Headphone Impedance is the electrical characteristic of the headphone voice coil and magnetic field coupling of the voice coil & magnet inside the headphone.

Impedance is an electrical unit which expresses the combined Resistance, Inductance and Capacitance of the headphone's voice coil.

In terms of matching to an amplifier, the headphone impedance is the "Load Impedance".

An amplifier drives a headphone with voltage, the current drawn by the headphone is proportional to the headphone's impedance.

As a rule of thumb, the load impedance (headphone) should be at least eight times higher than the amplifier output impedance.

This allows the amplifier to exert more control over the headphone, i.e. the lower the output impedance of the amp, the higher the damping factor and the higher the efficiency of the amplifier/headphone interface (See note 1 for additional detail).

Some basic electrical theory:

• Voltage = Current x Impedance,      in AC circuits, i.e. music, power lines, etc.
• short form for voltage is "V" and is expressed in Volts
• short form for current is "I" and is expressed in Amps
• Power =  Voltage x Current,
• short form for power is "P" and is expressed as "Watts"
• Impedance = Voltage/Current,       short form for impedance is "Z"
• Resistance = Voltage/Current,       short form for resistance is "R", this is for DC circuits, i.e. power supplies, batteries
• see note 2 for definition of Voltage and current, see note 3 for definition of resistance, inductance and capacitance.

Here's an example:

For a headphone to receive 2 watts of electrical power (this is almost never the case, just a simple example) it can be driven by:

1. 1 Amp @ 2 Volts , or
2. 2 Amps @ 1 Volt.

therefore if 1 Volt draws 2 Amps of current the headphone has an impedance of:

Impedance = Voltage/Current (or Z = V/I)

therefore Z =1/2 Ohm.

Based on Z=V/I, we can deduce that Case 1 will have an impedance of 2 Ohms, and Case 2 will have an impedance of 0.5 Ohms (see note 4).

Again, this is just an example, headphone impedances are actually much higher.

Headphone power levels are almost always much lower, for example: 1 - 10 milliWatts.

1 milliWatt (mW) is 1/1000 of a Watt.

If a 600 Ohm headphone draws 10 mW, Voltage = 2.44 V, current = 4.1 milliAmps (mA)

If a 32 Ohm headphone draws 10 mW, Voltage = 0.56 V, current = 18 mA

Pros and cons of high impedance headphones:

1. Headphone impedance is usually increased by thinner wire and most importantly more turns of wire in the voice coil. More turns or loops creates a larger field (area of magnetic influence). In layman's terms more magnetic force for the coil to move the diaphragm.  Thinner wire usually works out to a lighter, more responsive diaphragm. Depending on the headphone design, this may lead to more accurate response.
2. The displacement (amount of movement) of the diaphragm (the part that vibrates to produce sound) can be better controlled via a more accurate flux (magnetic field to pull and push the diaphragm).
3. Difficult to drive for small headphone amps with low output voltage and low gain.
4. Most high impedance headphones need an amplifier with higher voltage gain and higher output voltage, e.g. the 600 ohm Beyer DT770/880/990 series.
5. Allow Solid State Op Amps to work more efficiently with less distortion. Have a look at Op Amp data sheets and a graph of distortion vs. output impedance for most audio Op-Amps and you'll get the idea. This is a very complex subject, but most Op Amps are designed to output (typically) up to 10 Volts into 600 ohm loads or higher.

Pros and Cons of low impedance headphones:

1.  Headphone impedance is usually decreased by thicker wire and less turns of wire in the voice coil. The magnetic field is built up by more current.
2. Easier for small and/or portable headphone amps to drive. For example: an iPod or MP3 player headphone jack. Many small and/or portable headphone amplifiers are designed to output a volt or two into low impedance, high efficiency headphones, e.g. Grado headphones.
3. Low impedance, low efficiency headphones usually sound better when driven by a desktop amplifier, e.g. Audeze LCD-2 or AKG K70X.
4. Low impedance headphones usually sound better when driven by a solid state or a transformer coupled vacuum tube amplifier.  Low impedance headphone do not usually work well with Output Transformerless vacuum tube amps.

Generally, a properly designed desktop headphone amplifer can drive high and low impedance headphones and can drive headphones of any efficiency.

Do high impedance headphones sound better than low impedance ones?

No, there are many very good low impedance headphones available and many very good high impedance headphones available.

