MP3s are the GIFs of the ears.

Audio compression isn’t much different then image compression. Sometimes a compressed picture can be a fraction of the size and look as good as before, and sometimes a more colorful picture will compressed the same amount and look like crap. We have to realize that every beat doesn’t have any real significance it’s the bigger picture that counts. When we look at a web page we don’t take a moment to look at every pixel, we scan it for what we want. The same concept can be applied for music. I can honestly say I can tell a difference between 128kbs and 256kbs (65% of the time) but unless I am in a quite dark environment I couldn’t tell the differcne. I think that a better quality copy of a recording will amount to a better experience if the recording is of high quality, if the audio equipment is of quality, and if the environment and the mindset of the listener are in the right condition.

LAME also has optimizations in VBR mode to deal better with ‘transients’ btw. These are not new either, but were originally in 3.90.3, although limited to that version until the last year or so; as of 3.96.x they’re in by default and the –alt-preset (or any other preset) switch will simply default to whatever the comparable V setting is (V0-V2 are all acceptable replacements for 256-320k). LAME is and always has been suboptimial for CBR (320k CBR in other words) as most of the work that’s been put into tuning it over the last several years has been on its VBR mode.

Fraunhofer’s “improvement’s” are less well documented (and in many cases trade secrets) but CBR fares better there and is its continued claim to fame (especially for streaming data). AAC and other less known wavelet methods are again making strides over the ‘older’ tech, although modern mp3 codecs referred to here (not sure about Xing) are coming alone fine too. And then there’s Vorbis, which also sounds a lot better than it did several years back (Ogg is just a wrapper btw, like mp4).

CD’s and vinyl are largely fixed standards where the actual implementation is fixed (RIAA curve, redbook audio etc). Although the production and playback mechanisms are subject to refinement, outside of rediculously overpriced audiophile gear you’re unlikely to find anything available new that actually is refined.

Compare a higher end CD player from the early 80’s to one from the same (or similar company) today. You won’t even find a product category for that outside of an audiophile shop as it’s all based on the same highly integrated reduced set of cheaply made circuits, regardless of whether it winds up in a boom box, car stereo or standalone player for a component system. You’d have to go to the audiophile shop to find anything that has any real design put into it and that will last longer than a few years. The same thing for turntables, except you’re setting the benchmark for a ‘good’ mainstream turntable further back (late 70’s).

And their replacements will also be fixed, which gives the ‘industry’ a nice 10-15 year stepping between everyone ‘rebuying’ their favorite music in the ‘new’ accepted standard. Although I’m not sure that anyone really cares about DVD-A or SACD. Tape shouldn’t even be introduced in ’sound quality’ discussions imo, at least not the consumer implementation.

Mp3’s (and other digital audio standards) are not subject to these limitations and will steadily improve over time, while older versions will play back just fine without having to dig out that old 8-track or tape deck. Any attempt to compare a digital file format to a physical format is only relevant for as long as it takes the programmers to release another major stepping on the algorithm(s) they’re working on for the digital format.

Incidentally imho vinyl’s continuing existence is 2 fold, on one hand its lifespan goes back to the early spread of recorded commercial music and the physical form will hold up reasonably well as long as it’s properly stored. So my mom searches 2nd hand stores for listening to her ‘christmas library’ etc. On the other hand it’s the preferred method of distribution for indie punk/hardcore/rawk etc (7″) and dance music (12″). In the latter case certainly mp3’s would be preferred for ‘promotion’ and general (ipod etc) listening but most independant artists I know consider a vinyl release to be a certain benchmark. This is why they’re still done even though they’re quite often done at a loss of break even financially. Some will argue the ’sound quality’ point still but I always find that funny, considering that a 1200MkII isn’t a standard due to its sound quality. It’s built like a tank though for sure (and pitchable)…

MP3s have won the war of the formats because of technology, not because of their audio quality. “It’s like hearing through a screen door,” says neuroscientist Daniel Levitin of McGill University, author of “This Is Your Brain on Music.”

That’s why they put speaker grille fabric on speakers – to pre-slice the audio, so it will fit through screen doors.

Now, and already somebody has mentioned this, but i’ll state it again: 320kbps rips sound better than most CD’s, and often times converting a CD to 320kbps via LAME will even enhance the material. YES; the process of conversion from any format to mp3 does tend to “up the contrast” a bit — long, spacey reverbs do sort of fade away longer and fast transients are cleaned up where in the original they felt a little thick. The dynamics of mp3’s are often slightly more stepped, and the stereo field feels pushed apart… These are mostly good things! I watch my movies with higher contrast for greater impact, too.

As a producer, I’m happy to be able to rip work to 48khz, 320kbps mp3’s. They’re not nearly as bad as these weirdos with their broad generalizations seem to paint. Maybe they don’t realize they can actually have a bit more of the audio spectra recorded in an mp3, frequency-wise vs. CD’s?

The Devil is evil. As an expert, I find things that can be described as “sucking” or “good” are usually not facts. Also, I find that “difference” is not bad unless I am being a bigot.

Finally, much music has been created, recorded, encoded, and distributed digitally and what was left of the music moved my emotions quite vigorously. Perhaps emotions resonate to the song itself (good ol’ harmony, rhythm, melody, lyrics even) and not only the clarity or “goodness” of the sound that carries it. That would explain why our measurements of emotion are so varied.

And no format is perfect, even live. Unless you like destructive interference. And Jimi Hendrix’s guitar is kinda fuzzy.