in a miniDSP EARS, to actually measure my headphones.
The EARS are not factory standard for measuring headphones! Hell, even the best of reference measuring gears out there do not always agree, as conditions of tests, unit to unit variations etc. , even the seating of headphones and size of one's ears, all affect the graphs/results obtained.
However, my humble testing gear can serve one purpose! The effects of the venting on my headphones! i.e. the changes due to it.
So here are my "newbie" results:
1- this is what Audeze kindly has sent me. This is their QC FR curve for my pair:

2- This is the graph of frequency response, comparing "Factory" i.e. all four screws in place & one screw taken out & two screws taken out:

Blue is all screws in, Purple is one screw taken out and Green with two screws out.
- I can see, past 500Hz, the effects are non-existent. It is all happening from 20 to 500Hz.
- I can also see that Audeze engineers, got it right! The Blue is the smoothest, and flattest curve, bear in mind this is a zoomed graph. For those interested, here is a zoomed out version:

The effect of venting, boosts deep bass, but reduces the upper bass a little. With one screw out, I lose 1.5dB peak 250Hz, but gain 4dB at 40Hz. Two screws out, would give me another 1-1.5dB boost.
Here are the distortion curves, mostly noise (50Hz, 100Hz etc.):

First graph is factory, then one screw, and then 2 screws out.
I don't see much difference! Noise figures are a little different, and also don't forget the cups are now fastened more loosely with screws missing.
I am sure, If I could implement the right vent, with correct damping, while cups are secure, I would get better results.

My next project is changing the screws to these:

These are stainless steel screws, with a 1.2mm hole drilled into them. Both ends of the holes are flared out.
I shall report, when done.

 

[Robbo99999]

That's good that you decided to get a miniDSP EARS so you can measure it and see the result of your mods. I think that's a fine approach...I mean it is easier to EQ it, and that's still my advice to people vs doing hard mods, but at least now you can see the effects of your physical mods on the frequency response and also if it has any negative effects re distortion. If you're dead set resigned on going down the physical mod route, then I'm pleased you got a miniDSP EARS to measure it, that means you're not wasting your time.

You could also use the miniDSP EARS to do a channel balancing EQ, but considering you're somewhat against EQ considering you're going down the route of hard mods, then this probably isn't applicable to you!

 

[Ken Tajalli]

That's good that you decided to get a miniDSP EARS so you can measure it and see the result of your mods. I think that's a fine approach...I mean it is easier to EQ it,

Point is (was) that according to Amir, this headphone requires 9dB of boost ay bass. Which means an equal amount of preamp cut, if one is to avoid digital clipping.
9dB cut, reduces a 500mW headphone amp to a 60mW amp!
I managed to get at least half of that, with a simple, non-invasive , fully reversible mod.
Happy EQing.

BTW, I wasn't wasting my time, was I?
There's your proof and evidence.

 

[ObjectiveSubjectivist]

Nice I really appreciate your approach.
Mini dsp is completely fine when doing mods/tuning headphones for yourself.
Sanity check for little money.

 

[Robbo99999]

Point is (was) that according to Amir, this headphone requires 9dB of boost ay bass. Which means an equal amount of preamp cut, if one is to avoid digital clipping.
9dB cut, reduces a 500mW headphone amp to a 60mW amp!
I managed to get at least half of that, with a simple, non-invasive , fully reversible mod.
Happy EQing.

BTW, I wasn't wasting my time, was I?
There's your proof and evidence.

Yes, you were wasting your time before you got the miniDSP EARS, now you're not wasting your time, which I've said - you definitely sound like you want a pat on your back! I still think it's better to use EQ, but at least you're not blatantly wasting your time now.

 

[Ken Tajalli]

Yes, you were wasting your time before you got the miniDSP EARS, now you're not wasting your time, which I've said - you definitely sound like you want a pat on your back! I still think it's better to use EQ, but at least you're not blatantly wasting your time now.

