What’s the significance of headphones that have a frequency range of more than 20,000 HZ?

This question feels like it has been answered many times. Nevertheless, after looking at the eight answers that have been answered so far, I didn’t even mention the key points.

First of all, let’s give a brief comment: let’s not say if you can hear more than 20,000 Hz. First, ask this headset with nominal frequency response of 20-20,000 Hz. Does it really make sound when it is close to 20 and 20,000 Hz?

So my answer is also from two aspects: human hearing, and the frequency of headphones.

Let’s first talk about the range of human hearing, the so-called “20-20 kHz”.

People who know are not necessarily acoustics experts, so start with the basics. The so-called Hz is the reciprocal unit of time t. That is to say, the unit of T is seconds, and if we take the reciprocal of seconds, we get the Hz, which is one second, that is, the number of times per second. So, 20Hz means 20 times per second. Sound is a kind of vibration, but every vibration has a period, that is, a frequency, the so-called 20Hz sound, that is, vibration 20 times per second.

The human ear is stimulated by the vibration of the outer air through the eardrum, which promotes the lever structure composed of three small auditory bones. It transmits the vibration form with constant frequency and increased amplitude to the oval window of the cochlea. The fluid in the cochlea disturbs the hair cells in the vibration, resulting in the voltage difference between the two sides of the basement membrane, thus electric current is generated and sent to the auditory nerve. Vibration is converted into electrical signals of different frequencies, which are fed into brain neural networks, and the pressure generated by vibration is released from another elliptical window on the cochlea. This is a completely hardware implementation process. Since it is hardware, it has linearity and frequency response curve, so it has the so-called audible range measured by experiment.

Maybe many people know that there’s something called an isoloudness curve:

The significance of this curve is to let us understand that the human ear is not a linear system (linear “line” means straight line, that is to say, if a system is linear, then its input and output functions in the coordinate system will be a straight line, that is, equal proportion output independent of frequency), and it also has different frequencies. Sensitivity. For example, in this picture, a red line represents a subjective loudness. On the curve of 60 phons (phon is the unit of psychoacoustics loudness, 0 phon is just inaudible, 100 phon is the maximum volume that the ear can bear), we can see that a 20-Hz sound is about 110 decibels, while a 16-kHz sound is only about 68 decibels. The ears think the two sounds are the same loudness.

The decibel value here is an absolute physical quantity. That is to say, to make us hear a 20 Hz sound as loud as a 16 kHz sound, the current value used to drive the sound unit of the headphone is different, and the amplitude of the diaphragm of the headphone is also different.

Maybe you would say that the difference between 16kHz and 20kHz is 4000Hz. How could there be such a big gap? Then there is another lesson to be learned: the human ear’s perception of frequency is not linear. It’s in logarithmic form.

If the frequency is high, we will feel the voice is high and vice versa. However, the difference in frequency is not linearly reflected in our perception of sound level. Example: We know that an octave has twice the frequency relationship. That is to say, an octave at 20Hz is 40Hz, while an octave at 1000Hz is not 1020Hz, but 2000Hz. In this way, the higher the frequency, the greater the gap between our perception of sound level. 20 kHz is only 1.25 times higher than 16 kHz. What is the concept of 1.25 times? It’s almost a big three degrees, that’s the distance between do and MI in the do re mi we sing.

Of course, for the piano, for example, the lowest frequency is A2, the frequency is about 27.5 Hz, the highest is c5, the frequency is about 4186 Hz. Sounds that are not within this frequency range can hardly produce the sense of musical melody. However, the appreciation of music can not be separated from the human ear’s recognition of voices up to about 20 kHz, because there is overtone in the voice, and it is overtone that determines the timbre of the instrument. For example, the following picture shows the frequency spectrum of the single tones g,, g, A and B on the violin:

As can be seen from the chart, the four different pitches look very similar in the distribution of frequency energy (similar timbre), while our ears have a very clear understanding of their pitch (different frequency points with the most concentrated energy). This is because the auditory system has a very high time resolution, and can also clearly distinguish the very fine time structure (TFS).

I wrote an article in my column:

Mengwu: Can we really hear a 21 kHz sound?

