How do I select my headphones?
If we have a look at the current offer of professional headphones, we see that there are many different models, designs, impedances etc. How can I select a headphone from this offer that meets my applications and requirements best? What should be considered primarily? Which properties have what influences?
Let us start with the transducers. Headphones need transducers to convert the audio signal into air movements (sound). There are many different kinds of transducers; four different versions are described below:
The transducer is made of two grid-like electrodes with a diaphragm thickness: < 2 micron) between them, opposing each other directly. Depending on the audio signal the electrodes are always opposingly charged and pull or push the diaphragm depending on their charge. Electrostatic transducers are appreciated by audiophiles due to their extreme precision and low distortion factor. Disadvantages include the high operating voltage, the mechanical sensitivity and the relatively high cost price.
Another transducer is the almost forgotten orthodynamic transducer. The transducer consists of two grid-like ferrite magnets that are face to face in a certain distance with a concave diaphragm between them. The diaphragm consists of two layers between which the coil (flat) has been mounted. The diaphragm (coil) is moving between the two magnets (depending on the audio signal), the air is moved and sound is produced. Orthodynamic transducers also sound very precise and have a very low distortion factor. In the 80’s this kind of transducer was widely used, but today it is rarely found. The biggest disadvantage was that the small forces of the used magnets could only achieve a small sound pressure.
The third transducer is also the transducer mostly used: the dynamic transducer which is set up like a loudspeaker: on the rear of the diaphragm is a ring coil (sometimes it is also called moving coil), which moves in an air gap of a permanent ring magnet. This transducer provides a high reproduction quality, is mechanically very robust, needs only a low operating voltage and has a lower cost price compared to electrostatic transducers. For these reasons the dynamic transducer is still the mostly used transducer and found in almost every studio headphone worldwide.
(Electro-) Magnetic transducer
The fourth transducer is the magnetic or electromagnetic transducer. This transducer is frequently used in high-quality/expensive in-earphones. The electromagnetic transducer is similar to the dynamic transducer, the diaphragm, however, is the magnet at the same time and the coil is firmly under it (as the permanent ring magnet with dynamic transducers). The clear advantage is the sound pressure. These transducers produce more sound pressure than dynamic transducers, for example. The disadvantages include the higher cost price, the slightly worse distortion factor and the response peak. The response peak (depending on the model) is a defined frequency response which has been raised by some dB’s and affects the sound so badly because a certain range has been raised (for example voice range or bass range). This is the reason why 2- or 3-way systems are used for high-quality in-earphones. The complete frequency response can be more flat. (In a 3-way system the response peaks of the three systems are tuned to each other).
• Different versions (designs)
Concentrating on the market for dynamic headphones, we will see that there are big (normal) and very small designs depending on the application.
In-earphones and intra-concha earphones
The smallest designs are in-earphones (also in-ear monitors, briefly IEM) and intra-concha earphones. The difference between both designs is that in-earphones (such as the DT 60 PRO from beyerdynamic) are put into the ear canal like earplugs while intra-concha earphones (such as the MX series from Sennheiser) are only put into the concha in front of the ear canal. As a result in-earphones provide a much higher isolation from ambient noise, what can be of great importance for various reasons.
Let us start with the acoustics. If the in-earphone fits correctly in the ear canal and provides a tight seal, the space between tympanic membrane and diaphragm is closed and very small. The whole thing acts like a kind of spring system (or push-pull mechanism) and the diaphragm can move the tympanic membrane with a small deflection and less power, what results in very good bass response. When this system is not tight, the low frequencies are lost (as is the case with intra-concha earphones). The reason why that is the human ear is less sensitive for low frequencies (below approx. 150 Hz) than for higher frequencies. If we would like to hear low frequencies better, a lot of power has to be raised for the amplification. When using loudspeakers low frequencies can also be felt physically. This is not the case with headphones. Loudspeaker diaphragms are also bigger and more stable (thicker material), which moves a lot of more air compared to headphones. In order to make optimal use of the small power which the headphone system develops, you have to make sure that the headphone or in-earphone seals optimally.
The second reason, why it is so important that the in-earphones seal tightly, is that it is not necessary to set the volume so high to protect our hearing. An in-earphone is “plugged” into the ear, seals against ambient noise and can use the ear canal as a sound box to produce a better sound. Especially the bass response is significantly better than with intra-concha earphones. In order to improve the isolation from ambient noise, sound and comfort for wearing, many manufacturers offer the opportunity to produce so-called ear moulds, which are exactly fit to one’s ears and replace the supplied standard ear plugs. Now we leave the in-earphones and intra-concha earphones and have a look at supraaural headphones. But it makes sense to have a closer look at the acoustical operating principles first of all.
• Excursus: Operating principles
Open, semi-open or closed? What is the difference?