There you go again, posting posts after posts, to tell people they are wasting time.
Look in the mirror my friend.
Your posts are getting less and less constructive, more and more argumentative , for the sake of arguing.
You keep EQing.

 

[SRKRAM]

A while back, I asked on LCD-XC review page, for feedback on subjective effects on bass output, by removing one screw from the cups.
I got talked down to by some members, as to the pitfalls of such an experiment, and how without measurements, it would be a waste of time. And the wasting time argument carried on for a good few posts! (I know, I know, it takes two to tango, guilty as charged )
I did not get even one member reporting back.
Subjectively, I believe users have acknowledged, that this headphone lacks deep bass. Amir has said it, others have said it, and I agree with them.
So why not just EQ?
On ASR and I believe on our sister forum the Headfi, there is at least one member, who continually posts new EQ figures, with preamp cuts of up to 18-20dB!
When you need to EQ bass up by a large amount, one must apply preamp cuts, or you get into digital clipping. You can soon run out of gain! -18dB is alot!
Our +4dB gain in bass (see later) equates to 2.5X amplifier power.

So I invested on a miniDSP EARS, to actually measure my headphones.
The EARS are not factory standard for measuring headphones! Hell, even the best of reference measuring gears out there do not always agree, as conditions of tests, unit to unit variations etc. , even the seating of headphones and size of one's ears, all affect the graphs obtained.
However, my humble testing gear, serves one purpose! The effects of the venting on my headphones! i.e. the changes due to it.
So here are my "newbie" results:
1- this is what Audeze kindly sent me. This is their QC FR curve for my pair:
View attachment 270983
2- This is the graph of frequency response, between "Factory" i.e. all four screws in place & one screw taken out & two screws taken out:

View attachment 270958

Blue is all screws in, Purple is one screw taken out and Green with two screws out.
- I can see, past 500Hz, the effects are non-existent. It is all happening from 20 to 500Hz.
- I can also see that Audeze engineers, got it right! The Blue is the smoothest, and flattest curve, though, bear in mind this is a zoomed graph. For those interested, here is a zoomed out version:
View attachment 270963
The effect of venting, boosts deep bass, but reduces the upper bass a little. With one screw out, I lose 1.5dB peak 250Hz, but gain 4dB at 40Hz. Two screws out, would give me another 1-1.5dB gain.
Here are the distortion curves:

View attachment 270960 View attachment 270961 View attachment 270962

First graph is factory, then one screw, and then 2 screws out.
I don't see much difference! Noise figures are a little different, and also don't forget the cups are now fastened more loosely with screws missing.
I am sure, If I could implement the right vent, with correct damping, while cups are secure, I should get better results.

My next project is changing the screw, to these:
View attachment 270972 View attachment 270973

These are stainless steel screws, with a 1.2mm hole drilled into them. Both ends of the holes are flared out.
I shall report, when done.

I would have expected more responses to your post on the main forum thread than on the review thread. I think I was the only one to reply on the main thread.
As I wrote there, there was some discussion about using rubber o-rings as cup spacers. It's been a while since I sold my XCs, but I do recall leaving the screws a little loose as I preferred it that way.
I bought them used, and the cup screws were very loose when I got them, so the original owner may have been tinkering as well.

 

[IAtaman]

Yes, you were wasting your time before you got the miniDSP EARS, now you're not wasting your time, which I've said - you definitely sound like you want a pat on your back! I still think it's better to use EQ, but at least you're not blatantly wasting your time now.

I also think there might be good reasons to "tune" with physical mods if such mods are easy to do, reliable and effective, instead of EQ as well. Plus DIY has always been part of the hobby for me, and I suspect for many other people including Ken Tajalli as well, so modding has also that going for itself. I am waiting for the delivery of a STM32 development board myself so that I can (attempt to) build my own digital EQ hardware and software. I can use an EQ software as you suggested, can buy an RPI-4 and install CamillaDSP on it, or us the ADI-2 DAC or Qudelix 5K sitting on my desk, with even better results as well, but where is the fun in that.

