Headphone Fundamentals

[AudioStudies]

The main document has been thoroughly revised, but is getting a bit long for a single post. So I decided to post each revised section individually. This post is the revised Introduction. Next post will be Types of Headphones.

Introduction

Headphones are electroacoustic transducers (speakers) that convert electrical energy to acoustic energy; designed for a single person to listen, rather than a group. In the colloquial, many types of headphones are known as cans. Headphones differ from traditional room loudspeakers in that the driver elements to produce sound are smaller, and therefore require less power to get them to vibrate and produce sound at various frequencies. Headphones drivers have light membranes that are damped (both acoustically and mechanically) much better than drivers within room loudspeakers.

Historically, headphones originated from the telephone receiver earpiece, and initially they were the only way to listen to electrical audio signals prior to the advent of amplifiers that could power room speakers. All of the early applications of headphones were of monophonic sound directed to both ears.

In 1958, John C. Koss, an audiophile and jazz musician, produced the first stereo headphones to enable listening to music that had been recorded in stereophonic sound. Prior to the development of stereo headphones, the early monophonic devices were used exclusively by the US navy, telephone and radio operators, and individuals in similar professions.

Mass market headphones are mass produced headphones that may or may not be capable of good sound. Mass market headphones, particularly of the low grade variety, are often marketed with features that can include noise-cancelling, FM radios, waterproofing, and smart technology. Upper level mass market headphones typically do not include these features as often. Mass market headphones typically work well with mobile phones or other low powered devices.

High end headphones are luxury products often constructed with high quality components, and designed to show the state of the art of headphone technology. However, even with high end headphones there are some misses and not all of them end up sounding good. Many of the higher end headphones are designed to be used with accessories such as external amplifiers called headphone amplifiers and combination headphone amplifier / digital-to-analog converters (DACs). However, some high end headphones are designed specifically to be used with mobile phones and other low powered devices.

Many types of headphones are produced due to historical reasons and advances in technology that have led to better manufacturing techniques, smaller and more efficient transducers, and marketing for specific types of uses; gaming for example. The invention of many things such as the Walkman, mobile phones. Ipod, and laptop computers have influenced the headphone market, and subsequently the technology, considerably.

 

[AudioStudies]

Types of Headphones

Headphones designed to cover the ears, over-ear headphones, sit around the pinnae onto the skull; and are often characterized as either open-back (with an open grill on the back side of the ear cups) or closed-back, with no openings. Traditionally, open-back designs have allowed for easier production of high quality headphones; however many excellent products are now of closed-back design. A more limited number of headphones are designed as semi-open, and the back is only partially open.

There is no official standard or scientific distinction between open-back and closed-back headphones. The distinction, though colloquially common, is not ideal for characterizing all modern headphones. Problematic exceptions to the colloquial distinction between closed-back and open-back include some headphones that have a closed back to each cup but isolate noise poorly because both the rear volume and the front volume are not isolated from the exterior air; and headphones with an open back that have completely sealed ear chambers.

There are arguably two principle volumes from the standpoint of a headphone, the space in front of the driver (between the driver and the ear), and the space behind the driver (typically the cup, or open air for some designs). Either of these volumes may be sealed or conversely have some degree of leak relative to the outside air. The relationship between these two volumes is of great significance with respect to headphone performance. In most cases, with open-back headphones, sound will emanate to the nearby environment; and with most cases of closed-back use there is not nearly as much leakage. In general, closed-back has the advantage that they can be used when others are around, such as on airplanes.

Some headphones called in-ear monitors (IEMs) are designed to fit directly into the ear (not cover the ear). IEMs have become increasingly popular. The better IEMs have excellent sound quality, but may not feel as comfortable as over-ear headphones, especially for long listening sessions. IEMs are lighter, smaller and easier to store and transport than over-ear headphones.

Earbuds are small ear pieces that unlike IEM's do not go into the ear canal and do not sit on the ear but are tucked into the concha of the ear. The sound quality of earbuds is often lacking as it is very difficult to get a good seal. Earbuds are often included as freebies with portable gear; and consequently are not recommended for high quality sound. They have the advantages of low cost (or no cost) and portability; as for example the Apple Earpod.