In this paper, it is pointed out that the so-called 20 Hz to 20 kHz auditory range is measured by monotone method. In recent years, some scholars have found that human ears have the ability to distinguish between high frequency content and non-high frequency content by comparing music materials, regardless of gender, age and listening experience. The research method is simple, which is the ABX method commonly used in the industry: the subjects click on the three buttons of A, B and X on the software interface to listen to the sound material they play, and then choose the same material as X in A and B. Usually the subjects were asked to do 18 times (of course, ABX changed randomly each time). If the number of times they did the right thing was more than 12 times, they could think that the probability of “mismatching” was very low. These high-frequency content can even reach 21 kHz. How to explain this phenomenon? To put it bluntly, it is: a 20 kHz sinusoidal monotone (because the sinusoidal wave represents a simple harmonic vibration, its physical characteristics determine that there is no overtone in the sound wave generated by the simple harmonic vibration, and there is only one frequency in the spectrum, so it is widely used as a single tone or pure tone in the experiment) you can’t hear, but the sound of a bell is ideally recorded and made into two. One version is ideal for high fidelity (headphones with frequencies over 20 kHz are needed here, because bells must vibrate at frequencies over 20 kHz), and the other cut off frequencies above 20 kHz. You can distinguish the two recordings! In fact, for the percussion of metal instruments with high frequency energy concentration, such as bells and cymbals, the cold tone of the high frequency part depends greatly on the part above 16kHz. When about 16 kHz begins to be cut off, we can usually hear a very significant change in timbre. And the part of these frequencies is what we call TFS.

It is precisely because the human ear is very sensitive to TFS that the audio system needs to ensure as much linear output as possible in the audible range, otherwise it is easy to be considered distorted. So, let’s move on to the next step.——

Headphones with a frequency range of more than 20 kHz are not meant to be cool .

As mentioned above, although the range of “frequency response” of our ears is not linear, the recognition of TFS at high frequencies is unambiguous. This also requires that the linearity of headphones in these frequency ranges must be very good, and these parts should be restored as authentically as possible.

So, whether the questioner or the respondent, seems to think that a headset is nominal 20-20000Hz, then its frequency response suddenly starts from scratch at 20Hz, then stays upright until 2000 Hz, and suddenly collapses at 20001Hz? I drew a picture of your ideas:

Whether you are studying liberal arts, science or engineering, may I ask you that the frequency response shown above is possible? Do you have such headphones??? If there is, I must recommend it to me. I will buy a pair if I break the pot and sell iron.

In fact, the frequency response of the existing headphones is as follows:

The three diagrams are respectively the frequency response transfer functions of two earphones and one speaker in 40 subjects with ear canal obstruction. Like this one in the middle, Baia Dynamics DT-990 is already a very good headset with good frequency response. The subjective hearing linearity is already excellent. From a linear point of view, we can see that there are obvious ruggedness and nonlinearity from about 4000Hz.

Another graph from the network can more intuitively explain the nonlinearity of the frequency response of headphones as a physical vibration implementer.

Assuming you are the manufacturer of this headset, when you test it and get the frequency response shown above, you can naturally label it as having a frequency response range of 20 Hz to 2000 Hz. But what about its linearity? Few manufacturers print it on their boxes. Of course, I haven’t seen a manufacturer who is really foolish enough to sell only the range with good linearity as the frequency response range. Of course, the professional provision of experimental equipment is excluded. In fact, professional audio equipment can even draw frequency response curves on nameplates.

Therefore, it is not difficult to understand why there may be headphones with frequencies ranging from 5Hz to 80000Hz. Because 5Hz is actually two octaves lower than 20Hz, and 800Hz is two octaves higher than 20000Hz. First of all, we should widen the frequency range by two octaves, which is not an exaggerated range in terms of physical properties from the overtone range of general sound. Second, even if we use this headset not to listen to music, but to listen to sinusoidal monotones, we can know from its widened frequency range that its frequency response curve must be broader than that shown above, which means that it is in me. Within the perceptible range of 20 Hz to 2000 Hz, its linearity is bound to be better, because its rapid linearity decay occurs near 5 Hz and 80 000 Hz on both sides, which is beyond my audible range.

I drew another picture: it’s just a simplified sketch. The blue curve represents the response curve of a headphone with a frequency response of 20-20 kHz and the red one is the response curve of a headphone with a frequency response of 5-40 kHz. We ignore the ruggedness in the frequency range and only look at the extremes on both sides: the red headphones are obviously more linear in the human audible range.

< strong > To sum up:

In a complex voice (especially music), TFS over 16 kHz or even 21 kHz can be clearly recognized by human ears and can be used to judge whether it is distorted or not. Complex sounds are exactly what most consumers listen to with headphones. This requires earphones to have excellent linear frequency response at 20 kHz or even higher frequencies. The nominal frequency response of

  • headphones between 20Hz and 20000Hz does not mean that their frequency response is linear in this range, especially at both ends. In fact, the closer to the extreme, the worse the frequency response. Even some earphones can only measure extremely weak responses at 2000 Hz, but they can still nominally reach 20,000 Hz.
  • Because of the physical properties of the headphone diaphragm, it must have a great attenuation at the extremum of both sides of the nominal frequency range (in fact, it should be understood conversely here: because there is attenuation at these two values, it is nominated that these two values are extremum). It’s like an upside-down plate – that’s the law of nature. Then, the wider the frequency response range, the larger the plate, and when it is greatly larger than my audible range, it may be possible to get a relatively better linearity at the extremes of both ends of my audible range.
  • Leave a Reply

    Your email address will not be published. Required fields are marked *