a337a59fb8
Use alext.h for extension definitions
2010-03-16 16:25:30 -07:00
ec917e8e2f
Rename some struct members for consistency
2010-03-16 15:37:41 -07:00
89d84131a4
Avoid dereferencing a NULL buffer when unqueueing buffer 0
2010-03-15 16:10:21 -07:00
2033f60d24
Print an error when realloc fails
2010-03-15 14:28:38 -07:00
d4159430f0
Set the buffer length for Pulse when opening the device
...
So it will properly scale with context frequency requests
2010-03-14 22:20:55 -07:00
14255a38fc
Pulse's buffer size is not a good measurement for the update size
...
Lie and say 20ms instead, since there is no stable method to get such info
2010-03-14 21:45:23 -07:00
6c5f1b78d7
Reduce the mixing buffers to the next power of two
2010-03-14 03:30:42 -07:00
1f2d3a0132
Remove duplication of setting source angles
2010-03-14 03:23:18 -07:00
c23e4ed412
Add support for AL_EXT_DOUBLE
2010-03-13 21:32:45 -08:00
8c0118e7ae
Add support for AL_EXT_MULAW_MCFORMATS
2010-03-13 21:03:34 -08:00
9d4107a515
Add AL_EXT_MULAW_MCFORMATS tokens to alext.h
2010-03-13 18:39:12 -08:00
d07b9c2c11
Rewrite openal-info
...
Output is mostly unchanged, however there are differences. Device-specific info
(including context-specific info) is separated from NULL-device queries, and
there is now a distinction between ALC versions of the NULL device and the
opened device. Additionally, an alternate playback device can be queried by
specifying its name as the first parameter.
2010-03-13 18:03:01 -08:00
3108a6a997
Limit PulseAudio's buffer size to the requested length
2010-03-11 09:47:26 -08:00
b734cab6fc
Fix the retrieval of the RW offsets using the correct functions
2010-03-10 07:33:30 -08:00
08b1bed38b
Remove a couple incorrect comments
2010-03-10 07:16:53 -08:00
df6320eb0a
Remove duplicate copyright notice
2010-03-09 11:25:34 -08:00
80f1ec0d7e
Protect some capture calls with a mutex
2010-03-09 10:14:52 -08:00
98f4f5a4f0
Add EFX types and enums to alext.h
...
The types and enums are in efx.h, which is included by alext.h. It's done this
way because EFX has a lot if definitions which would polute alext.h
2010-03-09 09:28:22 -08:00
8feb089f5c
Build device lists only when needed
2010-03-09 06:09:29 -08:00
2ba3a88ace
Probe physical devices separately from appending them to the device list
2010-03-09 05:44:18 -08:00
1f10195c47
Use powf when available
2010-03-07 22:12:33 -08:00
dc40702b53
Remove unused source param count macro
2010-03-07 21:48:25 -08:00
a9e0e57797
Keep UpdateSize updated when apps request a frequency
2010-02-24 18:51:57 -08:00
9b47048231
Silence some warnings
2010-02-16 05:44:39 -08:00
c05a3d3549
Better handle run-time linking
2010-02-16 05:43:36 -08:00
2083884ae3
Properly check for functions in Win32 shared libs
2010-02-16 05:36:49 -08:00
2f644c7ed5
Constify some globals
2010-02-15 22:01:23 -08:00
d17c4a098c
Avoid probing for devices on lib initialization
2010-02-10 16:34:43 -08:00
96303893bc
Set ALSA playback buffer and period size based on time
...
So rate adjustments won't modify the latency
2010-02-08 17:42:44 -08:00
a329f9c84a
Fix retrieval of the current global context when changing it
2010-02-08 16:58:26 -08:00
fc96e965ae
Always disable ALSA's resampler for playback
...
This means the frequency config option won't have an effect if the ALSA device
can't set an alternate sampling rate
2010-02-08 16:38:53 -08:00
365a62d63a
Improve source write offset calculations
2010-02-07 20:11:24 -08:00
3a87c3c3ba
Avoid using periods with PulseAudio, and add an option to specify its buffer length
2010-02-07 19:29:48 -08:00
54dc26b145
Enable the WinMM backend only if the lib check succeeds
2010-02-05 12:07:51 -08:00
78c1629aeb
Disable ALSA's resampler when no specific output rate is requested
2010-02-05 11:45:17 -08:00
bd518ed264
Make sure all requested data gets written to PulseAudio
2010-01-24 00:03:23 -08:00
56379bffac
Use pa_channel_map_superset only when available with PulseAudio 0.9.15
...
pa_channel_map_equal can otherwise detect some compatible output modes
2010-01-22 21:20:08 -08:00
11d7a0e394
Specify the buffer size to ALSA, so the period count remains flexible
...
Under conditions where the period size is fixed (such as dmix), ALSA will still
attempt to provide the requested number of periods even if the requested period
size needs to be dramatically altered. In these cases, it would be better to
increase or decrease the period count, instead of growing or shrinking the
total buffer size and significantly changing the latency.
2010-01-20 01:19:32 -08:00
3ad86a1c1b
Release 1.11.753
2010-01-17 17:46:27 -08:00
931f5875cd
Fix speaker angle fixup
2010-01-12 09:11:46 -08:00
f22b02a762
al_print doesn't need to be inline
2010-01-12 09:05:57 -08:00
505f61ff43
Properly trace the invalid speaker angle
2010-01-12 08:50:32 -08:00
0c2efa1253
Use GCC's format attribute for al_printf
2010-01-12 08:48:25 -08:00
5921e18be2
Fix some typos with the solaris backend
2010-01-12 08:28:07 -08:00
5b9e6fe440
Only allow 8- and 16-bit, and mono and stereo, wave file output
2010-01-12 08:26:25 -08:00
35ba7d70dc
Consolidate some source offset handling code
2010-01-12 02:43:37 -08:00
026b4b8f4c
Better handle large second offsets
2010-01-12 02:26:38 -08:00
c924a50bed
Don't use a flag to set an error when applying the source offset
2010-01-12 02:22:38 -08:00
574792d2e1
Fix Windows compilation
2010-01-11 23:47:03 -08:00
ab2eb5a603
Check buffer size when copying padding instead of the data pointer
2010-01-11 10:58:33 -08:00
Chris Robinson