Another type of headphone is the on-ear headphone, that doesn’t completely cover the ear, nor does it fit directly in the ear; rather they sit on the pinnae. Depending on pinnae size and the available room within the pads, some over-ear headphones are partially on-ears for certain people; often leading to a poor seal and poor performance. On-ear headphones are not as popular as the other types; and not noted for sound quality. They may not feel as comfortable as over-ear; however they can feel less hot (or warm) for extended use and are more suited for on the go use.

 

[AudioStudies]

Impedance

Impedance is a measure of the opposition to current that a circuit exhibits when a voltage is applied to the circuit. Impedance is similar to resistance; however typically includes other aspects that oppose current, beyond just resistance, such as capacitance and inductance.

Most headphones have significant inductance and small amounts of capacitance, in addition to resistance. Therefore, when headphones are connected to a power source, there is a reactive load, and impedance is present; and the magnitude of this impedance is a measure of opposition to current that a particular set of headphones exhibit.

Headphone impedance is typically measured at frequencies of 500 Hz or 1 KHz and specified in units of Ohms (Ω). Almost all headphones fall within the impedance range of 16 Ohms to 600 Ohms, and most fall within the range of 16 Ohms to 300 Ohms. Anything over 100 Ohms is considered high impedance.

Knowledge a headphone's impedance can help determine how much voltage headphones will need to produce a reasonable listening volume. High impedance headphones require the use of a headphone amplifier (typically not battery powered) to perform properly. High impedance headphones cannot be powered well from the output jack of a smartphone or other low powered device.

Mobile phones and other low powered devices can typically power headphones with low rated impedances up to 32 Ohms (32 Ω). When headphone impedance is in the medium range between 33 Ω and 100 Ω it may (or may not) be possible to generate enough volume from a low powered device; and even if that is possible, it is likely that improved performance will be attained with the use of a headphone amplifier. Many headphones have rated impedance specifications in the range from 100 Ω to 300 Ω and these headphones will definitely require the use of a headphone amplifier.

In general, high impedance headphones require more voltage, the driving force to overcome the impedance. Although high impedance headphones produce less volume for a given amplifier output level they are coveted because of their ability to superbly handle electrical signals; thus enabling more accurate and vivid sound.

Typically, because they are coveted, high impedance headphones cost more than low impedance headphones. However, there are some excellent designs of headphones with relatively low impedance.

An amplifier used to power headphones, including headphone amplifiers and amplifiers within a low powered device, has an output impedance specification. The amplifier should have an output impedance less than of the impedance rating of the headphones; and the lower the better with respect to amplifier output impedance.

Impedance (continued)

An arbitrary statement that appears on many web sites is that an amplifier should have an output impedance less than 1/8th of the impedance rating of the headphones. Some experts believe 1/10th is a more appropriate ratio. Following the 1/10th guideline will help to ensure a flat frequency response when listening to the headphones.

Headphone impedance can (and often does) vary with the frequency of the incoming signal; thus in these cases, the actual impedance can vary from the rated impedance when the headphones are in use. However, some headphone designs have little to no impedance change over the audible frequency range.

In general, a low impedance headphone requires more current and less voltage; whereas a high impedance headphone requires more voltage and less current. The voltage is the driving force that produces current, and this driving force needs to be greater to overcome higher impedance loads. Conversely lower impedance loads do not require as much voltage to drive through the impedance, but the current requirement will be greater because there is less impedance resisting the current.

Power consumption of headphones is determined by the applied voltage and the drawn current. At the maximum amplitude in an AC circuit, power, P, is defined as the product of voltage and current: P = VI. It is therefore possible that a low impedance headphone and a high impedance headphone could have the same power requirements. Additionally, a low impedance headphone due to the need for current, may have a greater power requirement than a high impedance headphone. Therefore, impedance alone is not a determinant of how much power is required to drive a headphone to appropriate listening volumes. Headphone amplifier power is typically specified in milliWatt (mW).

 

[AudioStudies]

Sensitivity

Sensitivity is a measure of how effectively headphones (or other loudspeakers) convert electrical signals into sound; thus indicating how loud the headphones will be for a given electrical drive level. Sound pressure level (SPL) is typically measured in units of decibels (dB). There are two methods of measuring and specifying sensitivity:

• In decibels (dB) of SPL at one milliwatt = [ dB (SPL) @ 1 mW ]; or
• In decibels of (dB) of SPL at 1 volt = (dB (SPL) @ 1V).

