Steve Checkoway
0d10c76e87
Unscatter arch includes.
2006-11-17 18:25:20 +00:00
Steve Checkoway
5f9ee5c5bd
Set sample rate of the device, if possible.
2006-11-17 17:37:51 +00:00
Steve Checkoway
1ea6837c02
Header cleanup.
2006-11-17 17:13:39 +00:00
Steve Checkoway
e988ea9945
Survive device changes, (external) sample rate changes, etc.
2006-11-17 16:59:22 +00:00
Steve Checkoway
1f79f25240
Fix crash by stopping the audio _on_ the real time thread. This ensures that the current buffer gets played but that it never gets called again. I'm not convinced that the property listeners won't get called but I'm not sure how to handle those sanely just yet, at least in a lockless manner.
...
This code doesn't handle external format changes very well yet--but then, neither does the other sound driver.
2006-11-12 09:04:14 +00:00
Glenn Maynard
e130d73f88
DSERR_BUFFERLOST handling in DSoundBuf::GetPosition
2006-10-04 02:33:20 +00:00
Glenn Maynard
03ce23e9ea
ARRAYLEN
2006-09-18 23:27:09 +00:00
Glenn Maynard
b77024f3bc
include
2006-09-18 22:45:42 +00:00
Glenn Maynard
1f7fe1afd3
include
2006-09-18 22:24:18 +00:00
Glenn Maynard
62c114b49e
remove param
2006-08-21 00:20:13 +00:00
Steve Checkoway
a07e194cf1
Revert. Fixes Linux build.
2006-07-18 05:51:37 +00:00
Jason Felds
366c3440d4
Fix VC8 macro redef warning. (more to come)
2006-07-18 03:56:37 +00:00
Steve Checkoway
b1d08d1f8b
This is still a work in progress (although I haven't worked on it in over a month).
2006-07-17 00:44:38 +00:00
Steve Checkoway
a2dca95d01
Use native endian. Now Alsa works for powerpc linux.
2006-06-29 03:22:29 +00:00
Steve Checkoway
8d82bdfb82
Do not allocate or deallocate memory in the mixing thread. Doing so can block that thread which is possibly a realtime thread. Distinguish between STOPPED and AVAILABLE. In the STOPPED state, the memory for the sound is still around. It will be deallocated in Update() at which time the state will change to AVAILABLE.
...
The BUFFERING state is to keep all other threads from accessing the sound while StartMixing() is prebuffering. It corresponds to the old state STOPPED with available set to false.
2006-06-19 06:43:43 +00:00
Steve Checkoway
78d055dfad
Beginning of a sound driver that uses AUHAL instead of dealing with the HAL itself. Apple engineers claim it's easier to use (although, I note there's almost zero documentation) and it correctly handles things like device format changing. It currently works, for the most part (on 10.4 and above).
2006-06-18 19:05:40 +00:00
Steve Checkoway
8d9c59287d
Handle JackChanged getting called when the device is stopped.
2006-06-18 18:55:48 +00:00
Steve Checkoway
2572b8c5ea
That last commit was a bit more than I intended. Oh well. It (and this one) simplify the property listener code and add more listeners.
2006-06-18 09:40:56 +00:00
Steve Checkoway
1f72db5efc
Fix parameter order.
2006-06-18 08:28:18 +00:00
Glenn Maynard
3a70b367c0
temporary fix for crash (proper fix will take more time)
2006-06-11 19:07:54 +00:00
Steve Checkoway
bc6cf3b9a3
missed commit
2006-05-20 05:37:03 +00:00
Steve Checkoway
76e98bdaf5
More logging, but less noisy logging (log instead of info).
2006-05-19 20:26:25 +00:00
Steve Checkoway
59c663b238
Attempt to handle devices other than the built-in audio correctly. Not tested on anything but the built-in audio.
2006-05-19 08:09:38 +00:00
Steve Checkoway
060b68ff00
Add more logging: Check that the sample rate we just set matches what the device declares as it's sample rate; log if the device's sample rate changes while it's running.
2006-05-17 07:40:54 +00:00
Steve Checkoway
c2b857d01c
Remove bloated c++ wrapper.
2006-05-06 11:19:25 +00:00
Steve Checkoway
75d67fbe85
Spacing.
2006-01-22 07:24:18 +00:00
Steve Checkoway
3a4f3a8efd
Less noisy.
