While listening to the first couple of clips at http://www.ff123.net/training/training.html I noticed that at least the four first clips (the ones I have tried yet, applaud_radium128, applaud_insane, castanets_xing128 and castanets_insane) display wrong bitrate. Both 128kbps samples report as 129, applaud_insane reports as 322 and castanets_insane as 323kbps! As I said, it's not a big thing, but I thought I should mention it. Oh, I almost forgot, Winamp shows correct bitrates.

are they VBR, do they have VBR header ? (hmm dont look like it, so im looking at them. it may be they have bad xing headers for no reason)

Nope, they're all CBR.

They're mp3s with RIFF headers, foobar maybe taking the RIFF header into account when calculating bitrate ??

i dont know what else it might be..

edit: it still shows 129 without the RIFF header.. hmm.. so theres something else wierd about them, but I dont have any time to look at them now..

I think foobar reports bitrates of really short mp3s inaccurately.

i tested with some other mp3s and came to the same conclusions. e.g. 15 frame 192kbit/s reported as 222kbit/s. 6 frame 128kbit/s reported as 192kbit/sec.

its like its missing a frame in its calculation or something...

Wow, I think that makes it first "bug" I found in foobar

I wont call it a bug, its like if you calculate an integral, if you have much steps for calculating, then you get more accurate results than if you have less steps. So if a file has f.ex. 10 frames then its difficult to calculate the correct bitrate. Or?

are you sure ? im not saying im right, but the way I was looking at it is if foobar is miscounting the no. of frames in an mp3 by 2, the magnitude of error will increase, and only be significant with few(er) frames.

and (128*6)/4 = 192

(192*15)/13 = 221.5

Im not saying im right, but that's what I thought. But I dont particulary care anyway, I dont have any 15 frame mp3s.