Regrettably these are often (incorrectly) interpreted to mean the same thing and used interchangeably. Those who do know the difference will often distinguish between them as power sensitivity and voltage sensitivity.

Often sensitivity specs from manufacturers depict only a number and leave out the units of measure (dB/mW or dB/V) for proper interpretation to determine whether the value is the power sensitivity or the voltage sensitivity.

When units are provided, some manufacturers show dB/mW whereas others (who like to show high numbers) depict dB/V values. Yet other manufacturers do not even specify voltage sensitivity; rather they publish maximum sound pressure levels (SPLs), based on their maximum power rating. All this corresponds to an un-standardized chaotic mess. The end user should be aware that the most convenient efficiency number that can be used to directly compare headphone efficiency is the voltage sensitivity (dB/V value, which if unknown can be calculated from dB/mW rating combined with the rated impedance).

Headphone amplifiers are voltage sources; therefore the voltage sensitivity is most useful for comparison of speakers. However, if only the power sensitivity is known, a form of Ohm’s Law can be used to convert to voltage sensitivity or by using online calculators. When the voltage sensitivity is known, the maximum volume for a pair of headphones can be calculated if the maximum amplifier output voltage is also known. The appropriate voltage specification is root mean square (RMS) voltage, often denoted VRMS .

Consider the case of an amplifier with an output voltage VRMS = 1.0 volts; connected to headphones with sensitivity of 100 decibels SPL per volt = 100 dB (SPL)/V. The maximum volume that can reach the ears for this case is 100 decibels.

Clearly, sensitivity also plays a role in determining how much power is required to drive a headphone. The very hardest headphones to drive are those with low impedance and low sensitivity because of great current requirement due to the low impedance combined with the voltage requirement to overcome the low sensitivity and achieve listenable volume levels.

 

[AudioStudies]

Choosing Headphones

The most important considerations in choosing headphones are tonal balance and overall sound quality. Clearly, the impedance and sensitivity of headphones are important considerations; as well as the specifications of the amplifier intended to drive the headphones. There are other considerations as well such as build quality, comfort, aesthetics, value and affordability.

The intended use is of great importance, as closed-back may be better for going to the gym or travelling via public transportation, and open-back may be better to achieve great sound for private uses. A gaming headphone has different user requirements than one intended solely for listening to music on a mobile phone.

User preferences such as in-ear, on-ear, or over-ear certainly come into play. The reputation of the manufacturer, serviceability, and the warranty for a given set of headphones are also important considerations.

High sensitivity earphones, and some headphones, can exhibit audible hiss in the background that is independent of the volume control. Clipping of the amplifier is not problematic with these high sensitivity phones because of the very small voltage requirements. Conversely, there is a danger of playing music too loudly, when barely touching the volume control. Impedances of high sensitivity headphones are typically quite low, and this can cause tonal changes if the headphones are used in conjunction with an amplifier with high output impedance.

Medium sensitivity headphones typically have no problems in playback, as they can often (but not always) play loud enough from a low powered device such as a mobile phone, and they are easily driven by a headphone amplifier. Medium sensitivity headphones will most often have impedances in the range between 16 and 120 Ohms; and most likely in the range between 32 and 60 Ohms. Too much voltage is rarely an issue with medium sensitivity headphones.

Low sensitivity headphones typically have impedances over 60 Ohms, and will require higher output voltage from the amplifier than most other types of headphones. In these situations, the output resistance of the amplifier is rarely an issue. With low sensitivity headphones, the output impedance of the amplifier is rarely problematic.

 

[Scgorg]

Reactive Loads

The transducer technology chosen for a headphone design is related to the reactance and thus the impedance. ??Can anyone elaborate???

Moving coil transducers usually exhibit a lot of inductive behaviour, and for larger drivers the inductance prevents high frequency extension. This inductance also spikes around the resonance frequency of the transducer, leading to the typical impedance spike we often see in impedance plots of moving coil transducers. In short, moving coil transducers typically will have an impedance at their resonance frequency, and also have slowly increasing impedance with frequency. The Q of the driver determines how large this impedance spike at the resonance frequency is, with lower Q meaning a larger spike and vice versa. The electrical phase of moving coil drivers is typically not flat either, further complicating the impedance.

