aboutsummaryrefslogtreecommitdiffstats
path: root/Alc
Commit message (Collapse)AuthorAgeFilesLines
* Don't use periphonic FOA when the HOA decoder is not periphonicChris Robinson2017-02-195-41/+55
|
* Remove the sinc8 resampler optionChris Robinson2017-02-196-286/+26
| | | | | Perf shows less than 1 percent CPU difference from the higher quality bsinc resampler, but uses almost twice as much memory (a 128KB lookup table).
* Always lock the device backend before calling aluMixDataChris Robinson2017-02-1811-50/+63
|
* Return some device latency by defaultChris Robinson2017-02-181-2/+7
| | | | | | | A device will never have 0 latency. OpenAL Soft itself uses a sample buffer length of UpdateSize*NumUpdates, and during playback will have about (NumUpdates-1) periods filled, more or less. Without a more accurate measurement from the playback system, this is better than reporting 0.
* Use select() to wait for audio with OSS and SolarisChris Robinson2017-02-182-85/+137
|
* Reorganize ALvoice membersChris Robinson2017-02-152-88/+95
| | | | | This places the Send[] array at the end of the struct, making it easier to handle dynamically.
* Make ALsourceProps' Send array dynamically sizedChris Robinson2017-02-143-20/+102
| | | | | | ALsourceProps' Send[] array is placed at the end of the struct, and given an indeterminate size. Extra space is allocated at the end of each struct given the number of auxiliary sends set for the device.
* Make the source state atomicChris Robinson2017-02-133-7/+8
| | | | | Since it's modified by the mixer when playback is ended, a plain struct member isn't safe.
* Put BsincState in a generic unionChris Robinson2017-02-139-75/+75
|
* Porperly check for and use __builtin_assume_alignedChris Robinson2017-02-132-12/+20
|
* Clean up the bsinc mixer a bitChris Robinson2017-02-123-26/+28
|
* Add NEON-enhanced resamplersChris Robinson2017-02-123-4/+294
|
* Fix more uses of unsigned sizes and offsetsChris Robinson2017-02-102-9/+9
|
* Remove a couple context lock wrapper functionsChris Robinson2017-02-071-3/+0
|
* Properly capitalize NEONChris Robinson2017-02-071-1/+1
|
* Clear trailing whitespace from the cpu features stringChris Robinson2017-02-071-0/+5
|
* Use the correct IID for the opensl buffer queueChris Robinson2017-02-071-2/+3
|
* Convert the OpenSL backend to the new backend APIChris Robinson2017-02-073-177/+434
| | | | | This also removes the buffer queue callback's call to aluMixData, which could potentially block on a mutex.
* Fix for NULL JNIEnvChris Robinson2017-02-052-5/+9
| | | | Which can happen with native-only apps
* Call getSystemService as a non-static functionChris Robinson2017-02-031-4/+4
|
* Android's AudioManager.getProperty(String) returns a StringChris Robinson2017-02-021-1/+1
|
* Replace more ALuint with ALsizeiChris Robinson2017-01-293-15/+15
|
* Fix late reverb low-pass filteringChris Robinson2017-01-291-1/+1
|
* Use an all-pass series on each late reverb lineChris Robinson2017-01-281-261/+396
| | | | | | | | | This attempts to improve the smoothness of the late reverb decay by passing each line through multiple all-pass filters. Some work is still needed to work better in high-density and not-so-high-diffusion environments. This also removes the decay from the early reflections, since it's no longer continuous feedback.
* Add more traces for the Java calls being madeChris Robinson2017-01-271-7/+22
|
* Round and clamp the scaled update count with openslChris Robinson2017-01-271-1/+4
|
* Try to use the system sample rate with AndroidChris Robinson2017-01-261-6/+56
|
* Get the JavaVM handle on Android targetsChris Robinson2017-01-262-0/+72
|
* Improve the ambisonic upscaling methodsChris Robinson2017-01-243-129/+174
| | | | | | | | This now takes advantage of the differences seen in generated decoder matrices for first-order compared to second- and third-order, such that with the appropriate frequency-dependent scaling applied to first-order content, the result is identical with a higher-order decoder matrix compared to a first- order matrix for the same layout.
* Fix coefficient counts for the built-in B-Format decodersChris Robinson2017-01-211-2/+2
|
* Use a flat sqrt(2) scale for non-directional ambient gainsChris Robinson2017-01-211-8/+3
|
* Update the default basic B-Format decodersChris Robinson2017-01-211-47/+36
| | | | This also converts them to ACN/N3D format.
* Move the B-Format HRTF virtual speaker stuff to InitHrtfPanningChris Robinson2017-01-183-88/+76
| | | | | This keeps the decoder matrices and coefficient mapping together for if it changes in the future.
* Replace some ALvoid with voidChris Robinson2017-01-182-3/+3
|
* Use ALsizei in more placesChris Robinson2017-01-185-55/+55
|
* Pass the left and right buffers to the hrtf mixers directlyChris Robinson2017-01-174-32/+36
|
* Small update for the "virtual" Ambi2DDecoder coefficientsChris Robinson2017-01-171-5/+5
|
* More ALsizei, with the B-Format decoderChris Robinson2017-01-165-26/+35
|
* Fix a couple hard-coded array sizesChris Robinson2017-01-161-2/+2
|
* Use ALsizei in a few more placesChris Robinson2017-01-164-9/+9
|
* Use ALsizei and ALint for sizes and offsets with resamplers and filtersChris Robinson2017-01-167-71/+71
|
* Use ALsizei for sizes and offsets with the mixerChris Robinson2017-01-1610-171/+171
| | | | | | Unsigned 32-bit offsets actually have some potential overhead on 64-bit targets for pointer/array accesses due to rules on integer wrapping. No idea how much impact it has in practice, but it's nice to be correct about it.
* Use second-order ambisonics for basic HRTF renderingChris Robinson2017-01-155-23/+93
| | | | | | This should improve positional quality for relatively low cost. Full HRTF rendering still only uses first-order since the only use of the dry buffer there is for first-order content (B-Format buffers, effects).
* Avoid using some LP typesChris Robinson2017-01-101-5/+5
|
* Use proper atomics in the OSS backendChris Robinson2017-01-101-17/+17
|
* Hold Pulse's mainloop lock while calling capture functionsChris Robinson2017-01-101-2/+10
| | | | | | | Since commit c837484015e, the backend's lock is no longer implicitly held when calling capture functions. A separate mutex is used to ensure serial access, and its up to the backend to protect against races and reentrancy with the audio API.
* One more update for the HRTF B-Format coefficientsChris Robinson2017-01-091-14/+14
| | | | These should better represent the pseudo-inverse matrices with N3D scaling.
* Add missing AL_EFFECTSLOT_ properties for al(c)GetEnumValueChris Robinson2017-01-051-0/+5
|
* Update the B-Format HRTF coefficients to use the pseudo-inverse matrixChris Robinson2017-01-041-26/+14
| | | | | | It's hard to tell which is ultimately better, although this way does make the FOA output somewhat louder which will help when it's combined with direct HRTF rendering.
* Avoid writing to the same buffer that's read fromChris Robinson2016-12-231-36/+36
| | | | Also clean up comment formatting a bit.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-03 00:02:28 +0000