What is DETAIL in IEMs?

Introduction

We’ve all seen terms like ‘detail retrieval’, ‘true detail’, and ‘resolution’ thrown around by audio reviewers and hobbyists alike. But what does it actually mean when someone says one headphone is more detailed than another? It may seem straightforward, but can you take someone’s opinion at face value - can you trust someone - when they make this type of statement?

In this article, I’ll be breaking down my opinion (read: based on both supposition and fact) of what qualifies the idea of ‘true detail’ and how we can correlate it back to frequency response. A lot of the ideas I’ll cover in this article can also be applied to the way I believe we perceive other qualities of sound like soundstage and dynamics.

No two listeners hear exactly the same, so it's important to get your own ears on something.

We all hear differently

First, we need to establish that everything starts back with our hearing - which is different for everyone. This difference can mainly be attributed to two factors: HRTF (more on this later) and the audible spectrum.

What is the audible spectrum? We measure sounds by their frequency and loudness. The average person is born able to hear a range of frequencies from 20Hz to 20kHz; however, this ability diminishes with age and damage to the tiny hairs in our ears that pick up sound waves. This is why older individuals might struggle to hear high-pitched noises like the whine of a hospital’s machines.

Did you know? Many animals can hear much higher than humans!

Musical content in real-life has a broad spectrum that can fall outside and inside this range. When a cymbal strikes, the bulk of the sound is under 10kHz; however, the end of the note can carry all the way up to 20kHz, and even higher than we can hear. On the opposite end, some drums can go under 20Hz and are more ‘felt’ than ‘heard’.

Source: Independent Recording Network; see the interactive chart at this link, https://alexiy.nl/eq_chart/

But many IEMs can’t capture the entire audible range to begin with.

Generally, IEMs are most capable of reproducing the midrange frequencies like the sound of a person’s voice. They struggle with reproducing the extremes of the audible frequency spectrum, particularly upper-treble and a sense of airiness. IEMs that cover these extremes are considered to have good (bass or treble) extension.

On a frequency response graph, we would associate this with a ‘roll-off’ in either the lower-bass or upper-treble. Some IEMs might technically reach those frequencies but they’re not loud enough for us to hear. In the example above, note the strong drop to the purple line past 10kHz.  

When there is a disconnect between what we hear in real-life and what is capable of being produced by an IEM, we associate this with a loss of detail. For example, if an IEM is only capable of producing sounds under 10kHz, but a song has an instrument that goes up to 15kHz, that instrument will usually sound muffled to us. Or perhaps we’d say an IEM is missing the ‘rumble’ of the bass if the IEM rolls off under 50Hz.

You usually don’t need to have heard specific instruments in real-life to tell whether an IEM has good extension either! This is because in everyday life, we hear everything up to the limits of our audible spectrum. Take something as simple as a person’s voice or the snap of a finger. The bulk of these sounds is contained in the midrange frequencies which are lower and reproducible by most IEMs.

But this is just the threshold for audible intelligibility. As research and recording equipment has improved, it’s become clear that the highest frequencies, up to 20kHz and higher, are part of these simple sounds too. This is especially true of female voices and fricatives (certain, often sharp consonances). Assuming we are someone with average hearing, we instinctively feel like something is missing if an IEM can’t produce these frequencies.

On the other hand, let’s assume that we have two IEMs. They sound identical except one has extension up until 20kHz and one rolls-off after 10kHz.

This is a good example of where individual audible spectrums will have a dramatic influence on perception of detail. Listeners with average hearing will hear the IEM that rolls-off after 10kHz as being muffled and lacking detail. But let’s say we have an older listener, someone who can’t hear over 10kHz. In this instance, the lack of energy over 10kHz wouldn’t matter at all! He’d likely find no difference in detail between the two models.

What other implications does this have?

Anecdotes of listeners purchasing a new IEM and hearing tiny details they’ve never heard before are quite common. Heck, I’ve had this experience myself and I’ve often associated it with the idea of detail retrieval, or an IEM’s ability to bring out information from a track.

The Empire Ears x A&K Novus is probably the most recent IEM to have given me this impression.

However, assuming an IEM has adequate extension on both ends of the spectrum, it’s more common that the detail was always there and you always heard it. You just never noticed it! And this usually becomes apparent when you actually A/B the track between the new IEM with your old IEM; you’ll consistently pick up that new detail with the old IEM.

A common reason why we suddenly hear these ‘new’ sounds and think something is more detailed is the placebo effect. When we get something new, we expect it to sound better, so maybe we listen more closely than usual.

This experience can also be attributed to auditory masking.

A simple example of this is too much bass. When an IEM is tuned with too much bass, it begins to skew the frequency response and make the midrange and treble sound quieter (further away) by comparison. This can make it difficult to discern details in the midrange and treble even though they are technically still there (just quieter).

The Elysian Diva’s bass pot switch allows us to illustrate this neatly. Nothing but the bass response is changed between the settings. But the maximum bass setting will naturally sound less detailed because the midrange and treble will sound further away in the mix.