Planar magnetic headphones typically exhibit the characteristics of a pure resistor, just with a slight inductive element at very high frequencies. In general a planar magnetic transducer can be considered a very friendly load for most amplifiers, though they are typically designed in such a way that they have lower sensitivity and/or lower impedance than their moving coil counterparts.

Electrostatic headphones represent a purely capacitive load. Their impedance lowers with frequency. Going by the specs Stax give for the SR-009 (145KOhm@10KHz) tell us that at a frequency of 1000Hz the impedance will be an order of magnitude higher, and 2 orders of magnitude higher at 100Hz (etc. etc.). The voltage required for a given SPL is almost constant regardless of frequency, but the impedance varies from 72.5MOhm@20Hz to 72.5KOhm@20KHz. This naturally means the the power requirements for high frequency reproduction is much higher relative to low frequencies.

While there are other more novel driver technologies these 3 make up almost all that are used in headphones.

 

[AudioStudies]

Revised section "Reactive Loads" after the help from @Scgorg

Reactive Loads

Headphones typically do not exhibit the behavior of pure resistors; rather inductance is prevalent in most headphones, and to a lesser degree capacitance. Most headphones exhibit reactance, opposition to electrical current attributable to inductance or capacitance (or both), when connected to a source amplifier. This reactance represents a complex impedance.

In situations of reactive loads, the capacitive and inductive elements causes the impedance to change with frequency. Audio signals containing music are transmitted in AC current, thus impedance, rather than resistance, is how headphones are rated for use. Calculations involving reactive loads involve more complex mathematics than the simple forms of Ohm’s Law that can be used in circuits that are purely resistive and DC current only.

Reactance is similar to resistance in the sense that the greater reactance, the smaller the current for a given voltage. However, there are significant differences between reactance and resistance. Consider that reactance changes the phase, power is stored rather than dissipated in a purely reactive element, reactance can be negative, and reactance in a circuit element is frequency-dependent. The transducer technology chosen for a headphone design is related to the reactance and thus the impedance.

Phase is the time difference between the peak voltage and the peak current. A phase shift occurs in reactive circuit elements such that the current through the element is shifted by a quarter of a cycle relative to the voltage.

The Q factor (quality factor) is a measure of the quality of a resonant circuit; and is the ratio of the reactive power to the average (dissipated) power. The Q factor is a dimensionless parameter indicative of the energy losses within a resonant element; and it is smaller when the energy losses are greater. The voltage across an inductor or capacitor in a series resonant circuit is Q times the total applied voltage.

Moving coil transducers typically exhibit significant inductive behavior, and for larger drivers the inductance prevents high frequency extension. Inductance in moving coils spikes at the resonance frequency of a given transducer; and this spike is often depicted within impedance plots. Thus in addition to the impedance that slowly increases with frequency due to inductance of a larger moving coil transducer, an impedance spike occurs at the resonance frequency.

The Q factor of the driver determines the magnitude of the impedance spike at resonance in moving coil transducers; lower Q factors yielding larger spikes, and higher Q factors rendering lower impedance spikes. The electrical phase of moving coil drivers is typically not flat; thus further complicating the impedance.

Planar magnetic transducers closely resemble the characteristics of a pure resistor, with just a slight inductive element at very high frequencies in the audio band. Planar magnetic headphones typically exhibit a very easy load for most amplifiers; although typically the design has lower sensitivity and/or lower impedance than their moving coil counterparts.

Electrostatic transducers exhibit an almost purely capacitive load wherein the impedance lowers with frequency. Some slight resistance is attributable to radiation of acoustic energy and Ohmic losses. With electrostatic transducers the power requirements for high frequency reproduction is much higher with respect to low frequencies. The voltage required of a given SPL is nearly constant regardless of frequency for electrostatic transducers.

 

[JustAnandaDourEyedDude]

The Q factor (quality factor) is a measure of the quality of a resonant circuit; and is the ratio of the reactive power to the average power. The Q factor is a dimensionless parameter indicative of the energy losses within a resonant element. The voltage across an inductor or capacitor in a series resonant circuit is Q times the total applied voltage.

