I strongly suggest that you conduct blind ABX tests to determine that 320kbps CBR is what you need for your lossy music as many people, including myself, just don't hear the "benefits" when going from -V 0 up to 320kbps CBR.  In fact, many people perceive that Lame at -V 5 (~130kbps VBR) is transparent for them.

There are some forms of music (mostly classical) in which I simply can't listen to in lossy formats... but if I must, I have very high standards (only 320kbps w/LAME CBR).

Doesn't your curiosity extend to at least ABX testing LAME in VBR mode to see whether it meets your needs rather than just going for CBR320 without knowing whether you need to or not?

Cheers, Slipstreem. 

That way you can further solidify any audio quality statements that you do make by showing us your blind ABX test results (otherwise audio quality claims violate the terms of service here on hydrogenaudio).

Basically, it's more of a psychological thing... Even if my ears can never tell the difference I still feel as though there is one.

Quote from: Kratos on 2009-02-08 15:16:08

Basically, it's more of a psychological thing... Even if my ears can never tell the difference I still feel as though there is one.

And, frankly, the best way the "break" that line of thinking is by doing ABX tests. Select a handful lossless tracks you're extremely familiar with, encode them to 320kbps, V0, V2, and also to something much lower, such as V5. Start by trying to discern the lossless file from the 320kbps encode, then go down to lossless vs. V0, lossless vs. V2 and so on. See at which point you can start discerning differences, if at all (but take occasional breaks when you begin to feel fatigued). With LAME, you may even find that V5 is transparent for you with the handful of tracks you're most familiar with.

I used to be feel the same way about lossy as you do. I only had relative faith in bit rates higher than 192kbps, and I was certain that anything lower than that was universally terrible. I also always thought that I was missing "something" regardless of how the bit rate was, and that something about lossy encoding in general was "wrong". Then I participated in the 64kbps multiformat listening test last year, and all of my erroneous thinking flew right out of the window. Now, no bit rate is universally "too low" -- it's only what's transparent, not transparent but still adequate for the intended purpose, or both not transparent and inadequate. There's no more "this sounds good, but there must be something wrong with it" feeling any longer.

To break the psychological boundaries that don't allow you to ever trust low or moderate bit rates, you need only to compare lossless and low/moderate bit rate audio in a double-blind environment, finding out for yourself how low is "too low", and making all of your future encoding decisions based on that.

Personally, I use -V4 for portable use. It provides small files, and the artifacts are not particularly objectionable to me, if I even notice them. In a noisy environment, outside of blind testing, I simply don't notice the difference. Even in a blind-test environment, I frequently fail to differentiate the two on a random subset of my collection.