Commit graph

238 commits

Author SHA1 Message Date
James Rowe
e53a2ac411 Reenable AAC FFMPEG decoding
Simple cut/paste issue where initialized is only set to true when the
emulation attempts to init the Binary Pipe, but we used it to test if
the FFMPEG decoder was valid and disabled it if it wasn't. Just return
the value of have_ffmpeg_dl instead so when dynamic loading is added
it'll still work.
2019-12-16 19:23:43 -07:00
James Rowe
87facaa2e2 Prevent out of memory errors when the game passes in an improper length value
HACK

In Luigi's Mansion Dark Moon in HLE audio, the game mysteriously passes
in an extremely large value for length, which without any checks, causes
HLE audio to allocate an extremely large buffer.

This value seemingly is caused by some other HLE audio feature is missing,
and Luigi's Mansion subtracts two values to get a length, without
checking for overflow first. This appears to be caused by an incorrect
HLE audio emulation, as its fixed entirely by only changing to LLE. As
such, further investigation is required, but in the meantime, completely
eating up our users RAM is unacceptable.
2019-12-14 18:18:59 -07:00
James Rowe
337ac73915 Improve aac decoder selection 2019-12-08 16:18:39 -07:00
James Rowe
b395efe804 Dynamically load Media Foundation 2019-12-08 13:54:27 -07:00
vvanelslande
e1e136d72f service/mic: actually use the specified device 2019-11-11 17:15:10 -05:00
zhupengfei
f8b3950829
CMake: separate options for FFmpeg decoder/dumper
Separate options are now provided for FFmpeg AAC audio decoder and FFmpeg video dumper. This allows users to configure Citra with greater freedom.

