I decided to purchase a 64 GB iPod Touch 5G based in part on the reviews in this thread.
I've since returned it. While it is a beautiful, ultra-cool, and ultra-lightweight device, it sounds downright nasty. The worst sounding Apple device I've yet heard, in fact.
Specifically, it is so harsh and bright, especially at higher volumes, it's downright unpleasant and bordering on painful to listen to. Basically the same impression I had of the iPhone 5s, which has single-handedly kept me from upgrading from my 4S.
I listen to a huge variety of music from the 60's through the 90's: jazz, country, classic rock, vocal/easy listening, metal, new wave, folk, etc.
I first tried my Grado SR125e cans with this unit. The 125e's are on the bright side at higher volumes themselves, and the combination was almost as awful as getting a large dose of feedback in your ear while standing in front of a PA loudspeaker.
So, I tried my Grado SR60i phones, which have less high frequency extension than the 125e's, and don't sound bright to me at all on other sources. Still very unpleasant, though not quite as bad.
Then I decided to pull out my venerable Koss PortaPros. The PortaPro is an easy to drive, great sounding headphone for a Walkman-style lightweight portable, and they have a response dip in the upper mid/lower treble range. In other words, they are the opposite of bright and can benefit from some extra brightness at the source.
Even with the PortaPros, it was unpleasant and quickly fatiguing.
As a last-ditch effort to find a way to keep the iPod Touch 5G welcome in my household, I tried the 10-band graphic EQ in the excellent TunesFlow app. I hate using EQ, but I wanted to see if I could get this thing to sound good.
With any of the above-mentioned headphones, it took reductions of several dB on the 4, 8, and 16 kHz bands to smooth it, and it still sounded unpleasantly shrill on some recordings that otherwise sound good. I chalk that up to the fact that EQ will only affect the QUANTITY of sound at a given pitch in relation to others, and not the QUALITY.
Worse yet, reducing the outlandish highs made the character of the bottom end more perceptible. Which is to say, even weaker than it already seemed. It sounded like a caricature of the stereotypical poor portable device headphone out.
I don't know where people have gotten the idea that this device has punchy bass.
And, I postulate that those who have raved about the "clarity" and "detail" of this unit fall into the category of inexperienced listeners who believe that exaggerated upper midrange = "crisp" and that exaggerated treble = "detail." Much like a newbie who perceives the sweetest wine to be the finest, or a video "enthusiast" who believes that cranking the sharpness control up all the way adds "detail" when in fact the exact opposite is true.
I'm very disappointed in Apple. The sound quality of the phones has gone progressively down since the original 3G, and the quality of the iPods has gone progressively downhill since the original Shuffle and 1G Nano.
ON ANOTHER NOTE... in reference to the dog-ear comments earlier and the question of why do you need to have a device that can reproduce well above the range of human hearing.
The answer is, that while frequencies above 20kHz may not themselves be audible, they do affect the tonal character of audible frequencies, and their absence is a big part of why digital audio at 44.1 kHz is often described as "cold."
It's called modulation.
If you have ever played with an old analog synth, or if you have an audio oscillator / tone generator, you will know that a 1K sine wave sounds entirely different from a 1K sawtooth wave, which sounds different than a 1K square wave. Download an app and listen for yourself. If you start adding VCOs and EGs to the mix (assuming you have an analog synth to play with) you will notice the tonal character of the sound changes even if the frequency is exactly the same.
It is the harmonics of the wave (i.e overtones, i.e higher frequencies) that change the sonic character of the "root" 1kHz wave.
The same thing happens in audio with ultrasonic frequencies. They modulate the upper audible frequencies much like an audio signal modulates a radio frequency in Amplitude Modulation broadcasting, and thereby they change the waveform and the tonal character.
In CD quality audio (44.1 kHz sampling) according to the Nyquist Theorem the highest audio frequency that can be reproduced is 22.05 kHz. It takes a minimum of 2 points to define a wave- one at the top of the wave, and one at the bottom. But, with only two points, all you have is a triangle wave, regardless of the complexity of the original wave. (And after filtering, you are left with a sine wave) So, you can reproduce frequencies up to 22.5 kHz in theory (in practive a little less due to anti-foldback filtering), but they are all going to sound the same regardless of all the complex detail in the original wave, and regardless of what instruments originally created that wave. In other words, with CD-quality audio, the upper AUDIBLE treble is literally reduced to the approximate sound of a test tone generator.
This is why 192kHz digital audio exists, and why high sampling rates and ultrasonic frequency response are necessary for digital to come close to approximating the quality of open-reel analogue tape, and why your device cannot possibly reproduce accurate upper-treble unless its response (and the source material) extend well beyond the range of human hearing.