As already described, the bass response with in-earphones is very good, because the space between diaphragm and tympanic membrane is “closed”. In principle these are closed systems. The conclusion that closed headphones provide the best bass response, however, is not quite right, because the system operates completely different than in-earphones. But this is not easy to explain and would extend this document considerably. The biggest differences between closed and open headphones is the strong isolation from ambient noise (and vice versa, the environment cannot hear what is played in the headphone) with closed headphones and the better spatial sound with open headphones. Semi-open headphones (such as the DT 880 PRO) are more or less a combination of both and try to combine the advantages of both. If we look at it from the mechanical point of view, we find out that open headphones have an advantage to closed headphones: The air volume closed between the diaphragm and ear cap dampens the diaphragm vibration. Pressure equalization can take place with open headphones through the shell which affects the faithfulness of pulse. But the increased dampening of the diaphragm also reduces the risk of uncontrolled vibrations.
The selection of the headphone due to its sound strongly depends on what we want to listen to. An open or semi-open headphone is perfect for classical or jazz music, which doesn't need a strong bass response, but a very high faithfulness of pulse. A closed headphone is ideal for pop or rock music.
Finally the kind of application (where will we use the headphone?) determines the suitable operating principle. If the headphone is to be used in a quiet environment (for example studio) to mix or monitor music, the headphone can freely be chosen according to one’s personal taste. If the headphone however is to be used, by the musician for monitoring during recording, a headphone should be chose which attenuates ambient noise and is attenuated so that the microphone does not pick up sound reproduced by the headphone.
Now that we know the differences between open, closed and semi-open headphones, we can return to the supraaural headphones. As supraaural headphones only lie on the ear, the space between tympanic membrane and diaphragm cannot be completely closed. With supraaural headphones you only differ versions with open or closed housing, whereby the use depends on the music style and/or kind of application, as we meanwhile know.
Now we come to the ear pads, which affect the comfort for wearing. In addition to different materials there are different shapes: 1.) The “flat” close version, which is completely closed in the middle (known from old headphones for mobile cassette players). This version is quite rare today. 2.) The ring-shaped, open version (such as beyerdynamic DT 231 PRO), which provides today almost each supraaural headphone. When wearing the “flat”, closed version a lot of heat will develop what reduces the comfort of wearing. This problem is almost unknown with the ring-shaped, open version so that the headphone is still comfortable after wearing it for a longer period of time.
The circumaural headphone is the biggest design for studio headphones. The ear pads are large enough to fit closely around the ears to the head. As it were they enclose the ear and do not touch it. This results in a clear difference between closed, open and semi-open operating principles, because depending on the ear pad and the housing the spaces behind the diaphragm and between tympanic membrane and the diaphragm are closed or open.
The different ear pads (material, shape, thickness) also affect the acoustics of the headphone. If the ear pad, for example, does not attach and close correctly to the head, the space between tympanic membrane and the diaphragm is not closed and therefore the sound of the headphone is not ideal. The ear pads that fit most closely, however, are not always the most comfortable or hygienic ear pads. For example leatherette ear pads are easy to clean, but as they fit very close a lot of heat is developed what causes sweating. The sweat is not absorbed by the leatherette ear pads and can be removed with a normal, moistened cloth. Ear pads made of velour or cloth do not fit so closely and therefore are a bit lighter. Heat, however, is also developing after wearing them for a longer period of time. The sweat is absorbed by the ear pad. But this is not very hygienic when the headphone is worn by many different people. This is the reason why ear pads made of velour or cloth should be regularly cleaned. Of course there are other versions such as the gel ear pads which consist of a flexible foil and are filled with gel. When the headphone is put on the gel spreads over the whole area, the ear pad fits perfectly and an ideal isolation is produced. Genuine leather is also used for ear pads and behaves like leatherette regarding the development of heat and hygiene.
After having discussed transducers and acoustical principles, we now must decide which impedance do we need? Headphones are available with different impedances from 16 to 600 ohm and higher, but why?
In order to answer this question, we must return to the beginning; the actual task of a headphone is to convert the received signal into sound pressure. The stated nominal sound pressure level states how successful this conversion will be. The nominal sound pressure level is stated in “dB SPL” and states how much sound pressure is produced at 1 mW (0.001 watt) of electrical power. When we have a look at the nominal sound pressure of different headphones with various impedances, we can see that the number is always approximately the same value. But the source that is connected to the headphone, is different and ranges from a small MP3 player to a high-quality headphone amplifier. Portable devices (for example a laptop) operate with a lower operating voltage than permanent installations (for example a high-quality headphone amplifier). As the principles of electrical engineering teach us we need less power but more voltage with a higher impedance; i.e. an MP3 player that provides an operating power of 3.3 V, for example, will deliver a power of 42 mW to a 32-ohm headphone; with a 250-ohm headphone it is only 5 mW and with a 600-ohm headphone only 2.3 mW. This clearly shows that we need headphones with lower impedance for portable devices to achieve a sufficient output level.
