BTW I know this has been answered many times before but I feel the need to reconfirm this.
What exactly does an amp do? Other than providing the potential to drive cans at larger volumes? And is there a way to scientifically prove these effects?
And oh shoot I have one more question: does impedance affect how "much" it improves the headphones? Is there, for instance, a direct correlation between the improvements of an amp powering a 15 ohm headphone, as opposed to a 600 ohm headphone (other than perhaps, being able to sufficiently listen at preferred volume).
There's quite a lot of info, properly backed up, on this topic at nwavgy's "More Power?" page. Google for the two terms: I can't post a link.
There's a lot of talk of amps 'improving' the sonic quality, but really that shouldn't be happening: they should only be increasing volume. Unless there's a problem....
High output impedance of an older source, ie older ipods/iphones (or not so old Galaxy SII) is perhaps the most common. A decent amp can correct that.
In some high end, low-impedance, high-current headphones the source can distort as its basic amp is overloaded. An amp with adequate power can avoid that.
But really, perfect hifi amping has been a settled technology for more than 20 years and is not expensive. Expensive amps are more marketing than reality. Solid State changed the world. It's even got to the stage where talk of class A or class AB is absurd: a properly made, modern class D amp is much cheaper and genuinely hifi (and cheap on electricity). No one needs class A amps.
So why, then, are amp reviews so often 'expansive' beyond what the above can explain? It's as if the reviewer is saying that the amp adds to the audio, or improves it. Apart from the above issues, that shouldn't happening. Unless there's a problem...
Very few know that amp tech is settled and cheap. Magazines certainly won't tell you. Some premium amp manufacturers build in a bit of distortion increasing with cheapness of the model. Can you imagine why? It's sometimes justified with "2nd order harmonic distortion improves audio", which playing with a multi-order distortion plugin on my laptop: I can tell you it does not.
No wonder then, that old fashioned tube amps don't sound so bad, even though they also distort: the competition is worse. Some even have issues that give a 3D effect (which the reviewer will experience as a 'bigger soundstage'). In the headphone amp world there is another factor: consumer amps based on popular but flawed DIY designs (Mini3 being an example of one of those designs). These are very common, as are dodgy DIY amps sold on ebay.
People moving from one of these amps to another are going to experience different sonic qualities, sometimes improvements. In the highly subjective world of audio all this passes for normality. And all this chaos is further obscured because there are legitimate sonic improvements due to the common issue of sources having high output impedance.
The O2 was an attempt to address this by showing that you can build a perfect ('wire-with-gain), powerful, portable amp for $70 in parts. It's been an anti-climax because there is nothing for the reviewers to say beyond: it just increases the volume. Which is exactly all that it should do.
To summarise: source output impedance aside, not many headphones actually need amping. You can tell if they do simply because you can't get the volume.
I wish it was as simple as that, but with those rarer low-impedance, high current phones (ie. LCD2s, HE-500 etc) potentially causing distortion: how do you measure that as an end user? You'll be trying to guess whether your audio might have distortion that might be fixed by amping.