Age | Commit message (Collapse) | Author | |
---|---|---|---|
2010-11-24 | SCI: Moved handling of special reverb value 127 outside of the MIDI driver | Filippos Karapetis | |
svn-id: r54459 | |||
2010-11-23 | SCI: Also send a sustain off on a notes off event (patch from bug #3116608) | Filippos Karapetis | |
svn-id: r54429 | |||
2010-09-01 | SCI: set master volume correctly | Martin Kiewitz | |
and merge it together with global volume, fixes bug #3053104) svn-id: r52484 | |||
2010-08-18 | SCI: Only perform queuing of song signals for SCI0, where we handle the ↵ | Filippos Karapetis | |
updating of song queues ourselves. Newer games handle signaling on their own, thus we shouldn't interfere with this. Fixes bug #3045913 - "PHARKAS CD: Crash at the start menu" (a regression from rev #52043) svn-id: r52195 | |||
2010-08-12 | SCI: queuing signal, if signal not passed to scripts | Martin Kiewitz | |
fixes laura bow 1 when knocking at the door in the attic, scripts wait for 0xb in that case, sound resource sets 0xb and then immediately ends. This resulted in the scripts only getting the termination signal, so they waited endlessly. (bug #3042981) svn-id: r52043 | |||
2010-08-03 | SCI: not error()ing out on no free channels | Martin Kiewitz | |
instead we just ignore such channels. I'm not sure how sierra sci behaved in that case, they ignored channels as well, but maybe they removed them from earlier music svn-id: r51715 | |||
2010-07-20 | SCI: fix regression of r50463 | Martin Kiewitz | |
also check delta, because _play_tick does not include delta at that point - fixing sq1 ship takeoff svn-id: r51067 | |||
2010-07-14 | SCI: adding sound version 2.1, to properly add sci2.1 signature for ↵ | Martin Kiewitz | |
kDoSound(play) svn-id: r50878 | |||
2010-07-13 | SCI: erroring out on unsupported soundversion in midiparser | Martin Kiewitz | |
svn-id: r50849 | |||
2010-06-28 | SCI: removing ignoring of channel volume change code from r50463, the lsl5 ↵ | Martin Kiewitz | |
issue was actually caused by the old volume adjust code that was also causing the issues in gk1 svn-id: r50466 | |||
2010-06-28 | SCI: its lsl5 sound resource 294, not 274 (changed comment only) | Martin Kiewitz | |
svn-id: r50465 | |||
2010-06-28 | SCI: ignoring set signal on tick 0 directly in parseNextEvent instead of ↵ | Martin Kiewitz | |
filtering, also now ignoring channel volume changes on tick 0 (fixes lsl5 fading of sound 274, almost at the end), fixing also fading in gk1 and enabling fading in sci32 again svn-id: r50463 | |||
2010-06-28 | SCI: remove clipping of volume inside SciMidiParser::setvolume | Martin Kiewitz | |
svn-id: r50456 | |||
2010-06-28 | SCI: set volume all the time, fixes volume being 0 after fading in sci0 | Martin Kiewitz | |
svn-id: r50418 | |||
2010-06-28 | SCI: changed the way setVolume works for sci0 back to the way it was from ↵ | Martin Kiewitz | |
r50414, because adlib driver doesn't support channel volume changes for sci0 games, fixes fading in those games svn-id: r50417 | |||
2010-06-28 | SCI: reordering and cleanup, removing debug code from sci0 mix-code | Martin Kiewitz | |
svn-id: r50415 | |||
2010-06-28 | SCI: implementing real setVolume() support, fixing some fading in sci1 games ↵ | Martin Kiewitz | |
(like pq3 intro) svn-id: r50414 | |||
2010-06-28 | SCI: regression from implementing channel remapping and r50405 - center ↵ | Martin Kiewitz | |
pitch wheels and hold pedal on init instead of unloadMusic(), fixes lsl5 piano scene svn-id: r50413 | |||
2010-06-28 | SCI: fix regression of r50405 - sq1vga right at the beginning - scripts are ↵ | Martin Kiewitz | |
pausing sound and then sending manually commands afterwards svn-id: r50411 | |||
2010-06-27 | SCI: tell midiparser that he lost ownership of channels in case we take them ↵ | Martin Kiewitz | |
away svn-id: r50405 | |||
2010-06-27 | SCI: resetting curPos as well when mixing sound resources | Martin Kiewitz | |
svn-id: r50401 | |||
2010-06-27 | SCI: filter out set signal commands from midi streams, if they happen at ↵ | Martin Kiewitz | |
ticker 0. fixes castle daventry getting removed immediately in kq5 french/mac svn-id: r50385 | |||
2010-06-26 | SCI: fix warning | Martin Kiewitz | |
svn-id: r50341 | |||
2010-06-26 | SCI: do boundary checking when loading sci1 sounds and mixing them together. ↵ | Martin Kiewitz | |
The final end credits song (resource 699) in kq5 has some channels with invalid offsets, crashing ScummVM before svn-id: r50339 | |||