Also, previously for Linux builds, AAC decoder is accidentally enabled along with the dumper, which could potentially cause patent issues (?). This commit fixes it by only enabling video dumper.
2019-08-20 14:45:39 +08:00
zhupengfei
cf2c354fb9
CMake, CI: Add FFmpeg library dependency 2019-08-14 09:27:03 +08:00
zhupengfei
fff570012f
audio_core: Dump audio frames and samples
This is tested to work with all three current methods - HLE, LLE and multithread LLE.
2019-08-13 19:28:03 +08:00
Weiyi Wang
3dbd1fd4a8 audio_core/cubeb: silent signed/unsigned comparison warning 2019-07-14 14:30:24 -04:00
liushuyu
8021361bb4
audio_core: dsp_hle: use better f32 to s16...
... conversion by clamping and clipping sample to [-1,1] and use
different masks for negative and positive samples
2019-05-01 15:58:26 -06:00
James Rowe
f4d0bdd09a Destroy the callback after the stream is destroyed
Prevents a crash on close if the callback was running on a different
thread.
2019-04-05 14:16:55 -06:00
Vamsi Krishna
aedf5a84b0
Merge pull request #4671 from jroweboy/mic4
Microphone support
2019-03-20 23:12:38 +05:30
James Rowe
b4d538468d Change resampling s16 -> u8 to respect aliasing rules 2019-03-20 09:31:03 -06:00
James Rowe
9739e2b6fd Support signed 8bit pcm in cubeb input. Address review 2019-03-17 22:52:03 -06:00
James Rowe
f5df13eb24 Remove global state and add mic hot swapping 2019-03-06 20:59:58 -07:00
James Rowe
182d672c15 Add ifdef around cubeb for android. Address a few review comments 2019-03-06 20:03:22 -07:00
James Rowe
5c61d53047 Fix compiler error 2019-03-06 11:30:23 -07:00
James Rowe
ba4dfe4217 Increase mic stream volume to max in case the users mic is set low for some reason 2019-03-06 11:17:32 -07:00
James Rowe
5f532c2560 Address review comments 2019-03-06 11:16:43 -07:00
James Rowe
c669aa8d55 Rest of the owl 2019-03-04 23:07:05 -07:00
fearlessTobi
5c9e327ff3 Fix 2019-03-04 23:07:05 -07:00
James Rowe
7fccc995ce Initial Mic setup 2019-03-04 23:06:42 -07:00
Lioncash
d354a2ee3b audio_core/codec: Resolve truncation warnings within DecodeADPCM
The assignments here were performing an implicit truncation from int to
s16. Make it explicit that this is desired behavior.
2019-03-02 14:06:30 +01:00
Weiyi Wang
8f2bd97277
Merge pull request #4655 from tgsm/gcc-version
Remove GCC version checks
2019-02-22 16:21:21 -05:00
tgsm
433ab35e7e Remove GCC version checks
Citra can't be compiled using GCC <7 because of required C++17 support, so these version checks don't need to exist anymore.
2019-02-19 16:39:34 -05:00
tgsm
a400e49338 audio_core: use nested namespaces 2019-02-17 23:41:48 -05:00
liushuyu
a4ba35fe3a audio_core: hle: mf: lint 2019-02-13 15:15:11 -07:00
liushuyu
671ac441e9 audio_core: hle: mf: move transform initializer to ctor 2019-02-13 14:32:14 -07:00
liushuyu
6178cc08b7 audio_core: hle: mf: conform to RAII as possible 2019-02-13 14:08:34 -07:00
liushuyu
f0e041e27a audio_core: hle: mf: correctly handle stream change 2019-02-09 16:49:48 -07:00
liushuyu
01e0902fa4 audio_core: hle: mf: fix a memory accessing issue 2019-02-09 13:55:54 -07:00
liushuyu
6281660844 audio_core: hle: mf: address yet another batch of reviews 2019-02-09 13:55:42 -07:00
liushuyu
ab1f47ed15 audio_core: hle: mf: make DetectMediaType return a struct 2019-02-09 11:57:36 +01:00
liushuyu
168f2ee79a audio_core: mf: make initialize function return smart pointer 2019-02-09 11:57:36 +01:00
liushuyu
c91f5029ff audio_core: hle: mf: fix a regression...
... where the sample copying logic is incorrect due to the wrong usage
of `std::array`
2019-02-09 11:57:36 +01:00
liushuyu
6332e57069 audio_core: hle: mf: return values as std::optional 2019-02-09 11:57:36 +01:00
liushuyu
25fa10327b audio_core: hle: mf: use object proxy 2019-02-09 11:57:36 +01:00
liushuyu
26b3b41788 audio_core: hle: mf: wrap enum in class 2019-02-09 11:57:36 +01:00
liushuyu
633f02b9b2 audio_core: hle: mf: avoid name conflicts 2019-02-09 11:57:36 +01:00
liushuyu
452ac7b874 audio_core: hle: mf: use more enum 2019-02-09 11:57:36 +01:00
liushuyu
be764e4f88 audio_core: hle: mf: multiple fixes...
... more smart pointers and re-arrange code
2019-02-09 11:57:36 +01:00
liushuyu
4bc6bfd51f audio_core: hle: mf: re-arrange comments 2019-02-09 11:57:36 +01:00
liushuyu
972b527374 audio_core: hle: mf: cleanup headers 2019-02-09 11:57:36 +01:00
liushuyu
2c40ef2a90 audio_core: hle: mf: rename functions 2019-02-09 11:57:36 +01:00
liushuyu
ea8a1f8754 Media Foundation Memory Fix
* audio_core: hle: mf: more smart pointers

* audio_core: hle: mf: fix memory leaks

* audio_core: hle: mf: even more smart pointers
2019-02-09 11:57:36 +01:00
liushuyu
10f876653d audio_core: hle: mf: transition to use smart pointers 2019-02-09 11:56:51 +01:00
liushuyu
bee5ba3e33 audio_core: hle: mf: fix a regression...
... an introduction in 8f591d3b85
2019-02-09 11:56:51 +01:00
liushuyu
06316be8a7 audio_core: hle: mf: minor fix 2019-02-09 11:56:51 +01:00
liushuyu
c03861c2d9 audio_core: hle: mf: address another batch of reviews from @B3N30 2019-02-09 11:56:51 +01:00
liushuyu
7f5b54fda4 audio_core: hle: mf: address reviews from @B3N30 2019-02-09 11:56:51 +01:00
liushuyu
11e277149c audio_core: hle: fix compile 2019-02-09 11:56:51 +01:00
B3N30
80b4dd21d2 audio_core: dsp_hle: add Media Foundation decoder...
* appveyor: switch to Media Foundation API
* Travis CI MinGW build needs an update with the container image
2019-02-09 11:56:51 +01:00
B3N30
1581dea6de more reviews addressed 2019-02-09 11:55:12 +01:00
liushuyu
8fe3e37df5 audio_core: fix msvc include issue 2019-02-09 11:55:12 +01:00
B3N30
bd9984b5f8 Addressed reviews 2019-02-09 11:55:12 +01:00
B3N30
847003cc1c FFmpegDL: Added external_dlls folder to user folder
AACDecoder: addressed reviews
2019-02-09 11:52:31 +01:00
B3N30
45b5de7a18 Fixup: Merge conflict with dsp lle 2019-02-09 11:52:31 +01:00
B3N30
c521f3b7d6 CoreAudio::HLE: Add FFmpeg aac decoder 2019-02-09 11:52:31 +01:00
Weiyi Wang
f78512d6de
Merge pull request #4599 from wwylele/bitfield-endian
Common: make BitField endianness-aware
2019-02-01 10:48:56 -05:00
Weiyi Wang
104829ae58 remove all occurance of specifying endianness inside BitField
This commit it automatically generated by command in zsh:
sed -i -- 's/BitField<\(.*\)_le>/BitField<\1>/g' **/*(D.)

