| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
Currently unused, but ready to be hooked up to various classes using it.
|
|
|
|
|
|
If the driver id did not match the array index, the wrong driver entry would
be accessed causing a crash in the worst case.
|
|
set default volume of every midi channel to 0x7F right on startup
fixes Amazon Guardians of Eden logo jingle music
those values were actually sent to the driver right on startup
default value for panning was 0x40, this was adjusted as well
|
|
tries to detect OPL3 availability and prefers OPL3 timbre data
in that case.
|
|
- support AdLib + OPL3 streams
- also support stream(s) and filenames getting passed at the same time
in that case filenames are checked first, streams are used
as fallback
|
|
This protects it from any calls to stopAll() in the mixer, which
is needed for Return to Zork, at the very least.
|
|
|
|
Checks if at least one of those exists and reads the one that does
Needed for another version of Simon The Sorcerer 2, that only
features FAT.OPL
|
|
- not using byte pointer + size and instead now a
SeekableReadStream for factory
- adjusted ACCESS + AGOS engines accordingly
|
|
|
|
was added for testing purposes and was commited by accident
|
|
clean up, uses separate non-const pointer for reading file
and a const pointer for actually parsing the data
|
|
Sherlock Holmes was mentioned instead of Miles Audio
|
|
fixes broken music in Return To Zork (demo)
also added comments about those special cases (useful for testing)
|
|
simon 2 + return to zork are getting output on one speaker only atm.
both games don't get panning controller at all and the original
driver itself definitely did initialize it to 0.
We initialize panning to 63, so that this gets fixed for the moment.
needs to get investigated further.
|
|
add ability to pass timbre data directly to the AdLib Miles
Audio driver as well
|
|
calculate sysex address properly
|
|
sysex may not contain bytes above or equal 0x80, so we adjust
the address accordingly. fixes weird sound getting played
during 7th guest MT32 setup screen
|
|
implement support for TIMB chunk inside XMIDI-parser (forwarding of data to driver)
implement actual support for TIMB chunk inside Miles Audio MT32 driver
|
|
now an overflow of the embedded sysEx buffer will properly
execute a send.
|
|
initialize currentPitchRange (although it was properly initialized
later anyway)
|
|
also add warnings in case there are non-existant instruments
or no-non-protected timbre slots are available
|
|
|
|
use MILES_MT32_SYSEX_TERMINATOR instead of hardcoded 0xFF
missed one
|
|
use MILES_MT32_SYSEX_TERMINATOR instead of hardcoded 0xFF
|
|
now supports embedded SysEx as well
used by 7th guest
|
|
for games, that do not have a MT32 timbre file, simply pass an empty
filename to the Miles-MT32-factory.
|
|
channels were (wrongly) reversed
found and tested in discworld 1
|
|
- remove Miles Audio from Sherlock engine
- put Miles Audio into common audio (namespace Audio)
- Miles Audio is used at least by the engines
TINSEL, GROOVIE, TOLTECS, SAGA and KYRA
This way it can be used by the other engines
|
|
fixes non working audio when playing a File(Stream)
|
|
All callers requiring SeekableAudioStream have been adapted by using dynamic_cast
|
|
|
|
|
|
- rework 3DO audio decoders to decode into buffer only
- 3DO audio decoders also use streams without separate size arg now
- add support for ADP4 + SDX2 inside AIFC files
- add debug command "3do_playaudio" to play AIFC files
- remove audio flags and replace with stereo bool
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- queue up to 0.5 seconds of audio to avoid buffer underruns
- support for SDX2 codec
- put both audio codecs into audio/decoders/3do.cpp
- made movie player capable of playing EA logo movie
|
|
The way I've configured ScummVM, the default is to use FluidSynth
for MIDI, and any game where I want AdLib overrides that setting.
I have set the DOSBox OPL emulator as the default.
However, this causes the game's "opl_driver" setting to be "auto",
which overrides the default setting, and usually picks the MAME
OPL emulator. To counter this, if it sees that "opl_driver" is
"auto", it will check the default setting.
In other words, only use "auto" if "opl_driver" isn't set on any
level. I think this is the behaviour that should cause the least
amount of astonishment.
|
