[FlameGrilled]

A billion tests proves crap all, since the only thing that matters is finding a balance between compactness and quality for whatever format you use - if space/storage is a sensitive matter.

As far as quality goes - whatever gives you the results you need, allowing the encoder to allocate as it needs to regarding the content. If you are happy with the results as you get them.. noone else's two-penneth over whats sounds best at xxx kbps is highly redundant.

Look at it this way - if you were listening to a netcast, a classic bit of narrowband encoding, you could choose not to use it because the quality is poor and that's fair comment. But to then make your decision to listen on the basis of what codec is used and which sounds best at the kinda bit-rate used, is being a bit AP over something that you can take or leave as it suits you.

There are so many elements to encoding, reasons for doing things in popular ways and unorthodox ways, that this testing to work out which encoding sounds best at xxx rate is getting way past of being any real value - just a bunch of codec audio devotees doing their quasi-audiophile bit time and time again.