Today we’ll talk about CrinEar Reference, in-ear headphones worth of $350 on the manufacturer’s website or 23,000 rubles on Russian marketplaces. The in-ears are positioned as having ‘the most neutral sound in the world’, and this is, you know, a very bold statement.

What sound neutrality is, whether Reference are close to this very neutrality, and whether they actually sound good is what the today’s review will be about.

Yet another glance at the concept of sound neutrality and the target curves of headphones

Let’s remember the logic of reasoning about what sound we consider neutral. I’ve already written about it all in detail in the relevant articles.

TL/DR:

1. A neutral–sounding speaker is a speaker that, when measured by a microphone on the speaker axis in an anechoic chamber, provides a ‘flat’ frequency response, that is, the sound pressure at frequencies from 20 to 20,000 Hz is the same.
2. If we place this speaker in a prepared room (with a non-zero pulse decay time), then the measurement should result in the same smooth frequency response, but having 4 dB more low frequencies and 2 dB less high frequencies. The figures are approximate, and they depend on the room and other parameters. You can read more about all this, for example, in The Measurement and Calibration of Sound Reproducing Systems or in Sound Reproduction by Floyd Toole. In simple terms, the difference between the sound of a neutral-sounding speaker in an anechoic chamber and a more or less real room can be described by slightly turning the frequency response clockwise.
3. Studio rooms for sound processing and mixing are usually prepared rooms.
4. Thus, if a listener picks up a neutral-sounding speaker and listens to music on it in a prepared room, it is highly likely that they hear approximately what the sound engineer has heard in the studio.

The concept of neutral sound for speakers is very clear. There are no secrets here.

But there’s more to it when it comes to headphones.

Speakers are measured using microphones, but simulators of ear canals, auricles, heads, and even bodies (HATS – Head and Torso Simulator) are used to measure headphones. All this affects the recorded frequency response, especially in the 3 kHz range, where there is a significant emphasis when using the simulator. It’s important to note that we always hear this accent, and on speakers too, because we always listen with our ears, not with microphones. That is, if we take an artificial ear and measure neutral speakers with it, this emphasis will be present on the measurements.

And then, the search for a neutral target curve for headphones begins.

1. ‘Free field’ is a starting point. The frequency response is measured on a simulator head placed in front of a neutral-sounding speaker in an anechoic chamber.
2. Next comes the ‘diffuse field’. There is a room with a very high decay time, there is a simulator head, and the sound in the room is tuned so that it’s the same in volume at all frequencies and ‘comes’ to the simulator head from all sides at the same time.
3. In contrast to rather abstract concepts of free and diffuse fields (which admittedly don’t work very well), Sean Olive suggests going not by theory, but by practice and recreating in headphones the sound of neutral-sounding speakers in a good listening room. He takes the Harman reference room for listening (with decay time of 0.4 seconds), puts the simulator in the listening position, and measures the frequency response through artificial ears and auditory channels. This is how the Harman 2013 target curve comes in.

Thus, the concepts of free and diffuse fields are pretty theoretical constructs, whereas the Harman approach provides an answer to the question, “How can headphones transmit the sound of neutral speakers in a room prepared for listening?”

And here come a few smart people from headphones.com (Griffin Silver aka Listener, Andrew Park aka Resolve, and others), Corin Koh (aka Crinacle), and a few more enthusiasts who don’t like the Harman curve sound for in-ear monitors. Their key complaint is overly expressive high frequencies. Besides, the Harman curve doesn’t answer the question, “What are neutral-sounding headphones?”, but this question is interesting!

So, these guys do the following trick:

1. The numbers for the diffuse field are taken from the ISO 11904-1 standard (obtained on the heads of real people).
2. The B&K 5128 measurement rig is used.
3. In fact, the diffuse field curve, refined using ISO 11904-1, is calculated for the B&K 5128 rig, from which the contribution of the auricles is excluded because the goal is a curve for in-ears.
4. This curve is rotated clockwise by 1 dB per octave (by analogy with how the sound of speakers is transformed during the transition from an anechoic chamber to a real prepared room).
5. This curve is called JM-1 and declared neutral.

Then some other actions are performed to obtain the IEF Preference 2025 target curve, but it doesn’t matter at the moment.

