1226aec2fc
Don't change the format tag in MakeExtensible
2019-05-23 02:49:04 -07:00
99a55c4452
Handle a missing default WASAPI device ID
...
Backported fix from d1a86075
2019-02-20 00:22:19 -08:00
515bc4ba53
Workaround lack of roundf with early MSVC
2018-11-02 19:16:45 -07:00
25dfe4b57a
Use HUGE_VALF instead of INFINITY
...
Older MSVC lacks INFINITY, and we define a HUGE_VALF fallback when needed.
2018-11-02 19:16:26 -07:00
9a4a3d3977
Specify the correct array size for casting
2018-11-02 19:16:02 -07:00
7a79f09a31
Add a comment about waiting to kill the event thread
2018-10-07 17:18:50 -07:00
db65113c5f
Use a 24-bit dither depth limit
2018-10-03 13:51:21 -07:00
f3f94e478b
Don't limit output for ALC_DONT_CARE_SOFT and float samples
2018-10-03 13:48:04 -07:00
c39eeb9638
Don't try to get the JNIEnv on Android
...
It's currently not used. More stuff is needed anyway which may need a different
approach.
2018-10-02 12:40:26 -07:00
493c8bbc83
Add back an inadvertently removed static
2018-10-02 12:25:26 -07:00
052fdd67cd
Use < instead of != for some loop checks
2018-10-01 15:59:05 -07:00
ed8f44d102
Don't scale the reverb fade counter so much
2018-09-30 16:34:00 -07:00
1d1acc0c34
Ensure BUFFERSIZE is a power of 2
2018-09-30 12:12:27 -07:00
1860b2ec8a
Make the Compressor struct opaque
2018-09-30 12:08:01 -07:00
7ad4f753a4
Fix some length ranges
2018-09-29 19:39:49 -07:00
d5a78f0843
Improve a couple loops
...
Avoid masking the index with each iteration, and instead do up to when the mask
would apply. This allows for better optimizations, in particular fewer
instructions and better chances for vectorization.
2018-09-29 18:48:28 -07:00
32494e72a5
Don't use a ringbuffer design for the limiter's side chain
...
Rather than continuously wrapping when used, each update uses it from the front
and copies the tail to the front at the end. This allows for more effficient
accesses in loops.
2018-09-29 11:54:45 -07:00
0396c0ecd2
Add some assumes for the limiter
2018-09-29 09:47:05 -07:00
51a447852f
Simplify a lower-bound clamp
2018-09-28 18:34:21 -07:00
1684b2cc5f
Constify a couple more variables
2018-09-28 16:42:22 -07:00
79314c4461
Include the limiter's lookAhead delay in the device latency
2018-09-25 23:05:27 -07:00
2d6309d6fc
Don't hardcode the limiter threshold
...
It's now calculated from the output sample type and dither depth.
2018-09-25 10:33:20 -07:00
c69338bc0d
Update the output limiter/compressor
...
This provides better characteristics for an amplitude limiter. In particular,
it utilizes the peak amplitude instead of the RMS, and the used parameters
basically guarantee no output samples exceed the given threshold... almost, due
to floating-point errors as the threshold is converted from dB to log-e for the
envelope, then is negated and converted to linear amplitude to apply to the
signal. It's quite possible for some rounding errors to creep in and not
perfectly saturate the result.
2018-09-25 10:04:14 -07:00
39c3314d00
Only compare the reverb params that induce a need for fading
...
The offsets and coefficients are controlled by a relatively small set of input
parameters, just with different base constants or different calculations. This
lead to numerous redundant checks since if one value didn't change, others that
use the same inputs wouldn't have either.
2018-09-23 22:59:16 -07:00
db6905bf57
Clear reverb gain coefficients when doing a device update
2018-09-23 02:22:23 -07:00
800326d37a
Rename ALreverbState to ReverbState
2018-09-23 01:11:58 -07:00
77a53594ba
Improve the gain stepping difference check
...
Given the more stable stepping now in use, check that the total difference is
enough for perceptible transition, instead of the step itself.
2018-09-22 08:26:52 -07:00
36a6b9d42a
Adjust comment spacing
2018-09-21 06:06:29 -07:00
b6d0ff02c2
Use an internal event to more timely release old effect states
2018-09-21 04:14:15 -07:00
f21e2df4cf
Remove an unnecessary mutex
2018-09-20 22:53:16 -07:00
84a90b7cc3
Stop the event thread when releasing the context
...
To ensure no user callback gets called after alcDestroyContext.
2018-09-20 22:44:58 -07:00
ebbbeb0d66
Put user events in a union structure
2018-09-20 21:59:38 -07:00
cb8545346d
Always start the event thread with the context
2018-09-20 19:58:01 -07:00
5c6b8eda4f
Remove another duplicate function
2018-09-19 22:18:46 -07:00
ea95a8adef
Combine nearly-duplicate structures
2018-09-19 21:31:46 -07:00
6eb980d1b2
Remove a couple duplicate functions
2018-09-19 21:09:19 -07:00
48b7745a49
Add macros for commonly used square roots
2018-09-19 19:53:25 -07:00
9ef4dd4247
Use ALsizei for the source resample position
2018-09-18 19:08:13 -07:00
952ff84b99
Properly queue buffers for OpenSL capture
2018-09-17 23:43:23 -07:00
b77e6096b8
Fix some potential race conditions with OpenSL
...
For playback, increment the ring buffer's write pointer before queueing audio,
to handle cases where the callback is invoked, advancing the read pointer,
before the write pointer is advanced.
For capture, limit the number of re-queued chunks to the number of fully read
chunks.
2018-09-17 22:49:52 -07:00
7f4441ffbe
Handle the bsinc C resampler like the others
2018-09-17 04:07:56 -07:00
8bacb5dfb8
Fix buffer queue mixing logic
...
In particular, the source sample position was reduced by the size of the
next buffer list item when one is completed, rather than the size of the
one it just completed.
2018-09-16 17:38:55 -07:00
db452a19da
The last reverb loop update doesn't need an aligned count.
2018-09-11 19:05:49 -07:00
99737469e2
Ensure the max reverb update size is a multiple of 4
...
It's not an issue for the final mix, but if one loop has an unaligned count,
the next loop will have unaligned input and output buffer targets which can
crash the SSE mixers.
2018-09-11 18:39:50 -07:00
b13396cce2
Separate the delay line feeding from reading
...
Since it feeds a different line than it reads, the feeding could overwrite
what's subsequently read.
2018-09-10 13:30:20 -07:00
d1f8b78dd4
Avoid a couple line count assumptions
2018-09-10 03:19:41 -07:00
4bdab0051f
Combine identical loops into a separate function
2018-09-10 03:17:14 -07:00
634b13a630
Handle the early reflection delay separate from late refeed
2018-09-10 02:10:35 -07:00
275658b6db
Some suggested changes for iOS
2018-09-07 23:01:17 -07:00
9054f41434
PATH_MAX not MAX_PATH
2018-09-07 22:52:34 -07:00
Chris Robinson