The Q factor is the ratio of the reactive power to the average real (dissipated) power. The Q factor is a dimensionless parameter indicative of the energy losses within a resonant element, and it is smaller when the losses are relatively higher. The voltage across a single inductor or capacitor in a series resonant circuit with a resistor is Q times the voltage across the resistor, and Q/(1+Q) times the total applied voltage.

The Q factor of the driver determines the magnitude of the impedance spike at resonance in moving coil transducers; lower Q factors yielding larger spikes, and higher Q factors rendering lower impedance spikes.

I could be mistaken, but I would guess the opposite trend would be true.

 

[AudioStudies]

I could be mistaken, but I would guess the opposite trend would be true.

In your post you talk about the Q being smaller when energy losses are higher. I would think the energy losses would be higher around a spike in impedance at the resonance frequency. Also, if the opposite were true, it would seem to contradict @Scgorg Perhaps others can weigh in, or I can do more research online. Thank you for your comments.

 

[Scgorg]

I may very well have mixed up low Q and high Q in these cases. It's been some time since I read up on the subject.

 

[JustAnandaDourEyedDude]

In your post you talk about the Q being smaller when energy losses are higher. I would think the energy losses would be higher around a spike in impedance at the resonance frequency. Also, if the opposite were true, it would seem to contradict @Scgorg Perhaps others can weigh in, or I can do more research online. Thank you for your comments.

I may very well have mixed up low Q and high Q in these cases. It's been some time since I read up on the subject.

Yes, I think I was mistaken about the trend being opposite. I had never looked into Q factor before I read the "Reactive Loads" post above. After reading it, I looked up the quite informative (as always) Wikipedia entry:
Q_factor
After quickly skimming through the Wikipedia entry to get a basic understanding of Q factor, I responded to your post without a firm understanding of the impedance of a headphone or speaker. You (AudioStudies) are right in this case, the higher impedance at resonance represents a higher resistance and thus higher energy losses and a lower Q factor. In general, the impedance could be higher because of higher reactance which is not accompanied by energy losses. In my earlier post, I was confused about the reactance of the driver, falsely imagining that it rises at resonance to cause a rise in impedance and therefore a rise (my mistaken thinking; a fall would be the logical outcome) in reactive power (manifested in amplitude/excursion of driver oscillation) and not real dissipative power. Thinking further about it now, I realize that the reactance of the driver probably decreases while also its mechanical resistance and mechanical damping decrease at the resonance frequency leading to the larger excursions.

The details for headphones (more significant for large heavy diaphragm-coil assemblies in large speakers) are described by solderdude in the following and related posts in that thread
https://www.audiosciencereview.com/...helli-lab-erish-coming-soon.12963/post-468688.
In re-reading those posts, it seems that at and close to a resonant frequency of the driver (where the mechanical damping is lowest), the large excursion resonant mechanical motion of the driver gets electrically damped due to the coil motion in a magnetic field inducing a back emf that produces an opposing current leading to increased ohmic heating losses. Thereby also the current effectively drawn from the (nearly constant-voltage) amp decreases, which manifests as an increase (spike) in the effective impedance of the driver as seen by the amp, at and around the resonant frequency. Therefore, for the same reactive power in two driver designs differing in the amount of electrical damping power losses at the same resonant frequency, the one with the higher back-emf current and higher losses will have the taller impedance spike and lower Q factor (and smaller driver excursion and energy, other things being equal) relative to the other driver design, and vice-versa, exactly in line with what Scgorg posted. At other frequencies removed from the resonant one, mechanical damping dominates the driver damping, and both the reactive power and the ohmic losses are lower, and there is no impedance bump. solderdude or other expert on ASR would be able to clarify it one way or another.

Electrostatic transducers exhibit a purely capacitive load

Electrostatic transducers represent an almost purely capacitive load. There is some resistance that accounts for both acoustic energy radiated away and ohmic losses. You cannot violate the First Law of Thermodynamics; many have tried and failed; none has succeeded.

 

[AudioStudies]

Electrostatic transducers represent an almost purely capacitive load. There is some resistance that accounts for both acoustic energy radiated away and ohmic losses. You cannot violate the First Law of Thermodynamics; many have tried and failed.

