Why Headphones Shouldn’t Aim for a Flat Frequency Response

A ruler-flat frequency response on headphones won't sound good in real life.
A ruler-flat frequency response on headphones won’t sound good in real life.

We independently review all our recommendations. Purchases made via our links may earn us a commission. Learn more ❯

Headphones with a perfectly flat frequency response don’t exist for a reason.

One common misconception is the idea that headphones should aim for a perfectly flat frequency response.

At first glance, this notion seems logical. Besides, shouldn’t our gear faithfully reproduce sound exactly as it was recorded? But when it comes to headphones, this approach misses the mark entirely.

The reality is far more complicated and interesting. In fact, there’s a surprising lack of consensus in the audio industry about what the ideal headphone response should be. This leads to uneven sound quality standards, which audio expert Floyd Toole calls the “circle of confusion.”

Why Flat Doesn’t Work

A completely flat frequency response curve.
A completely flat frequency response curve.

The problem lies in the differences between how we experience sound from speakers versus headphones.

When I listen to music through speakers, the sound interacts with my room, bounces off walls, and reaches my ears from a distance. Because of this, my head and outer ears naturally shape the sound before it even hits my eardrums.

Headphones, on the other hand, skip all of this.

They deliver sound directly into your ear canals. So, if the frequency response were completely flat, you’d be left with an unnaturally sterile listening experience.

Let’s break this down a bit more.

In a normal room, speaker playback has a slight downward tilt in frequency response (about 6-7dB less treble than bass). This happens because room reflections absorb more high frequencies than low ones.

Flat headphones don’t account for this, so they may sound too bright in comparison.

Then there’s the issue of bass perception.

With speakers, you don’t just hear low frequencies – you feel them with your whole body. Headphones can’t replicate this physical feeling, which is why they often need extra bass to sound balanced.

It’s worth noting that truly flat headphones are extremely rare, if they exist at all.

Even headphones sold as “neutral” or “reference” usually have planned differences from a perfectly flat response. These tweaks are designed to make up for the factors mentioned above and create a more natural listening experience.

A pair of headphones with a ruler-flat response may seem impressive on paper. But, the actual listening experience is likely to be a letdown. The sound will feel lifeless and disconnected, lacking the engaging quality you’ll get from a speaker setup.

The idea that flat might not be best for headphones has led many audio engineers and researchers to study how we hear sound through these personal listening devices.

Understanding How We Hear

The Hearing Mechanism (From: Britannica)
The Hearing Mechanism (From: Britannica)

Our hearing system has evolved and fine-tuned to process sounds in the real world.

A key part of this is the head-related transfer function (HRTF), which is basically like a personal sound filter created by the shape of our head, ears, and torso.

When sound hits us from different directions, these body parts cause subtle changes in frequency and timing. Our brain then uses these information to locate sound and give it a sense of space.

The outer ear, or pinna, plays a particularly important role in this.

Its complex shape creates resonances and reflections that boost certain frequencies, especially around 3 kHz. This boost isn’t an accident – it makes us hear important sounds like human speech better.

Yet, there are still individual variations in the exact frequency and magnitude of this resonance depending on our ear shape, size, and design.

But there’s more to it than just the HRTF. Our brain uses several clues to find sounds:

Interestingly, we use these clues differently depending on the sound’s frequency. Below about 1kHz, we rely more on ITD and IPD. But above 1kHz, ILD becomes ILD takes center stage.

This crossover point relates to the wavelength of sound becoming smaller than the size of our head.

Headphones need to account for these natural processes to sound “right” to our highly evolved hearing system.

Historical Approaches and Their Shortcomings

Diffuse-field and Free-field (FF) frequency responses measured with HATS. (From: Measurement Apparatus and Modelling Techniques of Ear Canal Acoustics)
Diffuse-field and Free-field (FF) frequency responses measured with HATS. (From: Measurement Apparatus and Modelling Techniques of Ear Canal Acoustics)

The audio industry has long recognized that headphones need a different approach.

That’s why they have developed early attempts to address this including the diffuse-field (DF) and free-field (FF) response targets.

The DF response tried to simulate sound coming from all directions equally, like in a perfectly echoing room.

Researchers like Hammershöi and Möller measured this by placing a dummy head with microphones in a highly reflective room.

The FF response, on the other hand, tried to mimic listening to a single speaker in a room with no echoes, measured with a similar dummy head setup.

While these were steps in the right direction, neither quite hit the mark. One big problem was how hard it was to make headphones that could exactly copy the complex resonances and phase effects measured in these tests.

I’ve listened to headphones tuned to both DF and FF targets. While they sounded more natural than flat headphones, something was still off.

