flinkenick's 17 Flagship IEM Shootout Thread (and general high-end portable audio discussion)
Dec 7, 2017 at 4:44 AM Post #6,662 of 39,414
So a few interesting comments from Karthick Mannivan (research director at Audeze) about the advantages of planar technology:

Transparency: Other than electrostatics, none of the of the contemporary in-ear technologies (BA/DD) can match the speed of the ultra thin planar drivers we use. The reason is simple physics: for a driver to be able to quickly accelerate and also stop quickly, it needs to have very low inertia or mass. The nano grade diaphragm we use in i4 is orders of magnitude lighter than any DD or BA diaphragm. Which means the diaphragm is able to very closely follow the audio signal.

Distortion: Planar technology in general has very low distortion, but with our in-ear models such as iSine and i4, we measured extremely low distortion, likely lower than anything else in the market. The drivers used in i4 can take ear shattering levels and still not distort; this is because of the very linear response of our driver, and also the very fine control we have over the driver. There are no issues of related with bass and excursion either. This again has to do with the physics of the driver design in BAs and DDs; when you push the driver to do something that stresses them mechanically, they start distorting.

Phase coherence: i4 has a single driver design, full range at 10Hz - 20kHz. BAs employ multiple drivers targeting multiple frequency ranges and the gain for each of these ranges are tuned (not very different from a EQ), then the crossover is designed to put the sound together from all the different drivers so that they sound as a single driver. This is easier said than done; acoustically trying to do so by matching a crossover so they sound coherent is very hard to do, so a single planar driver that is creating the sound and has a very linear response will automatically have an advantage.

Planar bass: When it comes to bass, the large well controlled planar drivers provide the most natural and deep bass, flat from 10Hz through 900Hz. Many iem are tuned to provide a boosted bass, which is partly to overcome the limitations of the drivers and therefore seldom sound natural.

Short description of technologies:
Fluxor: Name for the magnetic technology used in i4..
Fazor: Wave guides used (in i4 there is one inside the sound port), to decrease any diffraction (reflections and cancellations) and provide a linear response.
Uniforce: The width of the traces in the diaphragm is varied so that the electro-magnetic force acting on the diaphragm is uniform across the diaphragm. This is again very important for a linear response, to eliminate distortion and provide good control. If the force is not uniform, then some parts of the driver will move faster than the other, which is not very planar.
 
Last edited:
Dec 7, 2017 at 4:52 AM Post #6,663 of 39,414
Dec 7, 2017 at 6:08 AM Post #6,664 of 39,414
Nice! how are you doing your EQ?
Currently on my Sony WM1Z (firmware 2.0) I have the EQ set as follows:
31hz +4.0
62hz +4.0
125hz -2.0
250hz -3.0
500hz -3.0
1k -3.0
2k -0.5
4k +8.0
8k +0.5
16k -5.0

I'm not good at describing sounds but this is how I like things to sound. :sweat_smile:
 
Dec 7, 2017 at 9:47 AM Post #6,665 of 39,414
Sorry to interrupt the eq discussion, but I just wanted to check in (was lurking, not posting) and say thanks to everyone here for all of the info re current TOTL IEM's. After months of due diligence, pm's to users and dithering about spending so much money on an IEM, I ordered a stock (stealth black) Universal Zeus XR from EE and am totally pleased with it. My concerns were cost, sound signature and hiss (not necessarily in that order). Several folks on Head-Fi using it with the Astell SP1000 assured me there was no hiss. Reading about its sound signature here (and on other threads) convinced me it would be "right" for me. As to cost, that was between me and my wife. :darthsmile:

I've had it for a few weeks now, gotten it nicely burned-in, gotten proper tips for it, and it is everything I hoped it would be. And I have found that I prefer the X(lV) setting over the R for most (but not all) of my music. Since I had been seriously considering the R, spending the extra money for the XR was the right choice.

