The "mp3s" sold by Apple, for example, are 256k AAC files. And I haven't seen any evidence that human beings can hear the difference between them and their source masters. I'll consider it interesting if you can show me a group of audiophiles who can do this. I'll be humbled--and admit I am wrong--if you can show me that a majority of average music listeners can do so. Frankly, I bet you yourself can't do it.
I'm confident that most people on Talkbass could hear this difference given the proper source material. I was confronted with this recently myself.
Anyone here can do this test providing they have the right instruments.
I was making a recording of my carved double bass using a Zoom H2 straight into Apple's Garageband at 16bit / 44.1kHz via USB. The H2, has stereo microphones, mic pre-amp, ADC converters and recorder all in a pretty case for $200. It's not expensive and not "top of the line" anything, but it's very handy.
So recording into Garage Band I get used to the sound of my double bass. Do my tracks as best I could (frustrating, as there is significant latency in the H2 that at the time I thought was me. "Why am I playing so late and out of time?") Anyways, I'm done in Garageband and send it to itunes as AAC at 256kps. Listening it in iTunes I'm shocked at how unclear the bass sound is compared to Garageband. So I re-export from GB to iTunes as AIFF and re-encode as AAC at 256kbps in iTunes 9. Still same sound degradation. I want to send the file to my friend at a higher sounding quality but in a reasonable way. So, I do it via Bandcamp. (I imagine the difference would be more noticeable had I used the better mics, mic-pre and ADC of a regular recording studio.)
Neil should use Bandcamp to reach his customers.
The main things is you need a good source. To me, any instrument where the sound comes from a speaker cone (ie. electric guitar, electric bass) is already too altered to be good enough as a source used to detect alterations in sound data storage. Acoustic instruments only please, and make them good ones. People can hear the difference between a $100 plywood acoustic guitar and $700 acoustic guitar with a solid wood top even though they are both guitars and one can play the same music with them. It's similar with lossy compression and uncompressed.
So anyone here can test this by recording a quality acoustic instrument as best they can into their computer. Don't clutter the recording with excessive overdubs or instrumentation that bring in their own frequency masking. Export the project as AIFF/WAV and 256 AAC and listen. I'm sure we'll all hear the difference that comes solely from the conversion of uncompressed audio to compressed audio.
Now if our projects are all about noisy guitars, digital effects, drum machine sounds from the 1980's, crazy EQ's and plugins, then it really gets hard to hear the difference because the sound source of the music itself, is so low. But if the project is all about one person playing an amazing acoustic instrument in a great sounding room, then it is easier to hear the difference.
==
I agree with the rest of what you wrote.
==
for those who remember, iTunes stared off selling m4p files at 128kps when the store opened in 2003. The move to m4a 256kps files happened in 2007 with the brand name of iTunes Plus. Previous purchasers could upgrade their purchased files for a small fee. Who knows, maybe the same will happen in a few years. But with the amount of songs sold off of iTunes growing so fast, a future upgrade of heavier files would be a massive amount of traffic for the iTunes servers.