 

[Robbo99999]

There you go again, posting posts after posts, to tell people they are wasting time.
Look in the mirror my friend.
Your posts are getting less and less constructive, more and more argumentative , for the sake of argumentation .
You keep EQing.

Really? I think you need to learn to read, you're misrepresenting what I'm saying. You enjoy your physical mods and I'm glad you've now got the miniDSP EARs to back up what you're doing with measurements, which is crucial.

 

[Ken Tajalli]

I would have expected more responses to your post on the main forum thread than on the review thread. I think I was the only one to reply on the main thread.
As I wrote there, there was some discussion about using rubber o-rings as cup spacers. It's been a while since I sold my XCs, but I do recall leaving the screws a little loose as I preferred it that way.
I bought them used, and the cup screws were very loose when I got them, so the original owner may have been tinkering as well.

Yes I saw your post, can't remember if I replied.
By using washers and o'rings , one can get a similar effect. But, as a first step, I wanted to know, if others can very easily, repeat the experiment and get similar results.
Frankly science says it should work, and as we can see now from measurements, it does.
But measurements at home, has cost me £300! OK, I got a new toy now, and I am sure I can get most of my money back on resale.
Going back to o'rings, I decided against it, as it would introduce extra compliance to cup and the housing, it might change the distortion performance.
Ultimately, pin holes and solid screw fastening the cups, would have been my choice.
To make sure, it is a reversible mod. the vented screws were chosen to accommodate that.
Regarding finger tight screws, I believe Audeze deliberately do that, since the assembly is made of soft Alu. , if you over-tighten the steel screws from the cups, it would be easy to bugger the Alu. threads, hence why I put a warning on original post.
Some people just do not understand the reason why EQing bass up has pitfalls , I had tried to explain.
To some, EQ cures a rainy day, with zero consequences.
There is a member, who posts newer and newer EQ profiles for this headphone, and they all start off with massive preamp cuts.
If you use his profiles, you will need a speaker poweramp to run your headphones!
With mobile devices, power output is at a premium. Even desktop models would struggle with gains at -18dB preamp reductions, as the first step.
Even @amirm 9dB bass boost suggestion, would reduce max. power output by a factor of 1/8X.
Luckily, LCD-XC is both efficient and has low distortion .

Final Look:
I think the stainless steel bolts look better than original black ones. If you look closely, there was already stainless steel metal against the black body on yoke screws, the cable inlet, the rod . . .

 

[Ken Tajalli]

Final results:

Zoomed in, 0.5dB/div.
From bottom to top.
- All bolts closed.
- One bolt vented, three closed.
- two vented, two closed
- three vented, one closed
- all four vented.
- Three closed, one bolt fully taken out.

Same graph, but 1dB/div :

Distortion charts:
More or less, same. Very low!

 

[Robbo99999]

Final results:

View attachment 271462
Zoomed in, 0.5dB/div.
From bottom to top.
- All bolts closed.
- One bolt vented, three closed.
- two vented, two closed
- three vented, one closed
- all four vented.
- Three closed, one bolt fully taken out.

Same graph, but 1dB/div :
View attachment 271464

Distortion charts:
More or less, same. Very low!

View attachment 271466 View attachment 271467 View attachment 271468 View attachment 271469 View attachment 271470

You might wanna tell people if it's important which of the various bolts they leave in or out or replace out of the 4 possible bolts, as maybe that would effect the results? It would probably also be interesting if you take the effects you're seeing there of your mods and then apply that to the GRAS measurement of the headphone (probably Oratory's), that way they could see the effect of the changes directly on the bonafide GRAS graphs vs say the Harman Curve standard - as most people are used to looking at GRAS measurements vs a Harman Target Curve.

 

[Ken Tajalli]

You might wanna tell people if it's important which of the various bolts they leave in or out or replace out of the 4 possible bolts, as maybe that would effect the results? It would probably also be interesting if you take the effects you're seeing there of your mods and then apply that to the GRAS measurement of the headphone (probably Oratory's), that way they could see the effect of the changes directly on the bonafide GRAS graphs vs say the Harman Curve standard - as most people are used to looking at GRAS measurements vs a Harman Target Curve.

