DOLLS wrote:Gil_Hamilton wrote:Also: Raw numbers also don't really give you an idea of what it SOUNDS like('s why every 5$ pair of headphones has really impressive-looking numbers on the back).
These numbers actually do, it's an objective measurement.
Especially not when they're generated under a loadless situation, which is totally unlike the real world, where you have, you know, headphones attached.
iPods can't drive most great sounding headphones anyway, almost no portable player for that matter.
I'd run the line through a portable headphone amplifier if I was so inclined, which puts considerable less stress onto the output circuit, or I'd plug it into my receiver which itself is connected to a preamp then to a power amp and finally to a couple of great sounding speakers, which puts the device in a similar situation where load is greatly diminished... In the worst case, I'd get a pair of good sounding low-impedance headphones knowing that they'd lessen the impact beforehand.
Which nicely evades the issue that your results are utterly meaningless, as your second link showed even the stock iPod ear buds do very bad things to iPod audio, or at least, to the bass portion.
http://hifiipod.co.uk/?page_id=52
Check out what happens when the iPod is actually driving headphones. The numbers change radically.
And that's the ONLY loaded test he's made.
The only considerable alteration being that of the stereo crosstalk measurement, which is not too bad an impact since it consists on a bit of channel crossfeed.
But your other link shows square waves crashing precipitously on TWO iPod models.
So who are we to believe?
I propose that BOTH are right. But the numbers presented don't paint the whole picture. It's a common problem with audio stats.
As is noted
here, where he notes the flaws his numbers hide.
Also, the numbers show 20x variance on "IMD + Noise", over double the THD, and "frequency response" skews 5x in one direction and 23x in the other. That's NOT a considerable alteration? Dynamic range and noise level are the only 2 that DON'T show a large change.
Also note that frequency response shows the peak and bottom frequences attained, but not how stable they are.
I gather that the iPod has a lot of variation across the board. I KNOW the bass collapses under load, though it doesn't show in frequency response tests, because there's still SOMETHING there.
He conveniently omits the response curve for his single loaded test. And
here he makes reference to a performance issue in his then-unpublished G4 review. The PUBLISHED, review, of course, contains no such comment. And omits the response curve that would show it.
I didn't say every iPod was great sounding, I just pointed out that such a generalization was bullshit.
A generalization I identified as a generalization. I didn't say every iPod sounded like crap, just that they have a reputation.
And you haven't really made your case, except in the specific instance of the Shuffle.
Which admittedly surprises me greatly, but I've seen it backed up several other places. How they can pull it off on the Shuffle and NOT the larger players is baffling.
snkcube wrote:DOLLS wrote:SPEAKING of headphones... no iPod yet made supports an adjustable equalizer, which is NECESSARY to compensate for headphone response variation. Different models of headphones respond quite differently to the same input(exaggerating some frequencies while damping others down), and without an EQ, there's no working around that.
No EQ = auto-fail.
FYI, those new iPods that came out does include an EQ. I checked on my friend's iPod Nano.
sweener2001 wrote:it may be a different EQ that he's referring to, because my 4G had EQ. and 4G's are OLD.
USER-ADJUSTABLE EQ.
Not predefined "rock", "pop", and "ZOMG BASS" settings. One of those things with a couple of sliders on them.
Even the 5th Gen doesn't have one. And given that it's been on the top of feature request lists since the FIRST iPod, I doubt the new crop have them either.
On top of that, the predefined EQ settings have been known to cause significant distortion all through the line, making the feature useless even if it DOES happen to offer a setting that more or less matches your hardware.