I've downloaded the whole iTunes beast (just for testing the encoder), encoded a few (problem) samples, and after my ears were ruined I've promptly uninstalled it. (there was e.g. very noticeable pre-echo at 320k and the quality of VBR encodes was not even comparable to Lame). No, thanks.J.M.(edit- hope i'm not violating T.O.S - the difference was really obvious)
The quality settings only determine the average bitrate bracket that the encoded files will belong to, and don't correspond to more advanced psymodels or anything. The number in the bitrate dropdown is the bitrate floor, so having a '192' kbps iTunes VBR MP3 compared to say preset standard will be very inefficient, as it will not use anything less that 192 kbps for easier passages. Think of the quality dropdown as the LAME '-Vx' equivalent, and the bitrate field as the LAME '-b' equivalent.
The VBR dropdown menu has been there for as long as I can remember, and was used during at least one of Roberto's multi-format tests (I think he later said he shouldn't have used it). I know it's common sentiment to say that the iTunes MP3 encoder is terrible but I sort of wish it could be tested at higher bitrates.
If someone has a few old versions of iTunes around, can you please encode some identical content to MP3 using the same settings and "prove" the encoder and output is different. ABX results would be nice, too.For the people that really want to how "how it competes" with LAME, just go ahead and test it yourself. 'nuff said.
For the people that really want to how "how it competes" with LAME, just go ahead and test it yourself. 'nuff said.
To me iTunes does a decent job when reconverting Apple Lossless to VBR 192kbps with High selected but with CD Audio it's terrible