How to read the frequency response of headphones correctly? What influences it, what frequency response is perceived as flat, what is the Harman curve? Finally, how to correct their sound in a minute and without an equalizer? More on this later.
Headphone frequency response problem
With the frequency response of the speakers, everything is simple – the smoother it is, the more neutral the sound will be in the end. However, in order to obtain maximum neutrality, you will need to drown out the room with a few cubic meters of mineral wool. Not an option for everyone.
And here headphones come to the rescue, which take the influence of the room out of the brackets. However, if you look at their frequency response, then almost always on the graph there will be a roller coaster with giant humps and dips. For example, below is the frequency response of the Beyerdynamic DT 990.Wired headphones Beyerdynamic DT 990 (250 Ohm) grayIn stores: in 5 stores
Why, despite the frequency response curve, are they used by many sound engineers to mix music? How to read their frequency response correctly and correctly predict which model is best suited to your tastes? And most importantly, how to correct the sound?
If you start to understand, then with headphones everything turns out to be much more complicated than with speakers. Their timbre balance is influenced by a bunch of things – from the design itself and the ear pads to the structure of the auricle, the volume level of listening to music, listening conditions and even the age of the listener.
Influence of the head and ears
If you sit a person in front of a sound source, then the audio waves will bend around the head and shoulders with distortion. Researchers have found that there is a whole set of body-related distortions called the Head-related transfer function (HRTF):
1. The influence of the head itself. It drowns out the high frequencies, but does not create problems for the mids and lows.
2. Shoulders and neck. They also create acoustic shade.
3-4. Auricle. It is not just that it has a complex shape – it is easier for a person to localize sound this way. However, the shape of the ears distorts the frequency response of the sound wave – it raises the upper middle and smoothes the highest frequencies.
5. The auditory canal. Together with the earpiece, it forms a closed system in which resonances appear.
HRTF will be slightly different for the left and right ear. The values will change depending on the direction the sound is coming from (waves blowing directly into the face behave differently than waves blowing sideways). For covering headphones, you need to simulate the influence of the head and shoulders, for in-ear headphones, you will have to add an imitation of the influence of the auricle.
All these studies resulted in compensating frequency response curves taking into account HRTF. They are sensitive to the position of the source. The speakers are usually placed in front of the listener and form a triangle with him. Headphone drivers shoot directly into the ear and form a straight line with the listener. Therefore, there are several compensating curves, depending on the type of measurements:
If you put a person in front of a column that blows directly into him, you get the Free Field, FF curve .
If you put a person in a room with several speakers pointing in different directions, but not at the listener, you get a Diffuse Field, DF curve .
In total, a sine with a flat frequency response on the approach to the eardrum turns into one of the graphs above. The frequency response of the headphones is measured on a mannequin with ears with built-in microphones. Therefore, to get the correct measurements, you need to subtract the effect of the artificial head from the raw graph using one of the curves above. The smoother the final graph, the smoother the headphones will be.
For example, one of the most sound-neutral Sennheiser HD 600 headphones, used in many studios as a reference. On the left is a raw chart without compensation, on the right – with compensation.
Firstly, all this introduces obvious confusion in understanding the frequency response graphs – it is not always indicated how measurements were taken, and whether compensation was applied. Secondly, even on the compensated graph, one can see a 1-2 kHz dip and irregularity at the top. Failure is an imitation of the influence of the shoulders and torso, and irregularity is the resonances that occur in the closed space formed by the ear and earpiece. A slight irregularity is the norm, it is not felt by ear.
Sennheiser HD 600 Wired Headphones grayThe influence of preferences and the Harman curve
Experimenting with HRTF misses an important thing – taste preferences. In the end, sound neutrality and balance are subjective things. Therefore, at the beginning of the 2000s, engineers Todd Welty and Sean Oliver Harman, since the tenths, have been conducting research on groups of listeners to find out which frequency response will be perceived as the most neutral and balanced, but taking into account the preferences of the majority.
For the experiments, the method of double-blind listening with different headphones was used. The listeners did not know their model or what they looked like. They even wanted to freeze their ears and whiskey to exclude tactile sensations, but the lawyers did not allow it. Based on the preferences of the listeners, an averaged curve was derived, which now bears the name of Harman. It is slightly different for in-ear and in-ear headphones.
64% of listeners preferred headphones with a curve close to the obtained one – it seemed to them the most even, neutral and balanced.
15% of listeners, mainly young and middle-aged men, preferred a 3-6 dB boost in bass, starting at 300 Hz relative to this curve, and a 1 dB boost above 1 kHz.
21% of listeners, most of whom are women and people over 50, would prefer, on the contrary, to turn down the bass by 2-3 dB, but they would also add 1 dB to everything that is above 1 kHz.