Due to the masking effect, it's possible that some information could be inaudible or less audible when there's a sharp peak in the treble, as the boosted frequencies could mask the others. Or comparatively, a sharp dip can direct attention to other parts. The way we hear frequency response as sound is a collective where small adjustments can have much bigger effects on the end sound.

In the case of the 64 Audio U12t, the 5kHz peak is balanced out by the 15kHz peak, so it doesn’t actually sound particularly peaky. The delta at ~8kHz leads to more emphasis on the ‘air’ frequencies and the peak at 15kHz.

Possible hot take: coloration is detail

If we follow what reviews say, we’d have the impression that the resolution of IEMs keeps increasing every year. But generally speaking, the idea that resolution can keep increasing is, in my opinion, a misleading tactic that is used to drive sales and propagate FOMO (fear of missing out).

In the current IEM landscape, there are many IEMs with excellent extension on both ends of the frequency spectrum. I have heard many of these IEMs and, overwhelmingly, it’s usually a case of “this sounds different” rather than “this actually sounds more detailed” when it comes to the most expensive ones.

There is a correlation between more expensive IEMs and sounding more detailed, but in my experience it's mostly based on these more expensive IEMs having unique tunings.

When an IEM is more expensive, it’s usually because it uses more drivers (the miniature ‘engines’ that create sound in an IEM). Similarly, a more expensive IEM can mean more research and development into advanced tuning techniques like un-lidding drivers, using tubing to carry the high frequencies, or manufacturing resonance chambers in the IEM's shell. Together, these afford a manufacturer increased flexibility with tuning and the ability to achieve superior high-frequency extension. 

The Subtonic Storm has a series of small, carefully tuned peaks in its treble response that I believe correlate to what I perceive as detail.

Our perception of detail is not solely based on what we hear in real-life, or what we’d define as a ‘neutral’ frequency response when it comes to tuning an IEM. Many listeners also appreciate tasteful colorations - the easiest example is a strong, controlled bass shelf. 

Tasteful is the key word. Tasteful adjustments to frequency response can lead to an increased sense of detail retrieval by accentuating sounds that we wouldn’t normally hear. To the point above, an IEM can sound like it has more detailed bass simply because the bass is more present in the tuning. 

If the adjustments are too liberal, though, sound begins to come across artificial or what many audiophiles describe as ‘plasticky’, ‘metallic’, or ‘muddy’. All of this also ties into the concept of auditory masking I covered above.

The Vision Ears VE8 is a popular IEM with lackluster treble extension that many listeners have reported as being quite detailed to them.

This also explains why a subset of listeners can hear some IEMs with poor treble extension as being more detailed despite it being contrary to what they’d hear in real life! A lack of treble extension creates the perception of graininess - what some might also call texture - to the end of notes. Some listeners perceive this as added detail rather than a lack thereof.

“I can read the resolution from a graph”

This is certainly possible with years of training, if you possess golden ears, and if you have mastered the skill of audiation...

In all seriousness, this is a common misconception. It is true to a very small extent and a ‘read’ usually falls into one of two buckets: 1) this IEM will likely not have good resolution because it has very poor extension and is extremely removed from conventional tuning or 2) we don’t know how the resolution of this IEM will sound.

By no means are either of these ‘reads’ concrete. Leaving aside the ideas of auditory masking and coloration corresponding to detail, this uncertainty can be attributed to HRTF.

HRTF stands for head related transfer function. When sound waves originate from a source, we don’t all receive the exact same sound at our eardrums. Instead, by the time they reach our eardrum, the sound is transformed by the shape and direction of our heads, ears and other physical geometry. IEMs can eliminate some of these factors because they are inserted directly into the ear, but our brain still expects the exact transformations that our anatomy makes—our unique HRTF—even if the IEMs don't interact with these external anatomical factors.

Bear in mind that HRTF and the audible spectrum are also continuous values that can change. You put on some weight? Your HRTF possibly changed. You got older and went to a dozen concerts without hearing protection? You’ve possibly experienced some hearing damage at this point and your audible spectrum has changed.

The takeaway is that it is impossible to read resolution from a graph with certainty because these are moving values that would affect individual perceptions of resolution. 

In the end, ‘true’ detail doesn’t exist.

If you take anything away from this article, it should be that perception of detail is highly subjective.

Ever heard the saying that there’s no right answer, but some answers are better than others? Similarly, there are IEMs that should be more detailed than others and I can tell you with a greater degree of certainty, based on my experience, whether some IEMs are more detailed. But I’ll never be entirely correct for every listener - only myself. Even for myself, I’m not afraid to admit that I often struggle to say between two IEMs which is more detailed!

And if this sounds like something out of a “you write your own story” pep talk, that’s because it kind of is. Don’t blindly trust the reviews. Don’t let other people listen for you. You need to decide for your own ears what qualifies detail and, hopefully, this article has lent more insight into why.