(I know this is a long post, but quite possibly worth it if you actually want to know the real causes of poor sound, etc.)
The problem is (other than trying to fit a triangular shaped object in your pants pocket) that the BIGGEST problem in music sound quality is at the mastering stage. You can have the best HiFi system in the world (mine cost over $6k here when I first got it and that's pocket change compared to stuff out there) and having to play an album that was basically mastered for FM radio or cheap car stereo systems or boom boxes in the late '80s or even were more or less straight vinyl masters transferred to CD without mixing the EQ curves back (result NO BASS; many early to mid'80s CDs were a victim of this and had awful bass response as a result) resulted in horrible sound quality.
But instead of blaming the mastering engineers and the record companies for ordering them to do it that way, they blamed the CD format itself. Redbook CD may be OLD, but it's still capable of sound nearly as good as even the youngest ear can hear and don't let some self-declared audiophool tell you otherwise. Anyone who knows anything about the SCIENCE of digital sampling will tell you that increasing the sampling rates will get you absolutely ZERO in terms of increased audio fidelity because the human ear can't hear that high. Yes, there's some advantage in starting at higher rates because you can make simpler, less strenuous audio filters, but the truth is things like simple oversampling solved this problem a LONG time ago and have only improved since then (Sony 1-bit, etc.).
There's some small benefit to increasing resolution from 16-bit to perhaps 18-bit theoretically as the ear can hear a full 120dB of dynamic range, but there's two problems left. One is that no one in their right mind that wants to keep their hearing would want to listen to recordings with actual 120dB range in them and listen to them at levels that you could hear it all (i.e. the loudest sound would actually have to be 120dB loud and your volume would have to be adjusted such that it would actually play back at 120dB in the room. That's like standing next to a jet engine at take off and the plane not moving. Your hearing will be damaged within 3 minutes at that level. You would never want more than an instant that loud and even then it would be way too loud for most people. 114dB is typically as loud as a movie is ever expected to get (120dB is 30x louder on a linear curve, which may not sound a whole lot louder to the ear, but is doing 30x the physical damage to your eardrum). In reality, most home theaters are typically calibrated for around 112dB maximum volume and even that is pretty scary if it goes on for any length of time. A typical "loud" listening volume in an otherwise quiet room might peak at 96dB A-weighted or 108dB C-weighted (bass carries a lot more punch and the ear far less sensitive to it).
The other issue is that 24-bit is pure overkill on the playback side (no one can hear 24-bit worth of resolution, but it IS VERY handy on the recording side since it's very easy to record without having absolutely perfect recording levels for maximum dynamic range. In other words, its purpose is headroom for recording, but the final mixdown will be just fine on 16-bit. In the car (and worse yet headphones) things might end up being louder because you're competing with other loud noises in the environment and they create a masking effect that make it hard to hear music detail so people turn the music up. As such, I believe headphones are the number one cause of hearing damage short of something like being around gunfire. I can't tell you how many people have hearing damage from wearing the old Sony Walkman headphones (horrible open-ear things to begin with).
In any case, 24-bit playback is unneccessary and doesn't actually help anything except for the marketing departments trying to convince you that bigger numbers somehow equal better sound. THAT is the #1 reason people like Neal Young think higher resolution playback is the key. Frankly, Neil's ears are so damaged by years of playing live and well AGE that I don't know how the hell he thinks he should even have an opinion on high fidelity sound. The real reason is that most of the "high definition" formats out there like SACD or DVD-Audio *SOUNDED BETTER* is that the music companies created another MASTER (i.e. re-mixed the album to sound good rather than sound LOUD on radio, etc.) for that format. So yes, that version sounds much better on a high end playback system, but it has NOTHING to do with the SACD disc itself. That same master mixed-down to CD levels would sound virtually identical. You only have to have a few examples to hear it for yourself to know this is true.
I make my own albums. I record in 24-bit. The final 16-bit mixes sound identical to the final 24-bit mix. There is no audible difference.
The biggest contributor to "bad" sound on the playback side are POOR TRANSDUCERS (i.e. speakers, headphones). It used to be that there were a lot of poor amplifiers out there as well that could be easily overdriven (into distortion) by a loudspeaker, but both loudspeakers and amplifiers have improved over the years to where this is unlikely to be an issue except in a CAR where there's a hard limit to the amount of power you can get out of a stock radio without a DC-AC-DC converter to increase the voltage output beyond 12-13V. If you have a very quiet car interior and efficient quality loudspeakers, it won't be a big deal. If not, a quality amplifier will make a real difference.
