I am looking to improve my OGG Vorbis encoder that I've written and I have two questions I'd like to ask you guys:

1. Is there any way to determine what the output peak value of a block of audio you've just encoded is (in LibVorbis) - so that after encoding, the program can add a gain tag to ensure a track peak level of x dB?

2. A PCM track that has a peak value of 0 dB often ends up having a far higher peak (+2 dB or worse) after being encoded (measured in Goldwave, etc. after using OggDropXP and my own program). What causes this exactly? If it's because of some sort of quantization going on after the MDCT is there anyway to predict what kind of variance over peak/avg track volume there will be?