Quote:
You guys ever tried an abx (blind test) using Foobar 2000 abx plugin (volume matched) .... or is it just a feeling ...... ?
Most people who say they can tell the difference between aac256 vs 320mp3, or even aac256/320mp3 and flac, have never actually tested themselves under proper conditions. The results are usually ..... enlightening.
+1
I was quite surprised when I realized that aac192 was my transparency point. That was back in 2009. Since then the QT implementation of AAC seems to have improved. Granted, I'm using
qtaacenc as my encoder backend (which allows VBR encoding, unlike the iTunes frontend), but my transparency point now is somewhere around ~172kbps on average. I've got a few problematic albums or songs (for some reason Quadrophenia seems to be giving the encoder trouble, and when I have the time I'll test with it to see if I'm actually hearing artifacts), but generally even trebly, overly bright recordings (like pretty much all of Yes) come through fine at my settings.
The last time I tested MP3 was the summer before last, and using LAME my transparency point is still V0 (~245kbps, I believe that comes out to), like it was the first time I tested. That's with my own personal killer sample, though, which is a really difficult part of one of my favorite songs. Usually I find V1 and even V2 to be perfectly listenable on most material. I use AAC to encode based solely on these results, since it's clearly a lot more efficient, at least to my hearing.
BTW, this is what that quote in my signature is about. I know what I can't hear, thanks to my testing. It honestly makes audiophile life a hell of a lot easier and more enjoyable when you're not constantly worrying that every little "off" thing you think you hear might be an artifact.
EDIT: For those interested in testing this for themselves, AFAIK the ABX component no longer comes by default with Foobar. You have to fetch it separately on the components page.
http://www.foobar2000.org/components