Quality speakers and headphones can't be stressed enough. Clean power to reach the levels you desire is a must as well (there are still esoteric high-end speakers that are hard to drive; my own Carver Ribbon speakers are only 8dB/1W@1meter from the factory and have a very reactive load. They are hard to drive like that. I have 250W on the bottom (10" subs) and 140W on top (Ribbons) PLUS and active crossover network that effectively gives me 90dB efficiency now and it's no longer a reactive load. In short, my speakers can now do 116dB if required with no problems. Any distortion from an overloaded speaker is going to sound worse than anything you might find on a typical pop record.
Most people don't want to spend the money required to get really good playback at home or in the car. Most people won't even spend a couple hundred bucks to get a really good pair of headphones (and I don't mean BEATS which are the new BOSE apparently for bang to the buck ratio in the lousy end of things). It's easier to blame the CD or the AAC formats than spend some money to improve things. Even so, unless the recordings released are properly mastered, you're going nowhere fast. The industry mixes for the lowest common denominator most of the time (unless you like mostly classical high-end recordings). They've at least fixed the bass problems and tinny highs over the years, but around the mid-to-late '90s they started destroying dynamic range to the point where there is less dynamic range on a modern rock recording than one from the '60s on a vinyl record by FAR. And this is not the CD or the AAC's fault. This is the fault of the record company and its mastering engineer purposely setting it that way because they know that psyco-acoustic studies show that people associate louder with BETTER, especially in a car environment where you're most likely to be listening to a radio in the first place. Most radio stations will further squash dynamic range (just in case it hasn't been already) so that volume levels are more or less equal for anything they play. You will notice that this is NOT the case when you play your own material.
"Remastered" typically means "LOUDER VERSION WITH LESS DYNAMIC RANGE" these days. One only need compare Ozzy Osbourne's "Crazy Train" song from the original CD released and the current one on his greatest hits album to see they've killed the dynamic range and introduced all kinds of overloading clipping distortion in the process. This is true of ALL the songs on that album, although some clip worse than others. Massive clipping is also found on the otherwise great album "Californication" by the Red Hot Chili Peppers. Sadly, this was on day one. These are just a couple of examples of what I'm talking about.
One of the best albums I have for sound quality on a pop/rock type album is Tori Amos' third solo album, Boys For Pele, which oddly was recorded mostly straight to DAT at a mere 16-bit/48kHz (save some studio drum work I believe). Despite there not even being a 24-bit master (there are two songs on the album where it clips just briefly, no doubt due to the lack of headroom when recording), the album sound unbelievably clean and life-like on a good system. Sadly, her followup album "From the Choirgirl Hotel" did NOT share this recording quality (studio squashed dynamic range). The fact her husband is a sound engineer makes it all the worse (someone else masters, though). Her newer albums are still squashed, but at least clean sounding. Her second album, Under The Pink, is also pristine sounding for the most part, although not DAT tape.
As for AAC, it has been proven in double-blind testing to be completely audio transparent at the 256-bit rate. I've compared it myself with lossless, CD and 256 AAC versions on my ribbon speakers and I have yet to hear a difference due to encoding in AAC. Again, I believe people are barking up the wrong tree and wanting to blame a format rather the masters because it's easier to blame a numbers game than something most people know nothing about and do not even consider at all. Sound quality starts in the studio (or wherever it's being recorded), after all and goes to mixing, mastering and ultimately to your personal playback equipment of which speakers are the most obvious cause of distortion and poor sound and often the least people bother with, followed by the listening room itself which can utterly RUIN the quality of good speakers if the room itself is terrible acoustically. But I've seen people stick speakers in the back corner of the room because it "looks better" there rather than where it should be for proper playback in front of the listening couch. And then these people wonder why it sounds so good at my house with a proper 6.1 theater setup downstairs (PSB speakers) and a high-end 2-channel upstairs (Carver Ribbons bi-amped with a custom active-crossover network).
Even so you can get reasonable computer speakers for $150 that sound pretty decent in a fairly small room. The Klipsch ProMedia 2.1 THX are an absolute bargain. I have two sets for two computes and one in my bedroom connected to an Airport Express for Airplay listening while sleeping, etc.