Thank you, a constructive suggestion.
Which screw to take out is irrelevant, makes no difference.
As to your second suggestion, here they are, to the best of my Photoshop ability:
Before plot is from "Resolve" on a different sample of the same headphone.

 

[LearningToSmile]

Nice work, and also good advertisement for the miniDSP EARS- people complain about it, but looks like it's doing a pretty good job here. I've been trying to EQ headphones that I previously swapped pads on, and without measurements it was quite annoying, as measurements I found online were no longer correct, and doing it by ear involves a bunch of trial and error. And I'm considering buying a second pair of headphones to do more drastic mods(Lawton cups) on, maybe I should start looking for a deal on a second hair set of EARS.
There's also the chinese clone couplers but that seems like a much bigger investment if you want to measure headphones and not just IEMs.

 

[Ken Tajalli]

I think ALL measuring jigs are just toys!
none are accurate or real world.
I mean, checking the bass output on a jig (rig?) is so temperamental, the slightest air gap throws the whole thing out of the window, yet listening to headphones, with or without glasses (i.e. air gap or not) does not send bass output into non-existence.
Those Harman curves, are only valid, if conditions of the test can be duplicated exactly! but so many people out there using them, obviously, it is not happening.
As I said before, these TOYS! are good for one thing - that is showing differences.
To get my results, I positioned the phones on to get a decent result, then ALMOST without touching the phones, or its position, I took screws out, or replaced them with vented screws (hole in the middle), and remeasured.
Once you remove the phones from the rig, the next reading you get is no longer valid.
After I had changed all screws with vented one and took my reading, to re-affirm my readings, I gently inserted the tips of four cocktail sticks into the holes and took one last reading!
Yep, the holes were doing their job, blocking them, revert the phones to "Factory".
Incidentally, the lowest THD figure belongs to mod. (0.797%) and highest to "Factory" (1.16%). The figures are not accurate, so I am not cheering, but all of them are very very low, kudos to Audeze.
A point has come to my attention.
By doing this simple mod. I am disturbing the default sound-isolation which is natural to all closed backs. Sound leaks out a little, and leaks in too!
So when my wife shouts at me while I have the headphones on, I can hear her a little, though I pretend I don't (I hope she is not a member here).
It is not a lot, but noticeable.
I am thinking of some kind of pad or sponge, to position on the inside of the cups, where the screws are, to act as a barrier for higher frequencies, yet allow slower bass frequencies to vent. Not high on my list, but I am wondering!!!

 

[Robbo99999]

Thank you, a constructive suggestion.
Which screw to take out is irrelevant, makes no difference.
As to your second suggestion, here they are, to the best of my Photoshop ability:
Before plot is from "Resolve" on a different sample of the same headphone.

View attachment 271614 View attachment 271613

Ah, you've just photoshopped the curves over eachother, that's not really very accurate, as for instance your one screw out only increases dB by +1dB in the bass but you've got it showing as around +4.5dB for the effect of the first screw (vs the Resolve measurement). What you'd need to do is apply the difference between your various measurements to the Resolve measurement, but you've literally just photoshopped the the various curves over eachother, which is not valid at all, and in fact misrepresentative, as you're giving the impression of way way more bass with your mods vs the Resolve measurement, which is not the case. It is somewhat complicated using REW to work out the difference between your curves and then to apply that to Resolve's measurement - the way I've done that in the past is by using REW to EQ from one comparative line to the other (for instance all screw vs one screw out), you then save that EQ in REW, you'd then load up Resolve's measurement into REW and then apply the EQ you've worked out - now that would then be the accurate effect of you taking the screw out, but you can't just photoshop it over like you have, that's really quite misrepresentative of the effects of your mods. (Sorry to be a pain!)

