For me the major change was when I did my first blind test. Quite a while ago I started experimenting with ripping all my CDs to FLAC, and then lashing together a mess of software, long cables, IR controller etc, to send the sound from my PC in another room to my preamp. I did this mainly to play with the software, and with the intention of using it for background music sometimes, always using the proper CD player for real listening. After it had been up and running for a few weeks I found I was using it more than I thought, but had a creeping impression that the music sounded slow. I was playing the music from aac files (the flac files were on an external backup disk, I was using the motherboard soundcard, and a 10m phono cable. So although I assumed it was the motherboard, or the long cable, I got curious to do a test of the flac vs the aac using foobar’s comparator. This was the most illuminating thing I ever did with audio. At first I could not hear the difference, but I persisted, tried to quiet my mind and just listen, and then there it was, each time I clicked on the aac it did sound slow. I was using the Nero AAC encoder as it was what came with dbpoweramps batch converter. I tried mp3 vs flac, mp3 was ‘messy’. I should point out I wasn’t guessing with these tests, I could tell the difference statistically, and after I had tuned in I knew I wasn’t guessing. I then tried using the apple AAC encoder in iTunes+ mode, hmm, now I was guessing. I remain confident that I could have passed this with more work, but the key for me was I had found a lossy codec that worked for me, and the difference if it existed was very small, and the double blind part totally re-calibrated how I assessed all future comparisons of kit, the question “how confident am I that I could ab these blind?” is the key question I ask myself, and the answer is usually not very.