The high-resolution videos play back at much lower volume than the
original ones. This adds hard-coded values for how much to amplify
each cutscene. It's all done by ear, and it does introduce some
clipping, but I think it should be acceptable.
Of course, it could also be a problem with the audio decoder, so
this may be the wrong approach entirely.
In all my attempts to get the audio and video to sync up in the
ZGI videos, there have always been 9-10 frames of video before the
audio even starts, even though the audio is timestamped to start
before. This attempts to fix that by prioritizing sending audio
packets to the decoder in a timely fashion.
I do not know if this is the correct way of doing this, and there
are still some things that need to be fixed. But pragmatically, it
does procude significantly better sync, so...
Sometimes (only at the very start of a movie?) there will be a
video packet that has a PTS but no frame to display. Save that PTS
and use it for the next frame. This doesn't actually improve
anything, as far as I can tell, but feels right.
This code comes from clone2727's now defunct (?) ac3 branch, with
some minor compile fixes. This represents the latest version of
the stalled AC-3 decoder work for Zork: Grand Inquisitor. Note,
however, that I have not yet asked for clone2727's permission to
use this. I'm just experimenting.
