diff options
| author | Eugene Sandulenko | 2005-10-19 04:59:18 +0000 |
|---|---|---|
| committer | Eugene Sandulenko | 2005-10-19 04:59:18 +0000 |
| commit | 4306c9344cd00edff5b9a04152d68a1a7b0862aa (patch) | |
| tree | 02978ce48b40c00efe4be9b5fc85e7a68edd4114 /saga | |
| parent | 13246d0dadd236fb36eeeb12228e53101cdabc43 (diff) | |
| download | scummvm-rg350-4306c9344cd00edff5b9a04152d68a1a7b0862aa.tar.gz scummvm-rg350-4306c9344cd00edff5b9a04152d68a1a7b0862aa.tar.bz2 scummvm-rg350-4306c9344cd00edff5b9a04152d68a1a7b0862aa.zip | |
Improved IMA ADPCM decoder. It appeared that MS violated yet another standard
and nibbles order in samples appeared to be swapped. Had to untemplate
whole thing over again because I have no idea how to speicalize one of two
parameters in templates.
Now voices are clean but have some ticks, looks like overload. ITE wasn't
broken ;)
svn-id: r19166
Diffstat (limited to 'saga')
| -rw-r--r-- | saga/sndres.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/saga/sndres.cpp b/saga/sndres.cpp index e508a00673..e84ea60242 100644 --- a/saga/sndres.cpp +++ b/saga/sndres.cpp @@ -216,7 +216,7 @@ bool SndRes::load(ResourceContext *context, uint32 resourceId, SoundBuffer &buff buffer.buffer = NULL; free(soundResource); } else { - voxStream = makeADPCMStream(readS, soundResourceLength, kADPCMOki); + voxStream = new ADPCMInputStream(&readS, soundResourceLength, kADPCMOki); buffer.buffer = (byte *)malloc(buffer.size); voxSize = voxStream->readBuffer((int16*)buffer.buffer, soundResourceLength * 2); if (voxSize != soundResourceLength * 2) { |