Lion

Almost, Lame 3.87 CBR Overall Delay is 1105 samples in table, it seems changed in lame 3.99.5. Lame CBR injects "Info" block in first frame, which affects resulting `lame --decode`...

Here latest lame decoder always compensate 529 samples: https://github.com/lingfennan/mp3lame/blob/master/lame/frontend/get_audio.c#L553 And additional encoder delay encoded in Info/Xing header: https://github.com/lingfennan/mp3lame/blob/master/lame/libmp3lame/VbrTag.c#L831

Hi) About DRMP3_MAX_FRAME_SYNC_MATCHES: mp3 sync points are very weak (much weaker than MPEG`s ones for example) and false positives can occurs. To make search procedure stronger, MAX_FRAME_SYNC_MATCHES frames is checked,...

So this is tag removal issue I mention above. This is tricky part to handle since mp3 can have: * id3v1 tag * id3v2 tag * APEv2 tag * VBR...

I`m working on opengl gui version of far2l and have same thought, but currently no ETA. As temporary solution you can assign fast viewer to images (Commands->File associations). I recommend...

I need this too. May be stencil buffer can be used instead of render to separate FBO? At least without AA.

Good point. I've added self-generated vectors which I use for performance testing https://github.com/lieff/minimp3/tree/master/vectors/performance . This is public domain. As for others, most of them from testing procedure supplied with ISO...

About .bit files format: ILL* - mixed layer 1\2\3 l1_* - MPEG layer 1 l2_* - MPEG layer 2 l3_* - MPEG layer 3 M2L3_* MPEG layer 3 MPEG versions...

@SirWindfield For ILL* and \*nonstandard* it's normal, they broken to trigger bugs in decoder. Others should be decoded\played normally. ffplay can play them, (but showed layer 1 and layer 2...

ILL* vectors is from Philips MPEG Audio Test Suite http://www.mpegmultichannel.free.fr/Mpeg.pdf