2006-01-22 07:23:05 +00:00
Steve Checkoway
4a8daa3671
Fix: plugging in head phones causes arrows to stop moving.
2006-01-22 07:21:31 +00:00
Steve Checkoway
f909c72376
Cleanup spacing.
2006-01-22 05:33:37 +00:00
Chris Danford
7f821e8cfc
CString -> RString
2006-01-22 01:00:06 +00:00
Steve Checkoway
9ed7548de1
Use new api.
2006-01-21 11:14:13 +00:00
Glenn Maynard
2084896f7d
RageSoundUtil::Attenuate
2005-12-30 19:04:18 +00:00
Steve Checkoway
99c41cde08
Cleanup registered threads.
2005-12-20 03:20:31 +00:00
Glenn Maynard
dcfff261ed
use RageThreadRegister. These should be freed properly, so things still work if the driver is recreated for some reason. There may be no guarantee that the same thread will always call the I/O callback--it may recreate the thread on overload, for example ...
2005-12-19 02:06:56 +00:00
Steve Checkoway
510254abb4
Add more property listeners, name the two threads created by the audio HAL.
2005-12-17 13:10:23 +00:00
Steve Checkoway
223c541ecc
This stops crashes when headphones are removed. If they're removed during gameplay, the arrows stop scrolling and the receptors stop flashing to the beat. After the hardware reconfigures itself (~.5s), the music resumes.
...
My guess (which I don't have time to test right now) is that the returned position is reset when the hardware starts back up.
2005-12-15 11:44:46 +00:00
Glenn Maynard
6497ac8f94
style
2005-12-07 03:03:21 +00:00
Glenn Maynard
321931c19a
not used
2005-12-07 03:01:23 +00:00
Glenn Maynard
b617542b62
cleanup
2005-12-07 03:00:27 +00:00
Glenn Maynard
b6ba17a7bb
fix busy looping on slower systems, inefficient buffer filling
2005-11-28 17:17:35 +00:00
Glenn Maynard
b948f686cc
cleanup
2005-11-25 20:38:03 +00:00
Glenn Maynard
9ed3cde95a
cleanup
2005-11-25 18:36:44 +00:00
Renaud Lepage
6e6e89276d
Fix for Xbox compilation. It was easier with HAVE_FOO, but oh well.
2005-11-16 04:38:28 +00:00
Glenn Maynard
3bbfe9ebc0
Two tests for not filling a buffer:
...
if( m_iBufferBytesFilled > m_iWriteAhead )
int iNumBytesEmpty = m_iWriteAhead - m_iBufferBytesFilled;
if( iNumBytesEmpty < iChunksize )
The latter, manipulated a bit, is actually:
if( m_iBufferBytesFilled+iChunksize >= m_iWriteAhead )
These do almost the same thing. This isn't what was intended. The
former checks whether we *want* to write another buffer (based on our
preferred writeahead). The latter was meant to check whether it's *possible*
to fit another chunk into the buffer. It should check against m_iBufferSize,
not m_iWriteAhead.
This fixes writeahead: 4096 means "write if we have less than 4096 frames
buffered", not "write if writing the new chunk would result in still having
less than 4096 frames buffered". This is important on slower devices, like
USB headsets.
2005-11-10 22:10:05 +00:00
Steve Checkoway
8310140755
Set new arbitrary values with the input handler above the sound decoder.
2005-10-24 11:40:43 +00:00
Steve Checkoway
05201a047c
Header cleanup. Use SetThreadPrecedence() and add potential replacement code for GetCurrentTime().
2005-10-24 11:21:17 +00:00
Steve Checkoway
aa0f90feae
__MACOSX__ --> MACOSX, keep __MACOSX__ for ogg.
2005-10-24 10:37:56 +00:00
Steve Checkoway
35ad7eb289
Support preferred sample rates and report the latency more correctly.
2005-10-24 08:55:25 +00:00
Steve Checkoway
cb229a8774
Remove redundant defines.
2005-10-24 08:17:12 +00:00
Glenn Maynard
539ec6aa5c
HAVE_WIN32 -> WINDOWS. HAVE_LINUXKERNEL -> LINUX. If some features originating in Linux work on other systems, then define them specifically (eg. HAVE_LINUX_JOYSTICKS).
2005-10-24 07:22:28 +00:00
Steve Checkoway