BitField is now aware to endianness and default to little endian. It expects a value representation type without storage specification for its template parameter.
2019-01-28 22:09:43 -05:00
fearlessTobi
24f966481f dsp_interface: fix sound being played while volume is 0
According to documentation, if the argument of std::exp is zero, one is returned.
However we want the return value to be also zero in this case so no audio is played.
2019-01-26 14:53:58 +01:00
Otávio Pace
6560fac5ff audio_core: remove unnecessary spaces on comments 2019-01-19 21:28:52 +01:00
Ben
4f23d5d69e
Merge pull request #4490 from wwylele/teakra-new
audio: implement DSP LLE
2019-01-11 16:45:00 +01:00
Weiyi Wang
443f4b964d DSP/LLE: add multithread mode 2018-12-21 20:49:24 -05:00
Weiyi Wang
969dc3b46f Update teakra and address more comments 2018-12-21 20:49:23 -05:00
Weiyi Wang
05c372bf6c Update teakra to fix macos issue; address comment feedbacks 2018-12-21 20:49:23 -05:00
Weiyi Wang
67213ca855 audio_core/lle: link ahbm and audio callback 2018-12-21 20:49:22 -05:00
Weiyi Wang
483680a124 audio_core/lle: only process pipe signal after loading 2018-12-21 20:49:22 -05:00
Weiyi Wang
6f6ffceec4 audio_core/hle,lle: implement UnloadComponent 2018-12-21 20:49:21 -05:00
Weiyi Wang
9b41e6f85f audio_core/lle: implement LoadComponent 2018-12-21 20:49:21 -05:00
Weiyi Wang
4add509b20 audio_core/hle: move implementation of LoadComponent 2018-12-21 20:49:21 -05:00
Weiyi Wang
9ab4e3c686 audio_core/lle: implement GetDspMemory and SetServiceToInterrupt 2018-12-21 20:49:21 -05:00
Weiyi Wang
5c13eca700 audio_core/lle: implement PipeRead/GetPipeReadableSize 2018-12-21 20:49:20 -05:00
Weiyi Wang
e40efbf37b audio_core/lle: implement PipeWrite 2018-12-21 20:49:20 -05:00
Weiyi Wang
c9736b3300 audio_core/hle,lle: implement SetSemaphore 2018-12-21 20:49:20 -05:00
Weiyi Wang
4671ccf416 audio_core/lle: implement RecvDataIsReady 2018-12-21 20:49:19 -05:00
Weiyi Wang
14eb3561de audio_core/hle: move implementation of RecvDataIsReady 2018-12-21 20:49:19 -05:00
Weiyi Wang
b609753fb6 audio_core/lle: implement RecvData 2018-12-21 20:49:19 -05:00
Weiyi Wang
7f568a3c19 DSP/HLE: move implementation of RecvData to audio_core, behind common interface 2018-12-21 20:49:19 -05:00
Weiyi Wang
6d51d95d44 audio_core: add teakra and lle interface 2018-12-21 20:49:18 -05:00
Lioncash
90082268dc audio_core: Make g_sink_details internally linked
We can hide the direct array from external view and instead provide
functions to retrieve the necessary info. This has the benefit of
completely hiding the makeup of the SinkDetails structure from the rest
of the code.

