iTea

[mishalak]

Bit of a weak pun there, but it was the best thing I came up with when my iPod asked to be named. I now own one, obviously. Early birthday present from my parents. So for some reason instead of my usual very utilitarian names for things I came up with this silly one (thanks Cat for inspiration).

Now I'm replacing a decade worth of .mp3s I ripped using winamp. I suspect the .aac format to be of a bit better quality. I know that it is proprietary and all, but I don't care, I expect by the time I replace the iTea they won't be that proprietary anymore. After all just the other day Apple announced it would be offering DRM-free music on the iTunes Store from EMI.

So I'm listening to the music from Waking Ned Devine as I figure things out.

Comments

[bovil.livejournal.com]

and yeah, EAC and LAME are a bit more work when you start doing new CDs, but the rewards are worth it.

[mishalak.livejournal.com]

Enlighten me as to the rewards. What makes variable mp3 so good?

[bovil.livejournal.com]

iTunes can do VBR, it's just a matter of settings. The rewards aren't just from VBR, it's a combination of things.

EAC has some really sophisticated error-checking functions that deal well with some of the "slack" in the audio CD specification. A perfectly mastered audio CD ripped on a perfect burner will always produce a stack of .wav files that's identical to the last rip, but we know how well "perfect" and "technology" go together. Factors including "gapless" CDs, mastering errors, dirty discs and player errors can all contribute to subtle (and in some cases, unsubtle) problems.

Once the .wav data is ripped, it needs to be compressed.

Now VBR is nice; it sets a "floor" bit rate for the least-complex groups of data frames, but increases the bit rate to provide more fidelity in complex frame groups. More detail where needed, less total storage than constant bit rate. This is a big deal in classical music (with its wide frequency range), but isn't inconsequential with any music.

Where LAME wins out is in its multi-pass compression process. It's more discriminating about what audio data is discarded during compression, providing a more "true" representation than other codecs at the same bit rate. It does a better job of adjusting the variable bit rate on the fly, again providing a more "true" representation in a smaller file.

That's not to say that other MP3 codecs couldn't do the same thing, it's just that many commercial developers choose to be a bit more sloppy.

According to the graphs, AAC is more efficient and precise overall. MP3 is universal, though, and newer MP3 codecs are approaching AAC in quality/compression.