LimitFPS now can be enabled, disabled or automatic. The latter option
doesn't wait for audio if the emulator is running slow (over the last 2
secs)
Renamed `Audio Backend` to `Audio Driver` in the UI.
The backend supports 44.1KHz and 48KHz (with resamping). The resampler
is not great, has some noise but no idea where it comes from.
This enables the switch port, since using SDL2 is the quickest way
to get audio working.
TODO: Add support in the cmake, once cmake is fixed at master/HEAD.
As we can now configure the alsa device in the GUI, it's no longer necessary to write the first working device back to the configuration.
Also there is now the "auto" device to automatically try to initialize the alsa device.
Trying to load the device from the emu.cfg and if not set (should be the default for everyone currently) use the existing procedure to try to determine the device (default > plughw:0,0,0 > plughw:0,0).
A new function was introduced to abstract the setup of input devices for each os: os_SetupInput().
I hope I implemented this everywhere correctly and the behaviour is identical to before.
The new function mcfg_CreateNAOMIJamma() creates the NAOMI Jamma interface and is hidden behind a compile flag (same as before).
The previous function mcfg_CreateDevices() was renamed to mcfg_CreateDevicesFromConfig() because it creates the devices based on the configuration (``nb`` setting).
This introduces the USE_ALSA flag (replaces the somewhat hacky
"HOST_OS==OS_LINUX && !defined(TARGET_NACL32) && !defined(ANDROID)"
if case).
The Makefiles for beagle, gcwz, lin86 and pandora have been adapted to
reflect this change.
This commit introduces PulseAudio support using PulseAudio's "Simple
API".
The PulseAudio backend will be included during compilation if
USE_PULSEAUDIO is defined.
This *should* add the Android audio backend. I do not have any idea how
to do this properly, but I figured that this might work for the moment.
There's probably a lot of testing required for this to work.
This commit adds the DirectSound audio backend, which is a modified
version of core/oslib/ds_audiostream.cpp.
Porting this was a bit tricky, since ds_audiostream.cpp defined 3
extern functions that the other backends didn't have:
* bool os_IsAudioBuffered();
* bool os_IsAudioBufferedLots();
* int os_getusedSamples();
I came to the conclusion that these methods are obsolete since they are
only used core/windows/winmain.cpp in the function void os_wait_cycl(u32
cycl) - which is used nowhere. Thus, I removed os_wait_cycl and the
headers of the three functions in core/oslib/oslib.h. I also removed the
functions themselves (except for int os_getusedSamples(), which is still
used inside the directsound backend, but can safely declared static for
that purpose).
The DirectSound backend will be included during compilation if HOST_OS
is OS_WINDOWS.
This adds the OSS audio backend, which is basically a slightly modified
version of core/oslib/alsa_audiostream.cpp from the skmp/linux-x64
branch (as of commit cdf9f3dc971506c2efd979275c2869f064e1027c).
The ALSA backend will be included during compilation if HOST_OS is OS_LINUX
and neither TARGET_NACL32 nor ANDROID are defined. This should be
changed to a USE_ALSA flag in the future (for constistency and and
simplicity reasons), but i didn't want to put too much stuff into this
commit.
Flyinghead