So, thanks again, and now back to more pages of discussion about eq settings. :)
 
Dec 7, 2017 at 10:58 AM Post #6,666 of 39,414
Sorry to interrupt the eq discussion, but I just wanted to check in (was lurking, not posting) and say thanks to everyone here for all of the info re current TOTL IEM's. After months of due diligence, pm's to users and dithering about spending so much money on an IEM, I ordered a stock (stealth black) Universal Zeus XR from EE and am totally pleased with it. My concerns were cost, sound signature and hiss (not necessarily in that order). Several folks on Head-Fi using it with the Astell SP1000 assured me there was no hiss. Reading about its sound signature here (and on other threads) convinced me it would be "right" for me. As to cost, that was between me and my wife. :darthsmile:

I've had it for a few weeks now, gotten it nicely burned-in, gotten proper tips for it, and it is everything I hoped it would be. And I have found that I prefer the X(lV) setting over the R for most (but not all) of my music. Since I had been seriously considering the R, spending the extra money for the XR was the right choice.

So, thanks again, and now back to more pages of discussion about eq settings. :)
Good to see you here. I've read your posts in the SE846 and Angie threads. Good times!
 
Dec 7, 2017 at 11:47 AM Post #6,667 of 39,414
Hey, @olddude, come a long way since the Pono and SE846! Still working my way up the TOTL iem food chain myself, nice to see a post from you!
 
Dec 7, 2017 at 12:51 PM Post #6,668 of 39,414
Sorry for being late to the EQ party. I guess I am not too irrelevant to this conversation, as I am the Vishnu that Nic referred to, in one of his posts. What follows is not to take sides, but rather some information that might shed some light on the situation.

I see that some are under the impression that the bump at 1.5 kHz is the sole reason for the i4/iSine earphones to sound unnatural. First of all, I am not comfortable calling the response at 1.5 kHz as a Bump. It is rather a discontinuity in the raise, that otherwise should continue to form a bump at around 3 kHz. The fact that the frequency response takes a dive after the 1.5 kHz is what gives an impression of a bump at 1.5 kHz when looking at measurements.

We do not know Audeze's headphone target curve. But I think it is somewhat safe to assume that, their headphone target curve resembles the raw measurement of iSine 20 with Cipher cable. If you look at the raw measurement of iSine 20 without the cipher cable, you’ll see that the delta between 500 Hz and 1.5 kHz is around 7 dB. And now if you look at the measurement of iSine 20 with the Cipher cable, you’ll see the delta between 500 Hz and 1.5 Khz is once again around 7 dB. So the response at 1.5 kHz is the same on both the curves.

Audeze_iSine20_Graph_FRPassive.jpg

Audeze_iSine20_Graph_FRLightning.jpg

One might argue that other widely popular and standard Headphone Target Curves only have a 3 or 4 db raise at 1.5 kHz. And for that, I would like to propose an exercise. I would like you to use one of your IEMs, that responds well to EQ. Try installing an app with PMEQ (Vox app for iOS or Neutron Player for Android) and try the following 2 scenarios.

1. Scenario 1: Now let’s try and roughly create a huge dip in the upper mid-range like what exists on the i4 and iSines: F = 3.5 kHz, Gain = -12 dB, Q = 1.5
2. Scenario 2: Now create another EQ preset with the following filter. Now, let’s try adding a slight bump at 1.5 kHz: F = 1.5 kHz, Gain = +4 dB, Q = 1.5

Here is roughly what you could expect to hear on the two scenarios. In Scenario 1, the vocals, pianos and other string instruments will lose some weight and forwardness forwardness and will start to sound delicate. The mid-range will lose a bit of transparency and articulation because, a good amount of their fundamental and overtone frequencies are heavily attenuated. In Scenario 2, all you’ll observe is, vocals gaining a touch of warmth & weight and the midrange brought forward by a very tiny bit. And for these reasons, Scenario 1 is the one that is likely to sound more unnatural.

But I'll let you try it for yourself and be the judge.
 
Last edited:
Dec 7, 2017 at 7:08 PM Post #6,669 of 39,414
Sorry for being late to the EQ party. I guess I am not too irrelevant to this conversation, as I am the Vishnu that Nic referred to, in one of his posts. What follows is not to take sides, but rather some information that might shed some light on the situation.

I see that some are under the impression that the bump at 1.5 kHz is the sole reason for the i4/iSine earphones to sound unnatural. First of all, I am not comfortable calling the response at 1.5 kHz as a Bump. It is rather a discontinuity in the raise, that otherwise should continue to form a bump at around 3 kHz. The fact that the frequency response takes a dive after the 1.5 kHz is what gives an impression of a bump at 1.5 kHz when looking at measurements.

