| Age | Commit message (Collapse) | Author |
|---|
|
sound command (including cmdUpdateCues, which is executed very often)
svn-id: r46983
|
|
currently not correctly filled out inside SoundResource class - because it wasnt previously needed
svn-id: r46963
|
|
objects, not handles. Also, made the warnings thrown when a slot couldn't be found more verbose. Finally, changed the logic of when warnings should be thrown in cmdSetSoundLoop
svn-id: r46871
|
|
respawns, like what the old music code does. Fixes the MIDI music in KQ5CD, Lonbgow and perhaps others
svn-id: r46843
|
|
MIDI worker thread. Some cleanup
svn-id: r46834
|
|
- Resolved a deadlock with the mixer, and added appropriate mutexes (a result of the fact that SCI mixes MIDI and digital audio in the same list)
- Fixed sound playing when loading games, by properly resetting the MIDI driver
- Reverted savegame version to 14 - the changes in versions 15 and 16 don't have any effect on the currently enabled old music code, and the new music code is disabled by default, and is still prone to changes
- Now saving/loading signal, loop and hold for each sound, as well as reverb
- Added stub code for setting reverb and channel hold
- The signal, loop and hold values of each song are cached, like in SSCI and like what happens in Greg's SCI implementation. This allows a clear separation of the engine code from the rest of the engine. Reverted commits 46792 and 46797
- Removed duplicate song list accessing code
- Song cues are now updated in kAnimate for SCI0, like the old music code does, to compensate for the fact that SCI0 didn't poll for music changes via cmdUpdateCues, like what SCI01 and newer do
- Cleanup
svn-id: r46812
|
|
from sound commands
svn-id: r46692
|
|
- Made the SciMusic class private, and added wrapper functions for invoking specific methods of SciMusic from outside the SoundCommandParser class
- Many SCI games keep creating and destroying sound effects constantly (i.e. many times per second). Therefore, another scheme has been devised, which replaces the mutex that was in place. Whenever a sound command is run which operates on a specific object in the play list, we disallow onTimer() from kicking in. This isn't ideal, but it does stop random deadlocks because of locked mutexes without any noticeable side effects
svn-id: r46681
|
|
- Switched to Common::StackLock for mutexes
- Renamed Music::stopAll() -> Music::clearPlayList()
- Implemented a better Music::stopAll method, which stops all sounds with proper signals, instead of destroying them
- Renamed the SCI0 command cmdGetPlayNext -> cmdStopAllSounds and implemented it
svn-id: r46665
|
|
svn-id: r46650
|
|
svn-id: r46549
|
|
music in sq3
svn-id: r46547
|
|
- Reverted the code to use local copies of certain sound object selectors, like what was done before
- Simplified and fixed the SCI1.1 digital sound effect playing code
- Removed some duplicate code
- Renamed some sound commands
svn-id: r46487
|
|
cleanup
svn-id: r46432
|
|
slight hack to the SoundCommandParser constructor
svn-id: r46430
|
|
svn-id: r46300
|
|
svn-id: r46196
|
|
now. Some cleanup
svn-id: r46020
|
|
svn-id: r45999
|
|
svn-id: r45944
|
|
svn-id: r45934
|
|
- Simplified some sound version checks
svn-id: r45923
|
|
svn-id: r45862
|