What matters is that JM1 is positioned as a neutral target curve that is:

1. Mathematically constructed on the basis of a diffuse field.
2. Valid for IEC 60318-7 measurement rigs (for example, B&K 5128).
3. Applicable at low background noise levels (below NR15; this is a consequence of how the diffuse field is measured according to the standard).
4. I couldn’t find the target sound pressure level, so if you can, please help me.

I’ll also note the following fact here: neutral speakers in a prepared room are unequivocally recognized by listeners as sounding the best (again, I recommend Sound Reproduction by Toole to those who want to go over the issue). However, this is definitely not true for existing neutral target curves for headphones (Etymotic Target, Harman 2013, IEM Neutral 2023, etc.).

Hence the question: if anyone manages to elaborate a truly neutral target curve for headphones, will the sound of the headphones corresponding to it be highly appreciated by listeners, by analogy with speakers?

Okay, it’s time to round off the theory.

Let’s move on to the review.

All but the sound

Here is the very kit:

1. the earphones;
2. eartips (11 pairs);
3. a cable with replaceable connectors;
4. a soft case;
5. some papers.

No revelations were found in the kit, but thanks for so many eartips, anyway!

The case is simple, with a zipper, upholstered inside, without any meshes or pockets.

The cable is thick and does not bend very well. But the connectors are fixed using threaded nuts. They are angled, 3.5 and 4.4 mm.

Four types of eartips are included: ones with narrow sound ducts (shorter and longer ones), ones with wide sound ducts, and foam ones.

I can’t say anything particularly good or bad about them. They are quite convenient to use with this model of in-ears, although they lose out to more expensive alternatives, if we talk about my personal experience.

And here are the earphones themselves.

The shells are aluminum, three-component (the shell itself, the sound duct, and the outer plate). There is no external coating, so there’s nothing to peel off. The connector area is recessed into the shell.

The compensation hole is located on the backplate.

The meshes of the sound ducts are cut crookedly, and their diameter is less than necessary.

The earphones look boring, although they visually support the concept of neutrality. Reference feel like a tool, not like an ornate gadget that must impress you with its extravagant design.

The diameter of the sound duct is 5.7 mm in its narrow part and 6 mm in its wide part. Therefore, almost any eartips will do. For fans of a deep fit, I’d recommend something narrow and relatively short, like Whizzer ET100AB. All others can try ddHiFi ST35. Specifically for culture vultures and those who want to get the most (from my subjective point of view, of course) from the Reference sound, try Final Fusion-G one size smaller than your usual one.

I didn’t find any ergonomic nuances when using Reference: the headphones just normally fit into the ears (with suitable eartips, of course), don’t fall out, and don’t squeeze in unlikely places. The sound insulation is very mediocre, which is probably due to the internal design of the earphones. I can’t recommend them for portable use in a noisy urban environment in any way.

Reference use a five-driver scheme for sound reproduction: two dynamic drivers mounted coaxially opposite each other, two Sonion midrange armature drivers, and one more armature driver by Knowles.

The manufacturer claims a sensitivity of 104 dB. Self-impedance:

At 1000 Hz, we can see self-impedance of 13.8 Ω without a cable and 14.1 Ω with an original cable. That is, the cable itself accounts for 0.3 Ω of impedance.

The sound

Standard links:

1. the description of my rig is here;
2. the audiogram of my hearing is here;
3. articles on measurement theory and the whole shebang are here.

Frequency response of Crinear Reference with standard silicone eartips:

Since the earphones were being developed by using data obtained from the B&K 5128 rig and it’s not possible to transfer measurements directly from IEC711 to 5128, it will be appropriate to provide the Reference measurement graph for 5128:

Frequency response measurements with other types of standard eartips:

Foam eartips, as always, smooth out all narrow peaks and dips in the high frequency range, and eartips with wide sound ducts add a significant rise in the ultra-high frequency range.

In the case of Reference, it’s difficult to comment on the frequency response graph in any special way: it almost perfectly coincides with the JM-1 target curve rotated by -1 dB/octave. Let me remind you that this curve was transferred from the 5128 rig through the averaging of an array of measurements of various headphones, so the match cannot be completely accurate. The Reference graph differs from the target curve only at the very bottom, in the range of 20-100 Hz, where a smooth 1.5 dB accent is added.

Nonlinear distortion at 94 dB with the ‘Use harmonic frequency as ref’ option turned off and on:

At 94 dB of volume, we can observe 0.7% distortion at the maximum. Everything is fine.