We do not know Audeze's headphone target curve. But I think it is somewhat safe to assume that, their headphone target curve resembles the raw measurement of iSine 20 with Cipher cable. If you look at the raw measurement of iSine 20 without the cipher cable, you’ll see that the delta between 500 Hz and 1.5 kHz is around 7 dB. And now if you look at the measurement of iSine 20 with the Cipher cable, you’ll see the delta between 500 Hz and 1.5 Khz is once again around 7 dB. So the response at 1.5 kHz is the same on both the curves.



One might argue that other widely popular and standard Headphone Target Curves only have a 3 or 4 db raise at 1.5 kHz. And for that, I would like to propose an exercise. I would like you to use one of your IEMs, that responds well to EQ. Try installing an app with PMEQ (Vox app for iOS or Neutron Player for Android) and try the following 2 scenarios.

1. Scenario 1: Now let’s try and roughly create a huge dip in the upper mid-range like what exists on the i4 and iSines: F = 3.5 kHz, Gain = -12 dB, Q = 1.5
2. Scenario 2: Now create another EQ preset with the following filter. Now, let’s try adding a slight bump at 1.5 kHz: F = 1.5 kHz, Gain = +4 dB, Q = 1.5

Here is roughly what you could expect to hear on the two scenarios. In Scenario 1, the vocals, pianos and other string instruments will lose some weight and forwardness forwardness and will start to sound delicate. The mid-range will lose a bit of transparency and articulation because, a good amount of their fundamental and overtone frequencies are heavily attenuated. In Scenario 2, all you’ll observe is, vocals gaining a touch of warmth & weight and the midrange brought forward by a very tiny bit. And for these reasons, Scenario 1 is the one that is likely to sound more unnatural.

But I'll let you try it for yourself and be the judge.

ezgif-3-bd2f5ee7a6.gif
 
Dec 7, 2017 at 8:03 PM Post #6,670 of 39,414
Sorry for being late to the EQ party. I guess I am not too irrelevant to this conversation, as I am the Vishnu that Nic referred to, in one of his posts. What follows is not to take sides, but rather some information that might shed some light on the situation.

I see that some are under the impression that the bump at 1.5 kHz is the sole reason for the i4/iSine earphones to sound unnatural. First of all, I am not comfortable calling the response at 1.5 kHz as a Bump. It is rather a discontinuity in the raise, that otherwise should continue to form a bump at around 3 kHz. The fact that the frequency response takes a dive after the 1.5 kHz is what gives an impression of a bump at 1.5 kHz when looking at measurements.

We do not know Audeze's headphone target curve. But I think it is somewhat safe to assume that, their headphone target curve resembles the raw measurement of iSine 20 with Cipher cable. If you look at the raw measurement of iSine 20 without the cipher cable, you’ll see that the delta between 500 Hz and 1.5 kHz is around 7 dB. And now if you look at the measurement of iSine 20 with the Cipher cable, you’ll see the delta between 500 Hz and 1.5 Khz is once again around 7 dB. So the response at 1.5 kHz is the same on both the curves.



One might argue that other widely popular and standard Headphone Target Curves only have a 3 or 4 db raise at 1.5 kHz. And for that, I would like to propose an exercise. I would like you to use one of your IEMs, that responds well to EQ. Try installing an app with PMEQ (Vox app for iOS or Neutron Player for Android) and try the following 2 scenarios.

1. Scenario 1: Now let’s try and roughly create a huge dip in the upper mid-range like what exists on the i4 and iSines: F = 3.5 kHz, Gain = -12 dB, Q = 1.5
2. Scenario 2: Now create another EQ preset with the following filter. Now, let’s try adding a slight bump at 1.5 kHz: F = 1.5 kHz, Gain = +4 dB, Q = 1.5

Here is roughly what you could expect to hear on the two scenarios. In Scenario 1, the vocals, pianos and other string instruments will lose some weight and forwardness forwardness and will start to sound delicate. The mid-range will lose a bit of transparency and articulation because, a good amount of their fundamental and overtone frequencies are heavily attenuated. In Scenario 2, all you’ll observe is, vocals gaining a touch of warmth & weight and the midrange brought forward by a very tiny bit. And for these reasons, Scenario 1 is the one that is likely to sound more unnatural.

But I'll let you try it for yourself and be the judge.