 

[Ken Tajalli]

Ah, you've just photoshopped the curves over eachother, that's not really very accurate, as for instance your one screw out only increases dB by +1dB in the bass but you've got it showing as around +4.5dB for the effect of the first screw (vs the Resolve measurement). What you'd need to do is apply the difference between your various measurements to the Resolve measurement, but you've literally just photoshopped the the various curves over eachother, which is not valid at all, and in fact misrepresentative, as you're giving the impression of way way more bass with your mods vs the Resolve measurement, which is not the case. It is somewhat complicated using REW to work out the difference between your curves and then to apply that to Resolve's measurement - the way I've done that in the past is by using REW to EQ from one comparative line to the other (for instance all screw vs one screw out), you then save that EQ in REW, you'd then load up Resolve's measurement into REW and then apply the EQ you've worked out - now that would then be the accurate effect of you taking the screw out, but you can't just photoshop it over like you have, that's really quite misrepresentative of the effects of your mods. (Sorry to be a pain!)

Photoshop is all I could do.
The curves are pretty much accurately portrayed, I overlaid the "factory" curve that I had, over "Resolves" curve, making sure all peaks correspond with his at the right frequencies and levels. I managed to get a better seal, so my Factory curve does not droop like his below 60Hz. My curve (also in blue) sits on his curve from 60Hz to 250Hz almost perfectly.
Take a look again, the levels portrayed are accurate. Compare his "before" curve, with curves I had posted before. Remember, taking a whole screw out does boost the bass by almost 5dB. With four 1.2mm vented screws, I got about 3dB of boost.

The replacement screws that I have made, which I have referred to as Vented had a 1.2mm hole drilled in. Now doing the maths, it would tell us that taking one 3mm dia. screw out completely open a surface area of 7mm2 ( 1.5x1.5x3.14), while each vented screw opens a surface area of just 1.13mm2.
So four vented screws together do not equate to even one screw being taken out, hence, even with all screws being replaced, I get no more than 3dB of bass boost, compared to 4-5dB achieved by removal of one screw.
I have since redrilled the holes to make them to 1.5mm. Now the four vented screws do equate to one missing screw, in surface area, and I am getting 4-5dB of bass boost.
It is fortunate, that the mod. boosts the bass, exactly where this headphone needs it, has no effect on the rest of the range, nor on distortion.
Frankly, I no longer need to EQ. the bass up! which means, no preamp cut, whatsoever.
The quality of bass is good, which was a selling point of the phones, at least to me. I just EQ the treble down a bit, and that's all I need to do.

 

[Robbo99999]

Photoshop is all I could do.
The curves are pretty much accurately portrayed, I overlaid the "factory" curve that I had, over "Resolves" curve, making sure all peaks correspond with his at the right frequencies and levels. I managed to get a better seal, so my Factory curve does not droop like his below 60Hz. My curve (also in blue) sits on his curve from 60Hz to 250Hz almost perfectly.
Take a look again, the levels portrayed are accurate. Compare his "before" curve, with curves I had posted before. Remember, taking a whole screw out does boost the bass by almost 5dB. With four 1.2mm vented screws, I got about 3dB of boost.

View attachment 271683

The replacement screws that I have made, which I have referred to as Vented had a 1.2mm hole drilled in. Now doing the maths, it would tell us that taking one 3mm dia. screw out completely open a surface area of 7mm2 ( 1.5x1.5x3.14), while each vented screw opens a surface area of just 1.13mm2.
So four vented screws together do not equate to even one screw being taken out, hence, even with all screws being replaced, I get no more than 3dB of bass boost, compared to 4-5dB achieved by removal of one screw.
I have since redrilled the holes to make them to 1.5mm. Now the four vented screws do equate to one missing screw, in surface area, and I am getting 4-5dB of bass boost.
It is fortunate, that the mod. boosts the bass, exactly where this headphone needs it, has no effect on the rest of the range, nor on distortion.
Frankly, I no longer need to EQ. the bass up! which means, no preamp cut, whatsoever.
The quality of bass is good, which was a selling point of the phones, at least to me. I just EQ the treble down a bit, and that's all I need to do.