Nonlinear distortion at 104 dB with the ‘Use harmonic frequency as ref’ option turned off and on:

At 104 dB, the distortion reaches 2.7%, which is a bit too much, and even falls at a frequency of 3.3 kHz. This result is not remarkable for the earphones in this price range.

Minimum phase response, group delay, and spectrogram in the ‘Burst decay’ mode:

Everything is fine on these graphs.

How do the Reference sound like?

The Reference sound is familiar, even recognizable because there are many earphones with a similar sound signature. We’ll talk about it later. However, at the same time, the sound is incredibly good and new because Reference still differ from all similar earphones. In my opinion, the sound of Reference is, yes, the most neutral of all I’ve heard, but at the same time the most diverting and balanced. After I picked them up, I sat down and listened carefully to about 150 tracks in a row. The feeling of ‘everything sounds right’ has not left me for about all these 7 hours. Sometimes my brain demanded an iota more bass, sometimes a little more top. At the same time, there was an infinitesimal overkill at the same frequencies in other tracks. And this is despite the fact that I listened to fundamentally different music, from the last album of Lost In Kyiv to Jazzmatazz from 1993, from dub to academic orchestrations, from Cradle of Filth to Black Chamber. The Reference’s witchery is that they are absolutely, stunningly, unconditionally omnivorous, which is what truly neutral earphones should be like, in my opinion. They are suitable for literally any musical content, without any exceptions. I’ve only added to my insight concerning this over the next month, during which I was listening to music via these in-ears for 3 to 4 hours every day and could not believe my own ears. The only thing I capriciously changed from time to time was the listening volume.

These are amazing earphones, I must say. They sound both full-fledged and natural and, at the same time, not boring, and the ear doesn’t get tired at almost any volume and term of listening.

I’m consciously not trying to describe particular frequency ranges because I don’t have any other words except that ‘everything sounds fine and right’. The same applies to the feeling of the sound width and depth; if they are in the track, you will definitely hear and feel everything.

The Reference tuning reveals two interesting points:

1. Judging by their reviews on the Internet, it seems to many people that Reference have too strongly articulated midrange frequencies. To my ears, there are just as many of them as there should be to give a feeling of natural sound delivery without the effect of the general ‘blatancy’ of the sound.
2. The 2.4 dB emphasis at 55 Hz is perceived by my ears in an interesting way: when using dense foam eartips, I’m ready to agree that it could have been removed (equalizer settings: [PK, 49 Hz, G=-2.4, Q=0.6] + [PK, 140 Hz, G=1.2, Q=2]) for greater neutrality. But when using silicone eartips, I would leave it on because eartips of this type still slightly skip the lowest frequencies.

Summing up my subjective impressions, I didn’t just like these earphones—their sound frankly delighted me. And it couldn’t have happened otherwise, as it follows from the comparisons below, as the industry has been revolving around the sound delivery like this for several years, and I personally praised the models that sound in a similar way. And now it seems that the last millimeter of this path has been made. Reference embody exactly ‘that very’ sound. Therefore, as a person who is well-versed in listening, I didn’t experience the wow effect from the sound of these earphones. Instead, I said to myself with satisfaction, “It’s official!”

As for the neutrality, these are really the most neutral earphones that I’ve ever tested.

Comparisons

I’d like to compare Reference not with other models of similar value, but with those ones, the sound of which I defined as ‘close to neutral’:

• Hisenior Mega-7;
• Moondrop Meteor;
• Moondrop Quarks 2;
• Myer Audio SLIIVO SL224;
• TRI I3 MK3;
• Truthear PURE;
• Twistura Sigma.

Frequency response of Reference compared to Hisenior Mega-7 (review):

The Mega-7 sound is a V-shaped variation of the Reference sound delivery, probably the best one I’ve ever heard. It has the same smooth top, the same magnificent middle, and a ‘fast’ but weighty subbass, but everything sounds funnier and brighter, broadly speaking. To consider this a pro or a con or is a pure matter of taste.

Frequency response of Reference compared to Moondrop Meteor (review):

The Meteor’s sound is devoid of those weighty low frequencies that Reference can boast of. That’s why Moondrop earphpones sound ‘colder’ or ‘deader’, if you will. Their sound delivery is perceived to be more analytical and saturated with conventional details. To my ears, Meteor is the only model that can raise the question of whether Reference are really neutral.