I don't think there are really any sides to take here, we pretty much agree on the same thing. I have no qualms with a gradual rise (found on the PP8 for instance, but it peaks at 3-4k) since the effect still remains that the sub-2k frequencies are not in relative emphasis as compared to the higher target bumps. Hell, even the QDC 8SH/Gemini have an emphasis at 1.5k, but the massive difference is that it plateaus at said frequency and is boosted all the way to 4-5k.

The Audeze portaplanars on the other hand have a PEAK (I think people are misunderstanding my wording here, peak = highest point of the frequency response and therefore will NEED an incline thereafter to be classified as such) of 1.5k which is causing the unnatural sound. The DSP masks that 1.5k peak by boosting the 3-4k frequencies to an even higher degree so the 1.5k PEAK shifts to a 2-3k PEAK.
 
Dec 7, 2017 at 8:44 PM Post #6,671 of 39,414
Crinicle, why didn't you just post your graphs? By the way Nick instead of finding some random graphs to your liking, try to compare stuff measured in the same rig.

NT6

jPZwdv9.png

i4

hQJ6BLz.png

iSine20

O7FBAkS.png

Anybody see the large hump at 1.5khz? When I posted this, Audeze admitted it's peakier than that. Crinicle's graph is how Headwacker figured out where was causing the funny mids or vocals with the 1.5khz peak. Hawaiian boy, by ears figured out somewhere there is something funny in the mids. Some guys do it by ears nick. I heard incoherency in the mids, and even after EQ with DAPs, software EQ on the computer, still didn't sound right. Only LPG sounded properly driven and with EQ, the mids sounded proper. Other than the LPG, many DAPs do not have proper PEQ. You can get PEQ with Android devices, and tried those as well, Didn't sound proper.

You can go on and on about how Audeze tech is all that. Cool.
 
Last edited:
Dec 7, 2017 at 8:45 PM Post #6,672 of 39,414
I don't think there are really any sides to take here, we pretty much agree on the same thing. I have no qualms with a gradual rise (found on the PP8 for instance, but it peaks at 3-4k) since the effect still remains that the sub-2k frequencies are not in relative emphasis as compared to the higher target bumps. Hell, even the QDC 8SH/Gemini have an emphasis at 1.5k, but the massive difference is that it plateaus at said frequency and is boosted all the way to 4-5k.

The Audeze portaplanars on the other hand have a PEAK (I think people are misunderstanding my wording here, peak = highest point of the frequency response and therefore will NEED an incline thereafter to be classified as such) of 1.5k which is causing the unnatural sound. The DSP masks that 1.5k peak by boosting the 3-4k frequencies to an even higher degree so the 1.5k PEAK shifts to a 2-3k PEAK.

@crinacle , I followed the dialogue between you and Nic and I realized that both you guys were pretty much on the same page by the end of the conversation. And I did read your post where you clarified your usage of the term 'peak' when referring to the response at 1.5kHz. That is the reason I did not quote your post or mention you. But I don't think everyone is on the same page as you we are. Let's look at @SilverEars ' posts here:

The vocal, mids issue was fixed with the 1.5Khz drop(there were additional tweaks in EQ, bit this was causing the major issue). It still didn't sound right after EQ.

The 1.5Khz bump was the major issue. Keep playing with it, you'll see it is.

The best EQ I've used was from a member here named Headwacker. He saw Crinicle's graph with the 1.5khz peak and reduced the band around it by 7dB, and made 4 additional filters to his EQ, and the mids, vocals issues was fixed.

It sounds very very odd with the 1.5khz bump. Once again, I will emphasize the 1.5KHZ bump nick if you have not registered with your ears.

You can imagine how many people might have a similar impression about the response at 1.5 kHz but did not post here. Also Silverears says that he EQd down 7dB at 1.5kHz on his LPG. LPG's Q value in its EQ does not go above 1. Meaning, it can't do a narrow band filter. And so when someone EQs down 7dB at 1.5kHz with Q=1, here is a visual illustration on what happens.

IMG_2708.PNG

This EQ would basically reduce the 2-3kHz region further. That's the reason, if you look at my i4 review, I would have advised a narrow filter of Q=3 at 1.5kHz to reduce 5dB. The purpose of this filter is not to kill the 1.5kHz response, but to compensate for the addition of a few dBs that filter at 3.2kHz would add.