In headphones, the sound also depends upon the following:

1. Frequency response, this is a representation of volume decreasing or increasing with frequency, this is actually the magnetic field pulling and pushing with different amount of force at different frequencies, impedance affects this in complex ways.
2. Distortion, represents the amount of "change" from the actual signal to the real signal (i.e. input signal to output signal).
3. Build and design, i.e. sealed vs. open, etc.
4. Diaphragm design, voice coil design

To conclude:

High impedance headphone:  requires more voltage but less current

Low impedance headphone:   requires less voltage but more current

Amplfier output impedance: the lower the better for any impedance headphone

higher efficiency (for example:  102 dB/1 mW) less power required

​lower efficiency: (for example: 91 dB/mW) more power required

How they effect sound is dependent on the magnetic field generated and diaphragm construction (physical response of diaphragm).

We hope this clears the air.

A few notes:

1. Some people prefer the sound of vacuum tube headphone amplifiers which usually have higher output impedance relative to solid state (transistor)  amplifiers, and therefore have poorer damping factor and hence, fuller bass.
2. Voltage is electrical pressure, current is flow of electrons
3. Resistance is constant with frequency, capacitance creates lower impedance at higher frequencies, inductance creates higher impedance at higher frequencies
4. Headphone impedance is actually quite complex and different driver designs have different impedances at different frequencies and (to a lesser extent) voltage amplitude
5. Assumes diaphragm construction to be similar and magnetic field to be similar
6. Assumes linear response at all frequencies (not possible for dynamic driver headphones)
7. Assumes that impedance increases with change in coil characteristics only.
8. Assumes speed of electrons in an electrical circuit is constant, it does not change with voltage or current or power.
9. Other headphone types which do not have a voice coil (for example, those sold by Hi Fi Man, Audeze and Stax) can also characterized by their electrical impedance.

Thanks, Rohan. Just intuitively, I feel that my higher impedance phones (Senns) have more subtlety than my lower (701s, Grados, Audio Technica). If I've grasped your explanation at all, it seems to fit that experience.
So does this mean that even lower impedance headphones would benefit from a headphone amplifier because it could deliver more, though slower, electrons? My Denons AHD2000s are only 25 ohms, but I couldn't imagine listening to them without a headphone amp.
Very informative, thanks for the info!
Do you know any good sources (papers, websites, books) of more information on headphone design, impedance, and the effects of amplification? I've had several undergraduate courses in EE and signals but never a really in-depth explanation of the physics of dynamic speakers. If my schedule wasn't so packed with computer engineering courses I would take some audio engineering courses, but alas time is not on my side right now.
The speed of electrons never changes, as a function of impedance or anything else. It's a constant. Also, the planar magnetics that are so popular in recent times show that low-impedance cans certainly can deliver subtlety. From time to time, ribbon speakers become all the rage in high-end hi-fi, usually for their "unbeatable" clarity, delicacy, etc. The same is said of electrostats, and so on. We need to be cautious of generalizations, especially when answering noobie questions..
PS: In the post above, I meant to point out that speakers like the Apogee Scintilla and Martin-Logan CLS approached a "dead short" (short-circuit) at high frequencies, yet each of these was all the rage for a period and considered state-of-the-art by its proponents. Farther back then that, electrostatic speakers with no input transformers were driven by direct-coupled OTL tube amps, which mated well because of the stat's multi-thousand-Ohm impedance. So, again, high or low impedance, transformer coupled or direct-coupled, solid state or tubes... all can work exceedingly well if properly designed and well engineered. Due diligence in research can protect against disappointments down the road.
i agree need to edit this, speed of electrons is constant in a conductor, its the kinetic energy which is responsible. The amount of work done in moving an e- form 1 point to another is voltage not it's relative speed. jeffreyfranz, it's a wiki please edit as you like.
Is this relate to the actual current rise of voice coil?
Wow, thanks for the Wiki. I'm finishing up my physics (electicity) course and I found this to be really relevant! (Both in the sense for headphone hunting and for physics review) :D
Just saying, whilst higher impedance headphones certainly require more power to achieve a high volume, it is lower impedance loads that are really punishing for an amplifier.
WOW. this cleared up things a bunch. and that's from an Electrical Engineer. Well, a freshman one :D
I am glad that this discussion has been helpful. However, I do not have a great deal of knowledge beyond what I have already shared. I am not an EE. What would be good for a true Wiki reference on this subject (and related ones) is for a veteran engineer with a talent for expressing complex subject matter in common language to lend a hand.
Well, Hugh Robjohns is a very experienced sound engineer with a degree in EE. He wrote an excellent article on impedance and hi-fi that you can find at
http://www.soundonsound.com/sos/jan03/articles/impedanceworkshop.asp
This is a simpler and practical expression and takes in a lot of assumptions, one could go onto more depth but then it wont be useful for the people who don't have an EE background. I kept it short for this reason only.
Thanks, Scuttle, I'll check out that article.
Isn't it so that a headphone amplifier has an output impedance, and that it should be matched as close to the headphone impedance as possible to ensure the best damping factor - but then what is "output impedance"