Since the publication of the research, many manufacturers have released models tuned to this curve. Of the in-ear headphones, these are JBL Live 200, 500, and 650, Samsung Galaxy Buds, JBL Reflect Flow, of the over-ear headphones, the most famous are AKG N700 NC, K361, and K371.
Frequency response of Galaxy Buds + tuned along the Harman curve
TWS Headphones Samsung Galaxy Buds + BlackDisputes about whether to focus on this curve when creating headphones do not subside, because the frequency response of some models known and loved by listeners strongly disagree with it. Some people think that the bass is too high in the curve, and a drop of more than 10 dB in the middle frequencies strongly colors the sound.
Influence of psychoacoustics
Our hearing is not linear and is most sensitive to the 1-5 kHz region, and the least sensitive to the lowest and highest frequencies. For example, a sound at 3 kHz with a volume of 20 dB will feel the same as a low frequency hum at 60 kHz and a loudness of 50 dB. In other words, at low volumes, bass and treble are less picked up. However, this sensitivity changes with the volume of the sound. At a volume of 100 dB, the perception becomes almost linear. It should be noted here that, according to the WHO , headphones are usually listened to at a volume of 75–105 dB.
Research on this topic took place back in the 30s of the last century, the measurements obtained are in the form of graphs – curves of equal loudness. These should be taken into account when choosing headphones – if you like listening loudly, then the bloated bass of some models will be even more bloated. Conversely, the lack of bass in open studio headphones will be compensated for by high volume.
Another element of psychoacoustics is the effect of masking one sound with another if they are both at the same frequencies. For example, the low-frequency hum of the train will drown out the bass and kick. Because of this effect, bass boost is common in outdoor and urban headphones with poor sound isolation. The better the soundproofing, the less bass boost will be needed. The best way to deal with the problem is the active noise cancellation system.
Bluetooth headset Sony WH-1000XM4 black
The headphones and ear create a closed system and act like a subwoofer. The more airtight the formed space, the more low frequencies can be expected due to the occlusion effect. – the bass accumulates in a closed space. The good news is that the influence of the design allows you to change the sound quite dramatically with a simple modification – changing the ear cushions.
Over-ear headphones can be open or closed. The latter are less comfortable when worn for a long time, but they have excellent sound insulation and, as a rule, a lot of low frequencies.
Bass and treble can be increased or decreased by changing the ear pads. For example, if open headphones with velor ear cushions lack low and high frequencies, you can replace them with leatherette ear cushions. They are more sealed, and bass will accumulate better in an enclosed space. The same with high – velor absorbs them, and leatherette reflects, so the headphones will sound brighter.
The thickness of the ear cushions also matters. If you replace the old ones with new ones from the same material, but thicker, the headphone driver will move away from the ear by a few centimeters. This will reduce the midrange, which will feel like a boost in bass and treble.
Finally, the position on the head has a big influence on the sound of the over-ear headphones. Some models are less sensitive to this moment, but some will sound much worse if the driver is not directed strictly to the ear. Due to the sensitivity to position, there is often a large variation in the frequency response measurements of the same headphone model.
As a rule, the frequency response of in-ear headphones has a more pronounced roll-off at high frequencies, simulating the influence of the auricle. The material of the ear cushions also matters here:
Silicone gives a balanced sound. There are double and triple silicone ear cushions, which provide better sound insulation and enhance bass. The size affects the tightness and the amount of bass.
Polyurethane memory foam. They are more sealed, better soundproof and more comfortable. They have good air-tightness and enhance low frequencies, eating out high frequencies a little. However, they cannot be washed, so they are not as durable as silicone.
Lyambda E1-BK replacement ear tips set, black
Recently, in-ear headphones such as IEM – In Ear Monitors, in-ear monitors are gaining popularity. Musicians use these to hear themselves on stage, but there are also models for consumers to listen to music. The difference is in frequency response – headphones for musicians can strongly color the signal, for example, accentuate the 2-5 kHz range to highlight vocals. Before buying such, it will be useful to know their frequency response.
Wired headphones FENDER IEM Nine 0 silver
Frequency response graph can be useful when choosing headphones. It should always be remembered that the variation in measurements is quite large, and the shape of the ears is different for everyone. Therefore, such changes are relative – for some, bright headphones will be dull. Finally, the frequency response only speaks about the timbre balance, but does not tell about the detail, dynamics, transients and other important parameters of the sound of the headphones.
Of course, the surest way is to listen to everything yourself, but there is not always such an opportunity. Therefore, much more useful than the frequency response curves will be tests of headphones on a breadboard with the help of music tracks and a direct comparison of several models with the original recording. So you can hear with your ears what this or that model does with the sound.