The decisions usually take about a couple seconds. I'm either very close to 100% accuracy or completely random based on the program material. At higher quality settings, it's very hard to hear the difference (in a blind listening test), if you can hear a difference at all.įWIW, I have actually done blind listening tests on systems ranging from decent RFZ rooms with genelecs to $15 sony earbuds. * Lots of people think MP3 "sounds bad", but most of these people have not done blind listening tests. You don't need 320 kbps to compress silence.) It's probably overkill but I'm not worried about disc space.
#LAME MP3 ENCODER 3.100 ALPHA 2 320KBPS#
V0 only goes to 320kbps when the algorithm thinks it's necessary. (At low bitrates, the sound can get pretty bad.) If you hear a compression artifact at the higher-bitrate settings it's usually related to the basic MP3 compression algorithm itself and it usually doesn't get better at an even-higher bitrate. It doesn't necessarily require a high-quality sound system to reveal compression artifacts. It turns-out that the program material and the listener's ability to hear artifacts are most important factors. (With VBR V0 is the "best" and V9 gives you the smallest files.) If it sounds identical to the original, you can't get "better" than that. With most audio files you can use a lower bitrate setting or a variable bitrate setting and it will still sound identical to the uncompressed original (in a scientific-blind ABX test*). No matter what settings you use, it's lossy and every sample is altered. And, if you don't care about file size there's no reason to use a lower bitrate.īut, we can only say it's "better" if it actually sounds better. If you are "paranoid" but you want to use MP3, 320kbps joint stereo is the "least lossy" setting.
0 kommentar(er)