Why are there headphones with higher impedance at all, if a device needs less power with headphones with lower impedance? There are two reasons for this from which the first is the operating voltage once again. In fixed installations (e.g. a headphone amplifier) the operating voltage is much higher, for example 15 V or 18 V. The output voltage is therefore considerably higher as with a laptop or MP3 player. The technical effort to drive high-impedance headphones, is less in this case as if driving low-impedance headphones, because the required output power is not so high (more voltage, but less power at a higher impedance). The second reason is mechanically: mass of the diaphragm and the coil sticked to the rear is less than with low-impedance headphones; and the mass can be moved easier and faster. The reason why is the lacquered copper wire (for isolation) of which the coil is made. It is a matter fact that thinner copper wire has a higher resistance (because less fits through) than thicker copper wire. The magnetic field, which a coil is to produce when an audio signal is present, depends among other things on the amount of turns. In order to achieve this minimum number of turns with low impedance, a thicker (and also heavier) wire is required and because the diaphragm cannot be endlessly light, the moved mass (diaphragm and coil) is relatively high compared to high-impedance system.
• Technical Specifications
Now we only have to explain a few technical specifications: frequency response, distortion factor and nominal power handling capacity.
The frequency response states the lowest and highest frequency, which the transducer can reproduce. These limit values are fixed to the point (the lower or upper frequency limit) where a drop of 3 dB results. Unfortunately, a standard does not exist and often some manufacturers are cheating a little bit. It remains a question, if the headphone really provides a frequency response of 20 to 20,000 Hz. The curve of this stated frequency response is also unknown. Is it linear or does it run up and down? You could see that from the frequency response of the headphone, but it is rarely enclosed because with most headphones it looks quite horrible. The explanation is found further down in this text. Ideally you should rely on your hearing and test the headphones in question to compare them. The best is to compare them with a headphone you already know.
We have briefly mentioned the distortion factor (also T.H.D. = Total Harmonic Distortion) already at the beginning. Without making it too complicated, the distortion factor (also referred to a T.H.D. = Total Harmonic Distortion) is the proportion of the original signal (fundamental sound) to the sum of the signals, which are not included in the original signal and therefore produced by the transducer itself or by the housing parts (harmonics). As the distortion factor is always smaller or at the most 1, it is mostly stated in percentage. So, the lower this value, the less harmonics are produced by the transducers and/or housing parts. This is what we want; as the harmonics do not belong to the original signal we have no control over them, if some are present.
The last topic I would like to mention in this article is the localization, one of the biggest problems with headphones in general. It is a problem in so far that the (headphone) loudspeakers are directly placed on our ears and not around or in front of us, what definitely sounds like that. Over the years many attempts at a solution have been presented, some work better, some are worse, but it did not solve the problem 100%.
Human beings determine the spatial sound or the direction of the sound with both ears. Travel time, frequency spectrum and volume differences are decisive when determining the direction, while the reflections (early and late reflections) of the room where we are located are decisive for the distance of the loudspeakers. But with headphones the sound comes directly into the ear and our head, the ears and the room have no impact any more. The “in-head localization” develops and we have the impression that the source is inside our head, what can be very inconvenient and, most of all, what sounds very unrealistic.
The sound of a linear loudspeaker is also influenced by our head and ears, but we perceive it as linear. This is not the case with a headphone either, because the sound goes directly into the ear; our head and our ears are not affected at all or only a little bit. With diffuse-field equalized headphones we try to solve this problem by capturing colourations, which are caused by the head and ears, with measuring instruments (with a dummy head) and adapt the frequency spectrum of the headphone to them. The description diffuse-field equalization derives from the fact that these colourations are measured in the diffuse field. The diffuse field is the location where reflections are louder than the direct sound. As the mechanical and electronically possibilities are limited, however, the frequency response can never be equalised perfectly. Furthermore, some headphones have also been adapted to personal taste. As already mentioned, directional hearing, depends on the shape of our head and our ears. The dummy head can only represent a non-universal average value. A diffuse-field equalization of the headphone reduces the so-called “in-head localization”, but this does not ensure to solve the problem completely. We recommend continuing to test the headphone. At last I would like to mention that there are further headphones which influence the audio signal with a DSP (such as Headzone® Pro and Headzone® Pro XT from beyerdynamic) and “manipulate” our brain. First of all the influence of head and ears are determined by measuring instruments and stored as so-called HRTF (Head Related Transfer Function). This HRTF contains information about the travel times to our head (e.g. how long it takes until a signal coming from the right hand side has reached the left ear after having reached the left ear, etc.). The audio signal is adapted by the DSP (Cross-Feed, HRTF, part of the reverberation chamber etc. is added) and reproduced as stereo signal to the headphone. This is possible with both stereo and 5.1 surround signals minimising the “in-head” localization.
© 2008 Peter Grooff
Product Manager Headzone® / Headphones / Headsets ProAudio
beyerdynamic GmbH & Co. KG