This cannot be recreated on LPG because for the lack of arrow band filter. Silverears did add that headwhackers EQ includes 4 additional filters. I am guessing that one of those 4 filters is adding some response at around 3kHz. On the LPG, reducing the 1.5kHz does more damage than good.
 
Last edited:
Dec 7, 2017 at 8:53 PM Post #6,673 of 39,414
Crinicle, why didn't you just post your graphs? By the way Nick instead of finding some random graphs to your liking, try to compare stuff measured in the same rig.

Anybody see are large hump at 1.5khz? When I posted this, Audeze admitted, it's peakier than that. Crinicle's graph is how Headwacker figured out where was causing the funny mids or vocals with the 1.5khz peak. Hawaiian boy, by hears figured out somewhere there is something funny in the mids. Some guys do it by ears nick.

You clearly do not have an understanding of headphone target curves. Come back when you do.
 
Dec 7, 2017 at 9:01 PM Post #6,674 of 39,414
@crinacle , I followed the dialogue between you and Nic and I realized that both you guys were pretty much on the same page by the end of the conversation. And I did read your post where you clarified your usage of the term 'peak' when referring to the response at 1.5kHz. That is the reason I did not quote your post or mention you. But I don't think everyone is on the same page as you we are. Let's look at @SilverEars ' posts here:





You can imagine how many people might have a similar impression about the response at 1.5 kHz but did not post here. Also Silverears says that he EQd down 7dB at 1.5kHz on his LPG. LPG's Q value in its EQ does not go above 1. Meaning, it can't do a narrow band filter. And so when someone EQs down 7dB at 1.5kHz with Q=1, here is a visual illustration on what happens.

[SPOILER = -7 dB at 1.5kHz]

[/SPOILER]

This EQ would basically reduce the 2-3kHz region further. That's the reason, if you look at my i4 review, I would have advised a narrow filter of Q=3 at 1.5kHz to reduce 5dB. The purpose of this filter is not to kill the 1.5kHz response, but to compensate for the addition of a few dBs that filter at 3.2kHz would add.

This cannot be recreated on LPG because for the lack of arrow band filter. Silverears did add that headwhackers EQ includes 4 additional filters. I am guessing that one of those 4 filters is adding some response at around 3kHz. On the LPG, reducing the 1.5kHz does more damage than good.

I've said it before but I'll say it again anyways for those still hazy about what's going on; it is the combo of the 1.5k peak and the 2-4k null that makes the Audeze portaplanars sound so weird. That is not to say that the reason for the weirdness is not solely due to the specific peak or null but rather both happening at the same time. The Campfire Jupiter is the most recognisable example of an IEM with a 2-3k null but does not sound as weird as the portaplanars because the latter have one very distinct difference: a 1.5k peak.

I guess my words got lost in translation at the start, I didn't say that the 1.5k peak is the only reason for the weirdness but rather is the reason for the unique weirdness that the 10/20/i4 exhibit.

Crinicle, why didn't you just post your graphs? By the way Nick instead of finding some random graphs to your liking, try to compare stuff measured in the same rig.

NT6



i4



iSine20



Anybody see the large hump at 1.5khz? When I posted this, Audeze admitted it's peakier than that. Crinicle's graph is how Headwacker figured out where was causing the funny mids or vocals with the 1.5khz peak. Hawaiian boy, by ears figured out somewhere there is something funny in the mids. Some guys do it by ears nick. I heard incoherency in the mids, and even after EQ with DAPs, software EQ on the computer, still didn't sound right. Only LPG sounded properly driven and with EQ, the mids sounded proper. Other than the LPG, many DAPs do not have proper PEQ. You can get PEQ with Android devices, and tried those as well, Didn't sound proper.

You can go on and on about how Audeze tech is all that. Cool.

I have a gentleman's agreement with Nic to not bring my data into this thread. At any case, I can find industry-standard graphs of the IEMs he listed so it's a non-issue.
 
Dec 7, 2017 at 9:09 PM Post #6,675 of 39,414
I guess my words got lost in translation at the start, I didn't say that the 1.5k peak is the only reason for the weirdness but rather is the reason for the unique weirdness that the 10/20/i4 exhibit.

That's what I am saying man. My post wasn't targeted at you because you clarified where you stood. But not everyone is on the same page as you. A lot of people seemed to be fixated on the response at 1.5 kHz, while the bigger elephant in the room is the lack of presence at 2-3kHz.
 

Users who are viewing this thread

Back
Top