This adds a MIDI driver for the Casio MT-540, CT-460 and CSM-1 devices. These
are supported by Elvira 1 and several Sierra SCI0 games as well as Altered
Destiny and Les Manley 1.
This commit adds the following functionality to the MIDI drivers:
- Add checking if a driver is ready to process MIDI events for a specific
source (rather than any source). To facilitate this, SysExes can now be sent
using a source number. This is stored with the SysEx data in the SysEx queue
and can be checked when isReady is called with a source number.
- Allow specifying controller default values per MIDI channel. Currently this
is only implemented for program.
- The OPL dynamic channel allocation algorithm will now respect statically
allocated channels, in case a subclass uses combined static and dynamic channel
allocation.
It also fixes the following bugs:
- Instrument remapping can now be specified using const arrays.
- OPL instrument writing code is refactored to a separate function.
- OPL note on with velocity 0 would be handled as a note off, and then
continued to be processed as a note on.
- OPL writeFrequency would always write key on bit, even if the note is not
active.
- MT-32 default channel volume was incorrect.
The auto loop functionality of the MIDI parser would call jumpToTick to jump
back to the start of the MIDI stream, then call parseNextEvent to read the
first MIDI event. However, jumpToTick already makes sure the first MIDI event
after the jump point is parsed. Calling parseNextEvent again would discard the
first and parse the second MIDI event.
Fixed this by removing the call to parseNextEvent after jumping to the start.
This commit adds a function to the multisource MIDI drivers for specifying an
arbitrary instrument map for remapping the instruments in the MIDI data.
This is used for remapping instruments from two Simon 2 GM tracks to MT-32
because they are missing from the MT-32 MIDI data set.
This commit adds MIDI parser subclasses for the Simon 1 GMF format and for the
SMF variant used by the Windows versions of Simon 1 and 2.
The GMF format was handled by the SMF parser, which can now be removed from
that class. It also fixes the tempos not matching the DOS interpreter.
The Simon Windows SMF variant was handled in the AGOS midi class. Moving it to
a separate parser allows for some cleaner code in that class. It now also
corrects the tempos to match those of the DOS interpreter.
This adds support for the OPL rhythm mode to the AdLib multisource MIDI driver.
Some games (f.e. Simon 1) use this to generate the rhythm sounds of their music.
This was previously disabled implicitly by the missing define
symbols in the SampleRateConvertor module. This has now been enabled
to use the internal rate convertor code.
This was generating undefined macro warnings from GCC when -Wundef
was passed.
This commit adds support for the SudoMaker RetroWave OPL3 sound card. Requires
the retrowave library.
Add the following settings to scummvm.ini:
retrowaveopl3_bus=serial
retrowaveopl3_port=<port> f.e. COM3 or ttyACM0
The output buffer size variable is unsigned so an assertion of
greater than or equal to zero will always be true, which generates
a type limits compiler warning.
This commit adds a feature to the multisource MIDI drivers which allows setting
a default value for most controllers. This default is then applied at the start
of every new track, which ensures controllers are in the correct state.
This can be used for games which use some controller, but do not reset it at
the start of each track, which can lead to incorrect playback.
Mostly done using the following Ruby script:
(Dir.glob('**/*.cpp') + Dir.glob('**/*.h')).each do |file|
s = File.read(file, encoding: 'iso8859-1')
t = s.gsub(/(([\w_.\[\]]+)\s*=\s*new\s+\S+?\[[^\]]+?\](?!\())([^\{\}]*?)\n\s+memset\(\s*\2\s*,\s*0\s*,[^;]*;/m, '\1()\3')
if t != s
File.open(file, 'w') { |io| io.write(t) }
end
end
Setting the Miles driver version property to 3 was meant to set the pitch bend
range to 2. In the AdLib driver, the pitch bend range was set during
construction, before the version property could be set, so this did not work
properly. This is fixed by moving the MIDI data initialization from the
constructor to the open function.
The pitch bend in Sam & Max was exaggerated when using the OPL3 driver. The
OPL3 functionality was apparently implemented by examining just the original
driver code and not the MIDI processing code in the interpreter. The pitch bend
values sent to the driver are not standard MIDI pitch bend values, but have
been processed by the interpreter, so the driver logic cannot be directly
applied to the pitch bend values from the MIDI data.
This commit fixes the pitch bend calculations to produce the same results as
the original interpreter. It also restores the pitch bend range functionality.
This was disabled for OPL3, probably because it is implemented in the
interpreter instead of the driver. It is used in several Sam & Max MIDI tracks.
The Roland GS MT-32 emulation option would not select the MT-32 drumkit on the
GS device, because the drumkit change was filtered out in the MT32GM driver.
This commit fixes this by adding an exception for this drumkit change when GS
MT-32 emulation is active.
This adds a command queue to the PC speaker emulator for PWM-style sounds that
need timing of less than a millisecond. The play method only supports
milliseconds and requires real time control, which is difficult to achieve with
microsecond timings.
The new playQueue method allows you to queue up playback instructions with
microsecond timing, which are executed when audio samples need to be generated.
The empty OPL instrument check would check only the ADSR envelope fields for 0,
assuming this would mean the instrument would produce no sound. However, some
of these fields are flipped, i.e. max value would produce no sound. So this
shortcut was invalid. It is replaced by checking all fields for 0.
This adds a timer proc to the null multisource MIDI driver to drive the fading
process. The null driver does not have any audio to fade, but without this the
fades will just hang and isFading will remain true. The fades are now
effectively timers so game behavior will remain consistent whether MIDI is
turned on or off.
Miles version 3 introduced support for the MIDI pitch bend range RPN using the
data entry MSB controller. The ScummVM implementation would also process this
controller for Miles version 2 games. This caused a problem in Ultima 8, which
used Miles 2 and which had MIDI data that contained some (probably leftover)
data entry MSB events. These had no effect in the original interpreter, but in
ScummVM they caused the pitch bend range to change, resulting in incorrect
playback.
This is fixed by only processing the data entry MSB controller events when
Miles version is set to 3.
The default source neutral volume for the Miles MIDI driver is set at 256.
However, the default source volume was the same as that of the regular MT-32/GM
driver: 255. This would cause a small volume decrease if a game did not set
source volume.
This commit fixes this issue by setting the default source volume for the Miles
MIDI driver at 256.
The ScummVM Miles AdLib driver would write some different initialization values
to the OPL registers than the original Miles drivers did. The most significant
difference was the amplitude and vibrato depth.
This commit changes the Miles AdLib driver to set the same values as the
original drivers. It also adds support for initializing a dual OPL2
configuration.
Miles 2 uses the default MT-32 pitch bend range of 12 semitones for its AdLib
driver. Miles 3 switched to the default GM pitch bend range of 2 semitones for
AdLib. ScummVM would always use the Miles 2 pitch bend range, which caused
pitch bend for Miles 3 games to sound wrong on AdLib.
To fix this, a property was added to the Miles AdLib driver to specify which
Miles version to emulate. Depending on the value of this option, the correct
default pitch bend range is set.
This fixes some issues which caused the fades in the Miles AdLib driver to not
work properly:
- _timerRate was not set
- end of track meta event was not handled, so active fades were not aborted by
deinitializing the source on end of track
The timerProc fields and onTimer function were removed because they duplicate
functionality from the MidiDriver_Multisource base class.
