Yeah, I've read the r3mix site, but it's old information and I have a newer problem.

On the PowerMac at home, iTunes presumably has some variant of the Fraunhofer encoder built in. It appears Apple did a bunch of performance tuning because it's screaming fast and it uses both of my CPUs, yielding a net compression speed of something like 10x. Lame, on the other hand, is blindingly slow in comparison (running at 1.5x real-time per CPU). Plus, the hack to use lame in iTunes is neither smart enough to rip in advance nor smart enough to run multiple lame processes concurrently.

So, my question is whether anybody has the appropriate analysis tools (i.e., the frequency response graphs from the r3mix site) to compare the MP3 quality of the current Fraunhofer codec (as used to by Apple) to Lame.

Or, just in general, is the quality of 'lame --alt-preset standard" noticably better than the "best" VBR quality out of iTunes? If nobody out there knows the answer, then it looks like I'll have some experiments of my own to perform...