[tabacaru]

I've often heard, including on this forum, that amplification means better sound quality. That is, say I'm using a pair of headphones that have a low enough impedance to not require amplification from a source such as an iPod, people claim that if I were to listen to the same source with the same pair of headphones but through an amplifier I would have better 'sound quality'.
 
I cannot personally fathom how this works - everything I've learned in school has taught me that amplification always comes at a price - whether that be power, or distortion.
 
And so I must ask the questions to you individuals that are more experienced than I (I am quite new here), how does amplification improve sound quality if it introduces things such as harmonic distortion, phase shifts, etc.?

 

[Uncle Erik]

First, impedance doesn't work that way. Impedance has much more to do with the efficiency of transferring power, not how much power an amplifier can deliver nor how efficiently a driver uses the power that gets through. Some low impedance devices - like some ribbon speakers and electrostats - present an extremely difficult load to amplifiers.

Next, even something like an iPod has an amplifier inside. The iPod is tuned for battery life since its screen and hard drive consume quite a bit of power. Thus, the power available to the amp section is cut back. However, this is a good compromise for the vast majority of iPod users, who listen with extremely efficient buds and IEMs. The point is that anything with a headphone jack has an amplifier inside.

Now, a driver is a physical device that moves back and forth. Some are easy to move and some are difficult. The difficult ones require more power. If you don't have enough power, it won't move. This can make an amp "clip" when it runs out of power. You'll usually hear clipping in the low end, because bass tones require the biggest movement and therefore the most power. Another issue is when there isn't clipping, but still not enough power for really good control over the driver. That will usually make the low end flabby, because the amp doesn't have the power to stop the driver from moving and pull it back in the other direction.

When it comes to power, impedance does play a role, but not in the simplistic sense kicked around here. Impedance varies in headphones. You can see the curves over at HeadRoom. Amplifiers also have an output impedance curve. Impedance (usually) varies by frequency in both headphones and amps. This means that delivered power varies according to what is in the music. This is where a lot of an amp's sonic character comes from. Two different 500mW amps will have different output impedance curves, so they will sound different with the same headphones.

There are other factors here, too, like the kind of circuit the amp has and the amplification devices it uses. Chips sound different and so do various tubes. Or you might be running a hybrid of the two.

Anyhow, to wrap up, you need an amplifier that has sufficient power to fully control a driver. An iPod's amp can drive some things, but putting those on a good desktop amp can result in much firmer control. You can hear that. And the circuit of the amp as well as its amplification devices lend a particular sound, as well.

 

[jasonb]

often you'll notice a difference with amping, but not all the time. depends on the equipment being used.

 

[tabacaru]

That clears a few things up, Erik.
 
On that note though, if I'm using the iPod as the source - wouldn't its amplifier have it's own impedance curve which will ultimately affect the curve of an external amplifier as well?
 
I guess your source is quite important.

 

[xnor]

Inputs have a high impedance of over 10k ohm, therefore, output impedance doesn't really matter anymore. 
 
Unless you have an ipod 3g I doubt that output impedance is a problem.

 

[Crazy*Carl]

I have never heard a headphone amp that made the sound quality any better than an ipod or computer motherboard.  I have heard some old receivers that had really bad sounding headphone out, but in 2010 headphone amps are not really needed anymore.  The only exception is if you enjoy spending lots of money on "hifi" gear that in reality doesn't do much.  If that sounds like its for you, then go buy a tube amp!

 

[CEE TEE]

I think this should be somewhere on Head-Fi easily accessed by new and old alike.
 
I am finally getting it.  My friend (mastering engineer) also explained it Uncle Erik's way to me:
 
Quote:

First, impedance doesn't work that way. Impedance has much more to do with the efficiency of transferring power, not how much power an amplifier can deliver nor how efficiently a driver uses the power that gets through. Some low impedance devices - like some ribbon speakers and electrostats - present an extremely difficult load to amplifiers.

Next, even something like an iPod has an amplifier inside. The iPod is tuned for battery life since its screen and hard drive consume quite a bit of power. Thus, the power available to the amp section is cut back. However, this is a good compromise for the vast majority of iPod users, who listen with extremely efficient buds and IEMs. The point is that anything with a headphone jack has an amplifier inside.

