Age | Commit message (Collapse) | Author |
|
external directory (and as such, allow you to have a debug, a profiling, and a release build from the same set of source files in parallel). Work in progress
svn-id: r12680
|
|
actually do; applied patch #886786 which corrects a bug in the balance code
svn-id: r12665
|
|
svn-id: r12559
|
|
svn-id: r12558
|
|
svn-id: r12503
|
|
(default, 8bit backends should define BACKEND_8BIT for fast colour remapping) and sound syncronisation.
svn-id: r12456
|
|
svn-id: r12176
|
|
care of that
svn-id: r12135
|
|
svn-id: r12110
|
|
svn-id: r12103
|
|
svn-id: r12102
|
|
used for MP3/Vorbis support -> this eases changing the implementations, and reduces header dependencies (and thus compile time) :-)
svn-id: r12097
|
|
variant with mad_timer_t was really only needed for MP3TrackInfo class)
svn-id: r12095
|
|
svn-id: r12094
|
|
loops
svn-id: r11992
|
|
svn-id: r11981
|
|
svn-id: r11980
|
|
svn-id: r11976
|
|
svn-id: r11962
|
|
svn-id: r11951
|
|
svn-id: r11942
|
|
svn-id: r11939
|
|
stream even after it has been given to the mixer for playback
svn-id: r11933
|
|
eosIntern will in some border cases return wrong results; some cleanup
svn-id: r11932
|
|
o Removed SoundMixer::bindToSystem()
o In scumm, replaced _silentMixer, _silentDigitalImuse and _noDigitalSamples by SoundMixer::isReady()
svn-id: r11893
|
|
'All note off' event, at least according to http://crystal.apana.org.au/ghansper/midi_introduction/midi_channel_mode.html ; and now, the quicktime MIDI code doesn't support it anymore, too ;-)
svn-id: r11885
|
|
mainly because people kept (accidentally and sometimes on purpose :-) misusing them
svn-id: r11881
|
|
svn-id: r11879
|
|
svn-id: r11858
|
|
slightly more efficient)
o Fixed potential memory leak in SoundMixer::playRaw
o Channel not anymore friend class of SoundMixer
o misc cleanup
svn-id: r11857
|
|
svn-id: r11829
|
|
playSfxSound_Vorbis; made scumm/queen engines use it
svn-id: r11812
|
|
this by playInputStream(), this cleanup eases this a bit)
svn-id: r11803
|
|
the sound data
o This allows us to get rid of the ChannelRaw class
o Removed the sound index return value from several methods
o Removed all methods dealing with sound indices (i.e. stopChannel and pauseChannel)
svn-id: r11801
|
|
report #860067); added some comments; LinearMemoryStream now can auto-dispose the data passed to it
svn-id: r11798
|
|
svn-id: r11756
|
|
o Added SoundMixer::playInputStream and made some of the other play* methods use it
o Added ProcInputStream stub (not working yet) which one day may allow us to replace the premix code, and allow other fancy stuff
o Remove AudioInputStream::readBuffer default implementation (subclasses should always provide it for max. performance)
o Some minor cleanup
svn-id: r11754
|
|
svn-id: r11752
|
|
svn-id: r11742
|
|
svn-id: r11741
|
|
pointless MusicStream class; removed various specific Channel subclasses and instead generalized the base class some more
svn-id: r11699
|
|
svn-id: r11697
|
|
WrappedAudioInputStream; this allows further streamlining of the channel/mixer code (can you already guess what I am working towards? :-)
svn-id: r11696
|
|
svn-id: r11694
|
|
svn-id: r11686
|
|
svn-id: r11665
|
|
svn-id: r11597
|
|
Initialise _cd.playing to false. Otherwise ScummVM may create a savegame
where a CD track appears to be playing, but everything about it is
undefined, causing ScummVM to crash when loading it.
Initialise _track_info[] with NULLs, otherwise ScummVM crashes for me when
I start the CD version of MI1.
There's probably a lot more that *should* be properly initialised, but this
seems to take care of the most serious issues, and is all I have the time
to do now anyway.
svn-id: r11573
|
|
Painfully detailed in its description of how to
get a MidiParser object hooked up and running.
Hope this helps ya, Joost! ;)
svn-id: r11559
|
|
now based on the READ_LE_* and READ_BE_* macros.
Very slight performance gain, woo hoo.
svn-id: r11558
|