Given that this makes the array hidden, we can also make the array
constexpr by altering the members slightly. This gets rid of several
static constructor calls related to std::vector and std::function.

Now we don't have heap allocations here that need to occur before the
program can even enter main(). It also has the benefit of saving a
little bit of heap space, but this doesn't matter too much, since the
savings in that regard are pretty tiny.
2018-12-16 02:44:04 +01:00
Weiyi Wang
296c458e0e Memory: move GetPhysicalPointer and IsValidPhysicalAddress into class 2018-12-05 20:21:14 -05:00
Weiyi Wang
ec01975549 audio_core: pass in memory reference
Because HLE::Source is initialized as an array in the member initializer, it is hard to let it accept the reference on ctor, so it has a second init stage performed by DspHle::Impl::Impl
2018-12-05 20:16:42 -05:00
RoadrunnerWMC
008ef1fd71 DSP: Add a comment about physical address masking
See @wwylele's comment on PR 4483 for more details on what causes this behavior.
2018-12-03 12:22:11 -07:00
RoadrunnerWMC
3aff585f32 DSP: Add address mask for physical pointers to audio data buffers
Hardware testing indicated that FFFFFFFC is the correct mask for all audio formats (mono and stereo PCM8, mono and stereo PCM16, and ADPCM). This fixes broken audio in Luigi's Mansion: Dark Moon and a few other games.
2018-12-03 01:07:55 -07:00
Weiyi Wang
9458e4d8ec CoreTiming: wrap into class 2018-11-04 10:26:38 -05:00
Weiyi Wang
b4062abc11 cubeb_sink: ignore null-name device when selecting
We already ignore them on listing devices. We should do the same when selecting devices. This fix a crash when opening a specific device while there is a null device in the list
2018-10-26 09:17:44 -04:00
Subv
1d59c8f792 Logging: Change the TimeStretch::Process log from debug to trace level.
This function is called too many times and makes the debug logging basically unusable due to the spam.
2018-10-06 16:55:05 +02:00
Weiyi Wang
7db46da4fc
Merge pull request #4247 from FearlessTobi/port-1314
Port yuzu-emu/yuzu#1314: "audio_core/time_stretch: Silence truncation warnings in Process()"
2018-09-22 12:02:51 -04:00
Lioncash
bcb1aaf05c cubeb_sink: Get rid of variable shadowing within CubebSink's constructor
The parameter of the lambda was shadowing the variable that was being
assigned to.
2018-09-22 12:55:40 +02:00
Lioncash
553bdb2ae4 audio_core/time_stretch: Silence truncation warnings in Process()
The SoundTouch API only accepts uint amount of samples.
2018-09-22 12:52:13 +02:00
Merry
bb9e92c77c
Merge pull request #4194 from MerryMage/audiofifo
audio_core: Simplify sink interface
2018-09-21 13:30:51 +01:00
Lioncash
93bd630d40 audio_core/sink_details: Change std::string parameter into std::string_view
The given string is only ever used for lookup and comparison, so we can
just utilize a non-owning view to string data here
2018-09-15 15:48:24 +02:00
MerryMage
a6cf2e1f9d cubeb_sink: Improve logging 2018-09-09 08:37:34 +01:00
MerryMage
675ffc1024 dsp_interface: Move volume control to audio thread
We also clamp the linear volume value to [0.0, 1.0].
Do nothing if linear volume is 1.0.
2018-09-08 22:24:51 +01:00
MerryMage
eed55a813e time_stretch: Simplify audio stretcher 2018-09-08 22:09:26 +01:00
MerryMage
f34711219a audio_core: Simplify sink interface 2018-09-08 22:09:26 +01:00
Weiyi Wang
7d8f115185 Prefix all size_t with std::
done automatically by executing regex replace `([^:0-9a-zA-Z_])size_t([^0-9a-zA-Z_])` -> `$1std::size_t$2`
2018-09-06 16:03:28 -04:00
Merry
b130dd6842
Merge pull request #4161 from FearlessTobi/port-1139
Port #1139 from yuzu: "bit_field: Convert ToBool() into explicit operator bool"
2018-09-02 09:53:50 +01:00
fearlessTobi
488694d01f Replace ToBool() with static_cast() 2018-09-01 15:50:57 +02:00