* Changed MidiDriver::createMidi so that if MD_ADLIB is passed, it
now *does* instantiate the adlib driver.
* Rewrote MidiDriver::detectMusicDriver (code should be clearer now,
and hopefully easier to adapt/maintain). Note that the behavior
changed slightly (this may require some tweaking).
* Added MidiDriver::findMusicDriver which works similar to the old
MidiDriver::parseMusicDriver but returns the full MidiDriverDescription
and ignores all characters in the passed driver name following a colon ":".
* Changed MidiDriver::parseMusicDriver to use MidiDriver::findMusicDriver.
* Changed game engines to match the changes made to MidiDriver.
* Modified SCUMM engine to not record the selected midi/music driver
(in _midiDriver), but rather the music *type* (in _musicType).
svn-id: r19859
changes. There are a few things that could use a bit more work, and I've
only tested it on my Linux box. I have verified that ScummVM still compiles
when it's disabled, though, so it shouldn't break anything too badly.
svn-id: r17512
- Replaces multiple identical MT-32-to-General MIDI mapping tables with a common one in MidiDriver.
- Changes Sky's GmChannel class to allow NULL instrument and velocity mapping tables, giving a 1-to-1 mapping without creating a dummy table.
svn-id: r17361
o added configure option
o mi2 intro doesn't freeze anymore and has no sound glitches
o missing instruments in many titles are fixed
o numerous memory overwrite bugs are fixed
o code is cleaned a lot and splitted into many smaller files
o mt32.cpp went to backends/midi
o synced with upstream code
o reverberation fixed
* don't complain about File class wrapper :)
* all custom types are back
* #pragmas are to do
* maybe some indentation is wrong too
I prefer smaller commits, but this thing came in one piece.
svn-id: r15715
It is not a valid MIDI control change message.
It is used internally in the SCUMM MIDI tracks
to indicate part priority.
The only MidiDriver that needs it is Adlib.
It should be noted that the priority values are
out of bounds for a regular MIDI message anyway.
Valid control change values are 0-127, but the
priority value ranges from 0-255.
This fixes a crash with a certain software MT-32
emulator running under Linux. Thanks to olki for
bringing up this issue.
svn-id: r13645
As defined in Patch [811623] MT-32 patch for Bug 810564
Added a channel mask to MPU-401 devices so that --native-mt32
may force the device to use only the subset of MIDI channels
actually supported by the MT-32. Also added a best-guess
interpretation of iMuse Part priority in the SysEx 0x00 msg,
since part priorities become more of an issue when the
channel count is cramped.
svn-id: r10409
This circumvents problems doing generic send()
calls to MidiDrivers that support more than
16 MIDI channels (i.e. Adlib). Because of the
way it interacts with MidiDriver, Simon could
have run into a problem if it tried to
allocate more than 15 Adlib music channels
(though this would only happen in very, VERY
rare circumstances).
Also fixed a problem with the channel
numbering scheme used by MidiDriver_Adlib,
in particular the percussion channel number.
svn-id: r9604
Miscellaneous upgrades and fixes to MidiParser, including the Smart Jump (which could not be tested before iMuse started making use of the MidiParser).
*** THIS IS A BIG UPGRADE! EXTENSIVE REGRESSION TESTING IS NEEDED! ***
This has been tested through the intros and a number of other scenes from MI2, FOA and S&M.
NOTE! This upgrade introduces savegame format version V19. Earlier version savegames will load, but the music will simply start over from the beginning. Only V19 and later games will properly restore the position of the music! Don't say you weren't warned....
svn-id: r7849