Now, a driver is a physical device that moves back and forth. Some are easy to move and some are difficult. The difficult ones require more power. If you don't have enough power, it won't move. This can make an amp "clip" when it runs out of power. You'll usually hear clipping in the low end, because bass tones require the biggest movement and therefore the most power. Another issue is when there isn't clipping, but still not enough power for really good control over the driver. That will usually make the low end flabby, because the amp doesn't have the power to stop the driver from moving and pull it back in the other direction.

When it comes to power, impedance does play a role, but not in the simplistic sense kicked around here. Impedance varies in headphones. You can see the curves over at HeadRoom. Amplifiers also have an output impedance curve. Impedance (usually) varies by frequency in both headphones and amps. This means that delivered power varies according to what is in the music. This is where a lot of an amp's sonic character comes from. Two different 500mW amps will have different output impedance curves, so they will sound different with the same headphones.

There are other factors here, too, like the kind of circuit the amp has and the amplification devices it uses. Chips sound different and so do various tubes. Or you might be running a hybrid of the two.

Anyhow, to wrap up, you need an amplifier that has sufficient power to fully control a driver. An iPod's amp can drive some things, but putting those on a good desktop amp can result in much firmer control. You can hear that. And the circuit of the amp as well as its amplification devices lend a particular sound, as well.

 

[wankski]

to answer the question if you connect your iPod to the desktop amp thru the headphone jack yes: a high gain desktop amp will also amplify the signal noise and essentially exacerbate the poor sounding source.
 
but with iPod you can sort of bypass the ipod's integrated amp using a LOD - line out dongle. You would then line that into your headphone amp to bypass the iPod's amp...
 
there is a fair bit of difference b/w the low cost iPod amp circuit that would likely feature a single low cost opamp to drive the stereo output compared with something portable like amb's credit-card sized mini3 that features quality analog devices/burr-brown or national semiconductor opamps x3. 1 for each channel plus an active ground to 'sink' the return current floating around the circuit.
 
anyway you cut it it's a superior amp to most DAP's internal amp circuits whether or not you need the extra voltage for higher impedance cans...

 

[Uncle Erik]

That clears a few things up, Erik.


 


On that note though, if I'm using the iPod as the source - wouldn't its amplifier have it's own impedance curve which will ultimately affect the curve of an external amplifier as well?


 


I guess your source is quite important.

You can bypass the iPod's internal amp by using a LOD. If you connect it through the headphone jack then, yes, its output impedance will interact with whatever it is fed into.

 

[xnor]

Quote:

You can bypass the iPod's internal amp by using a LOD. If you connect it through the headphone jack then, yes, its output impedance will interact with whatever it is fed into.

 
Something I want to add:
Everything has an output impedance, including the line out (dock).
 
What matters is the ratio and variations. To give an example:
100 ohm output impedance + a 16 ohm in-ear can be very bad for frequency response, especially if it's a multi driver earphone
but if you connect 300 ohm headphones or a 10+k ohm line-in the output impedance doesn't matter much anymore
 
Ime, portable devices can't afford to have a 100 ohm output impedance, so usually it's much lower. Around 20 or lower is a more realistic value.
 
Now the funny thing is that line out's usually have an impedance of around 100 ohms but I don't know if this is the case with those ipod LOD constructions..
Measurements are sparse on that matter. If you know of any, please let me know.

 

[Anaxilus]

Quote:

I have never heard a headphone amp that made the sound quality any better than an ipod or computer motherboard.  I have heard some old receivers that had really bad sounding headphone out, but in 2010 headphone amps are not really needed anymore.  The only exception is if you enjoy spending lots of money on "hifi" gear that in reality doesn't do much.  If that sounds like its for you, then go buy a tube amp!

Hate to differ with you but everything from my DBA-02 to the D7000 sounds clearly better w/ an Arrow than just a Clip+.  The improvement in bass control alone is quite audible on both phones.  I wish I could just sell all my sources and my amp and live off the Clip+.  I guess my ears just like wasting money as opposed to my head that wants to save it.  Hopefully you aren't confusing improving the performance of a phone w/ the improvement of a source.  Although it sounds like you are.