That one was a bit trickier than expected; a bug in max_seeds causes the psychoacoustics for very short frames to discard bin 0 in masking, causing residue dynamic range overload. The bug triggers at about the same point the frame size makes frame mismatches audible... Just to avoid alarming anyone, the bug doesn't affect the bitstream format; the encoder simply screws up the near-DC amplitude level. I'm a bit burned out tonight, a fix will get checked in tomorrow. Monty --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/