2010-06-26 | SCI: set _driver to 0 in MidiParser_SCI destructor to stop the MidiParser ↵ | Martin Kiewitz | |
destructor call allNotesOff() which would affect all channels all the time - fixes sq1 ship flying away sound getting stopped when ladder falls down svn-id: r50313 | |||
2010-06-26 | SCI: removing fixme (issue got already solved) | Martin Kiewitz | |
svn-id: r50311 | |||
2010-06-25 | SCI: only send new volume to currently mapped channels, fixes assert in sq1 ↵ | Martin Kiewitz | |
because setVolume is called on an object that's currently not playing. added fixme svn-id: r50274 | |||
2010-06-21 | SCI: instead of queueing inside midiparser, we are queueing now globally in ↵ | Martin Kiewitz | |
SciMusic, also some little cleanup svn-id: r50130 | |||
2010-06-21 | SCI: added fixme about r50110/r50073 | Martin Kiewitz | |
svn-id: r50115 | |||
2010-06-21 | SCI: fix regression of r50073, allNotesOff() now directly sends to driver ↵ | Martin Kiewitz | |
again. If we send to queue, queue will never actually get processed and even if it was, the channels wouldnt be mapped anymore anyway svn-id: r50110 | |||
2010-06-20 | SCI: storing all manual midi commands now and actually sending them to ↵ | Martin Kiewitz | |
driver during onTimer() - fixes mt32 emulation crashing during lsl5 piano scene svn-id: r50073 | |||
2010-06-19 | SCI: fixing last uninitialized variable issue in MidiParser_SCI (hopefully) | Martin Kiewitz | |
svn-id: r50063 | |||
2010-06-19 | SCI: fixing another uninitialized variable usage issue - also limiting reset ↵ | Martin Kiewitz | |
velocity to used channels only, same is true for setting voice count svn-id: r50062 | |||
2010-06-19 | SCI: stopping scripts from sending to unused channels manually (fixes sq1vga) | Martin Kiewitz | |
svn-id: r50060 | |||
2010-06-19 | SCI: reverting r50058 - we can't actually send to the channels at that time, ↵ | Martin Kiewitz | |
because we do not own the channels at that point svn-id: r50059 | |||
2010-06-19 | SCI: move setVolume down inside loadMusic so that it actually affects the ↵ | Martin Kiewitz | |
used channels svn-id: r50058 | |||
2010-06-19 | SCI: fixing uninitialized usage in MidiParser_SCI | Martin Kiewitz | |
svn-id: r50057 | |||
2010-06-18 | SCI: implementing fading for sci1 only for used channels of the object, ↵ | Martin Kiewitz | |
fixes lsl1 music issues when going right at the start svn-id: r50018 | |||
2010-06-18 | Added information regarding a FIXME | Filippos Karapetis | |
svn-id: r50002 | |||
2010-06-18 | Changed an assert into a warning with a FIXME (happens in SQ1) | Filippos Karapetis | |
svn-id: r49994 | |||
2010-06-17 | SCI: implemented channel muting for sci1, finally fixes lsl5 paino scene ↵ | Martin Kiewitz | |
with patti svn-id: r49926 | |||
2010-06-17 | SCI: fixes regressions of r49905 (thx to digitall) - first we now find out ↵ | Martin Kiewitz | |
what channels are used in sci0 as well, we are doing straight mapping for sci0 though - this fixes music in sci0 games, also fixing crash in sq1vga because of using null pMidiParser svn-id: r49922 | |||
2010-06-17 | Fix another valgrind warning, thanks to digitall for spotting. | Matthew Hoops | |
svn-id: r49918 | |||
2010-06-16 | SCI: fix debug command "show_instruments", got broken by r49905 on purpose | Martin Kiewitz | |
svn-id: r49906 | |||
2010-06-16 | SCI: implement channel remapping for SCI1 | Martin Kiewitz | |
svn-id: r49905 | |||
2010-06-14 | SCI: Added a version of allNotesOff() which sends messages only to the ↵ | Filippos Karapetis | |
channels used by the associated song instead of all channels svn-id: r49673 | |||
2010-06-12 | Added debug output to kDoAudio and some music commands. Also added a warning ↵ | Filippos Karapetis | |
when an audio stream can't be created. Finally, the debug level of the MIDI parser debug output has been raised to 4, as it's too verbose svn-id: r49605 | |||
2010-06-11 | - Added a new debug command, verify_midi, which can be used to check all the ↵ | Filippos Karapetis | |
songs of a game for unmapped instruments (still WIP and disabled) - Fixed a bug in the verify_scripts command (it was loading the script resource twice) svn-id: r49597 | |||
2010-06-10 | SCI: adding current delta to tick position when getting the setsignalloop ↵ | Martin Kiewitz | |
command - fixes sq3 music slight delay before actual looping svn-id: r49585 | |||
2010-06-04 | Some more work on channel remapping: Moved the remapping code to the music ↵ | Filippos Karapetis | |
loading code (still disabled) svn-id: r49430 |