Frequency response of Reference compared to Moondrop Quarks 2:

Comparing models with a twenty-fold difference in price does not look reasonable. However, Quarks 2 are not lost on the background of Reference. Yes, there is some overkill with upper frequencies, and the 3-10 kHz segment is not so smooth by ear. But damn it!

Frequency response of Reference compared to Myer Audio SLIIVO SL224 (review):

This is another variation on the theme of well-implemented metatuning. The sound is close to Reference, but has significantly expanded upper frequencies. Those who lack ‘detail’ and ‘air’ in the sound of the today’s review’s hero can take a closer look at SL224. However, the reverse of the coin is the undisguised excessive sound sharpness on some tracks.

Frequency response of Reference compared to TRI I3 MK3:

I’ve repeatedly praised the I3 MK3 sound, and I’ll praise it again: they’re not neutral compared to this review’s hero, but they’re expressly close to it. The key difference is a smooth, neat 2 dB emphasis on the upper midrange, as well as an extended upper frequency range giving additional volume to the virtual sound stage, which makes some music tracks sound wide and airy, while on others the I3 sound clearly lacks ‘anger’, but acquires ‘blatancy’ instead.

Frequency response of Reference compared to Truthear PURE:

“Why buy Reference when you have PURE and an equalizer?”, the above charts seem to ask us. Yes, the PURE sound delivery is similar to Reference, although PURE seem frankly ‘warm’ compared to the latter, and they ‘mumble’. Therefore, their sound is nothing more than ‘similar’. They are different enough to never confuse them by ear. And while it’s easy enough to adjust the lower frequencies of PURE using the equalizer, it’s unlikely to achieve the smoothness of the upper frequency range that Reference have.

Frequency response of Reference compared to Twistura Sigma:

+5 dB at 3.3 kHz is not as scary as it seems. Sigma sound brighter, more lightweight, and their high-frequency range loses to Reference in smoothness and balance. Just don’t forget that the Sigma’s price is some $55.

Summary

What questions should we ask after the release of CrinEar Reference?

• Can Reference really be considered neutral earphones?
  Objectively, taking into account the theoretical basis that gave rise to the JM-1 target curve rotated by -1 dB/octave, as well as taking into account the proximity of the Reference frequency response to this very curve, it’s 95% yes. I leave 5% for a detailed answer to the question of whether it’s valid to extend this turn from speakers to earphones ‘just because’. It still requires further explanation. Subjectively, having had the opportunity to listen to several thousands of headphone models and having more than 60 different pairs of them on hand, I personally tend to rate the Reference sound as the closest to what my brain perceives as neutral.
• Are these good earphones for their $350?
  From my point of view, yes, absolutely. I’ll say more: if I were asked to get rid of all the earphones in my collection by choosing the only ones, I’d settle on Reference. I’ll also assume that for many people this model will become the last one in a long search for ‘the right sound’, especially when it comes to listening to multi-genre compositions.
• What will Reference be bad for?
  From my point of view, for outdoor use: their intelligent low frequencies get ruthlessly masked by urban noise with their frankly weak sound insulation. And I’d recommend a different frequency response tuning for competitive games, too.

As for the disadvantages, there are some of them: a number of quality control problems (crooked mesh design, loose connectors, faulty dynamic drivers, and significant channel imbalance – look on Reddit if you’re interested), as well as rather high distortion at about 3.3 kHz at high volume.

The year 2026 has brought very few (so far, at least) interesting in-ear models. It seems that after a rapid 7-year development (from 2018 to 2025), when manufacturers were looking for a ‘new sound’, the industry has finally entered the phase of finding ways to reduce the cost of reproducing previously obtained results. As part of this approach, CrinEar Reference is a final chord. It’s a decisive end put to the discussion of what the sound of good versatile in-ear monitors should be like.

If we look at it that way, it gets a little sad.

Does this mean that we are dealing with sort of ‘the end of the in-ear sound’? Of course, not. Human ears and brains are different, materialism is insatiable, so the debate about the best headphones will last forever, and the industry will keep fueling this discourse with the release of new and new models.

Besides, it would be interesting to see and hear the final version of the latest Harman curve, wouldn’t it?

To buy or not to buy: you bet your ass, go and buy!