The DF ones often seemed a bit bright and thin, while the FF models could sound uneven in the treble. Neither fully captured the experience of listening to great speakers in a good room.

However, these are generalizations, and the actual sound can vary significantly between different implementations.

Even Our Headphones Need a Little Boost

The instrument frequency chart shows that a lot of instruments lay heavily on the low frequencies (below 700 Hz)
The instrument frequency chart shows that a lot of instruments lay heavily on the low frequencies (below 700 Hz)

This brings us to an important point: certain frequency boosts are not just okay in headphones, but beneficial.

Remember that 3 kHz boost from our ear’s natural resonance? Good headphones should copy this to sound natural.

Another important factor is bass. When I’m at a concert or listening to my home speakers, I don’t just hear the bass – I feel it. As mentioned before, headphones can’t copy that physical feeling, so a slight boost in the low end helps make up for it.

It’s not about exaggerating the bass, but rather making it sound full and satisfying without the aid of sound waves actually hitting my body.

Some research, like the Harman studies, has shed more light on these preferences. They found that most listeners prefer a bass boost below about 300Hz in headphones.

But, they also discovered different listener groups. Some, particularly younger males, preferred even more bass. Others, often older listeners and more females, preferred slightly less bass and a bit more treble.

Personally, I’ve found that headphones with a well-implemented bass boost and a natural-sounding treble are the ones I use most often. They manage to sound both accurate and enjoyable, striking a balance that flat headphones simply won’t achieve.

The Challenge of One-Size-Fits-All Targets

While we can establish general principles for good headphone sound, creating a single, perfect target response is hard. Why? Because we’re all unique.

The shape of my ears is slightly different from yours, which means our individual HRTFs vary.

Some people prefer a bit more treble energy, others a touch more bass.

This variability is why headphone manufacturers and researchers continue to refine their approach.

One of the most important developments in recent years has been the Harman target curve. Developed through many listening tests by Sean Olive and his team at Harman International, this curve tries to copy the sound of high-quality speakers in a good listening room.

The Harman target is interesting because it’s not based on theoretical ideals, but on what listeners actually prefer. It usually includes a slight bass boost below 200 Hz, a mostly flat midrange, and a gentle treble emphasis.

And, based on their research, this curve was preferred by a majority of listeners, regardless of age, gender, or listening experience.

Yet, it’s not universally accepted as the ideal. Even the Harman target isn’t a one-size-fits-all solution.

There are several variants of Harman target curves, owing primarily to the differences in listener preference (within a constrained range) as well as seal between different formats (circumaural, versus in-ear monitor, for instance). (From: Harman)
There are several variants of Harman target curves, owing primarily to the differences in listener preference (within a constrained range) as well as seal between different formats (circumaural, versus in-ear monitor, for instance). (From: Harman)

As previously mentioned, their studies revealed distinct listener groups with slightly different preferences. So, it’s clear that there’s a challenge of creating a universal target.

Measuring individual HRTFs is complicated. It often involves tiny microphones placed in the ear canal or advanced 3D scanning techniques.

And, even if we could easily measure everyone’s HRTF, there’s another problem: headphone placement. Small changes in how headphones sit on your ears can cause big variations in high-frequency response.

That’s why, some researchers have suggested other methods to address these challenges.

David Griesinger, for example, suggested using pink noise and loudness comparisons to create personalized target curves.

His method involves comparing the loudness of pink noise at different frequencies (every one-third of an octave) to a reference frequency of 500Hz, first with a known flat speaker and then with headphones. The difference between these measurements creates an individual calibration curve.

On the other hand, Siegfried Linkwitz developed a technique using sine wave sweeps to identify and equalize resonant peaks.

His approach involved using continuous sine wave sweeps through the entire audible frequency range to identify sudden changes in loudness, which indicate resonances. He then equalized these peaks to achieve a more uniform perceived loudness across frequencies.

In my experience, the best headphone makers use a target curve like Harman’s as a foundation, then fine-tune the sound based on careful listening tests. I realized that, while we can use science to get close to ideal headphone sound, there’s still an art to those final adjustments.

Embracing a Headphone-Specific Goal

TL,DR: The goal shouldn’t be a flat frequency response, but rather a carefully shaped response that considers how we hear in the real world.

So, the next time you’re shopping for headphones, don’t be swayed by claims of perfectly flat response. Instead, look for models that aim to sound natural and engaging. Better yet, trust your ears.

The best headphones are the ones that disappear, letting you connect with your music on an emotional level.

While I’ve focused on the challenges of headphone design, they do have some advantages over speakers. They’re less affected by room acoustics, can provide excellent isolation, and can deliver high-quality sound in a portable package.

Leave a Reply