I wonder if the resistance is so small that it is negligible in comparison with the capacitance?

 

[JohnYang1997]

I wonder if the resistance is so small that it is negligible in comparison with the capacitance?

At very high frequency yes. But for audio frequency it's almost non conductive. Or you can say the "resistance" is too high. And the cable resistance/impedance will play much larger roll in this.

 

[AudioStudies]

Electrostatic transducers represent an almost purely capacitive load. There is some resistance that accounts for both acoustic energy radiated away and ohmic losses. You cannot violate the First Law of Thermodynamics; many have tried and failed.

Updated.

 

[AudioStudies]

And the cable resistance/impedance will play much larger roll in this.

Thanks, John. If I wanted to add a paragraph about cable impedance, what would be some appropriate things to say? I know there is resistance, but in those short cable lengths, I don't know how problematic the capacitance of the cable would be. Also, is the same gauge typically used for the cable, or if not what is the gauge range used for headphone cables? What is the maximum length of cable used? Do the cable properties interact differently with the various types of transducers?

 

[AudioStudies]

The Q factor is the ratio of the reactive power to the average real (dissipated) power. The Q factor is a dimensionless parameter indicative of the energy losses within a resonant element, and it is smaller when the losses are relatively higher.

Updated

 

[AudioStudies]

Does anyone think there will eventually be a lossless codec for wireless headphones?

 

[raistlin65]

Does anyone think there will eventually be a lossless codec for wireless headphones?

I'd be more interested in an industry practice of making the batteries replaceable in Bluetooth headphones. Rather than the current planned obsolescence in non-changeable battery design.

Until then, I'll stick to my wired headphones, which could still be working well 10 years from now.

 

[JustAnandaDourEyedDude]

An arbitrary statement that appears on many web sites is that an amplifier should have an output impedance less than 1/8th of the impedance rating of the headphones. Some experts believe 1/10th is a more appropriate ratio. Following the 1/10th guideline will help to ensure a flat frequency response when listening to the headphones.

The statement appears arbitrary only because the websites do not cite the basis. I believe it is based on a study by the mysterious and controversial NWAvGuy, who also designed the low-cost high-performing Objective and O2 amps as well as the ODAC to back up his claims. You should still be able to download his paper regarding the recommended damping factor of at least 8, if you search for it. Ensuring the damping factor is at least 8 or better yet 10 will ensure the frequency response or tonality of the headphones is not perceptibly changed, whether it is flat or not. You do not need to follow the damping factor recommendation in the case of headphones or IEMs with planar magnetic drivers, because their impedance is largely independent of tone frequency. Using even a low damping factor with planar magnetics (using an amp with high output impedance or a high-resistance cable) will not change the tonality significantly; though the voltage across the driver would be less than it potentially could be, other things being equal.

 

[AudioStudies]

The statement appears arbitrary only because the websites do not cite the basis. I believe it is based on a study by the mysterious and controversial NWAvGuy, who also designed the low-cost high-performing Objective and O2 amps as well as the ODAC to back up his claims. You should still be able to download his paper regarding the recommended damping factor of at least 8, if you search for it. Ensuring the damping factor is at least 8 or better yet 10 will ensure the frequency response or tonality of the headphones is not perceptibly changed, whether it is flat or not. You do not need to follow the damping factor recommendation in the case of headphones or IEMs with planar magnetic drivers, because their impedance is largely independent of tone frequency. Using even a low damping factor with planar magnetics (using an amp with high output impedance or a high-resistance cable) will not change the tonality significantly; though the voltage across the driver would be less than it potentially could be, other things being equal.

Over in the Objective and Subjective thread that I started, I was provided that source and have used some of the info from it. However, Solderdude disagrees with some of what that guy says. For example, for small drivers as used in headphones, the electrical damping should not be of concern because these small drivers are already damped well in the mechanical and acoustic domain. It was too late to edit my post after Solderdude informed me, but I updated the main document. I am sticking with Solderdude on this one. He still recommends the ratio (1/10 th) because there are other reasons (beyond electrical damping) to keep the output resistance of the amplifier low and hence adhere to the ratio.