That's still not valid unfortunately, miniDSP EARS can vary in bass even vs a flat plate GRAS system like Amir's, from his review of the LCD-XC 21:

You can see that bass is a fair way off Harman, yet in your photoshop you have your stock curve ("Factory") already very close to Harman bass levels. There is more bass in Amir's measurement vs the Resolve measurement you chose to show, but it's certainly disingenuous to show your measurements directly superimposed on Resolve's measurement in particular because your mod doesn't give that kind of comparative uplift in the bass vs his measurement. There's no way round it, it's not valid to overlay your miniDSP EARS bass measurements directly over someone else's GRAS measurement. It's only valid if you compute the comparative differences between your various miniDSP EARS bass curves and then apply that difference to a GRAS measurement of your choice for the LCD XC 21 (Oratory / Amir / Resolve for instance). I'm sorry, I know it's fairly complicated to actually do what I'm saying, but it's not valid to superimpose your miniDSP EARS measurements on top of a GRAS measurement.

(Oratory measurement has less bass, and he shows best case bass, as in perfect seal):

EDIT: actually it's not good to use Resolve's measurement in terms of showing his graph, because he uses a different target to Harman (less bass), so that will throw people off as they're expecting the target curve in the graph to be Harman, which will screw up their comparisons, so best not to use that Resolve graph like you did for that reason too, as it confused me for a while as well, which is why I crossed out part of my post above.

EDIT: I'm currently typing out a workflow on how to do the proper work in REW........done - it's the following post.

 

[Robbo99999]

Here's how you'd apply the comparative difference between your curves to a GRAS measurement of your choice using REW:

1. Choose a GRAS measurement of your choice, I suggest Oratory for best accuracy, most units measured as a general rule so enhanced relevance/accuracy.
2. Capture Oratory measurement into data by either tracing using VirtuixCAD or download the measurement data from AutoEQ.
3. Import the captured data measurement into REW. (set to one side, we'll use this later).
4. Have all your different miniDSP EARS measurements in REW.
5. Work out the comparative difference between your miniDSP EARS measurements, described in following sentences. Set your 1-vented screw measurement as a Target Curve in REW. Go into your Factory Stock miniDSP EARS measurement, then access the EQ function in REW to EQ it to your Target Curve (which is 1-vented screw measurement). Save that EQ as "1-Vented Screw". That EQ is your comparative difference, and is showing you the effect of your vented screw mod.
6. Go into the GRAS measurement in REW (that you created in Step 3). Go to the EQ function in REW. Load up the EQ that you saved in Step 5 "1-Vented Screw" and apply that EQ. That will show you the effect of the 1-vented screw on that GRAS measurement. Then select the option in REW to "Save Predicted as Measurement" - it's called something like that, which will save that as a new measurement in your REW project.
7. You'd do Steps 5 & 6 for each of your different screw mods vs Factory Stock (obviously changing the target each time in Step 5 to your new mod). You'll now end up with GRAS simulated measurement curves for each of your screw mods which you can show altogether on one graph.

That would be the valid way to do it, even though it takes a lot more effort. Unfortunately you superimposing miniDSP EAR measurements over someone else's GRAS measurement is not valid or accurate and is misrepresentative.

(you don't have to go to all this trouble though, in which case just use your miniDSP EARS graph comparisons that you showed in your very first posts without any reference to GRAS measurements or Harman......just if you're gonna reference it to GRAS then you can't just superimpose your miniDSP EARS measurements on top of a GRAS measurement, in which case I suggest you delete your superimposed graphs in your earlier post to avoid misrepresentations).

 

[Ken Tajalli]

Thanx, I shall do it in proper time.
So far, i haven't found oratory's fr curve for LCD-XC 2021 model.
The newer model is different to older versions.
Can you send me a link, if there is one?