1 /*****************************************************************************
2 * qtsound.m: qtkit (Mac OS X) based audio capture module
3 *****************************************************************************
4 * Copyright © 2011 VLC authors and VideoLAN
6 * Authors: Pierre d'Herbemont <pdherbemont@videolan.org>
7 * Gustaf Neumann <neumann@wu.ac.at>
8 * Michael S. Feurstein <michael.feurstein@wu.ac.at>
10 *****************************************************************************
11 * This library is free software; you can redistribute it and/or
12 * modify it under the terms of the GNU Lesser General Public License
13 * as published by the Free Software Foundation; either version 2.1
14 * of the License, or (at your option) any later version.
16 * This library is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 * Lesser General Public License for more details.
21 * You should have received a copy of the GNU Lesser General Public
22 * License along with this library; if not, write to the Free Software
23 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110, USA
25 *****************************************************************************/
27 /*****************************************************************************
29 *****************************************************************************/
35 #include <vlc_common.h>
36 #include <vlc_plugin.h>
39 #include <vlc_demux.h>
40 #include <vlc_dialog.h>
42 #define QTKIT_VERSION_MIN_REQUIRED 70603
44 #import <QTKit/QTKit.h>
46 /*****************************************************************************
48 *****************************************************************************/
49 static int Open(vlc_object_t *p_this);
50 static void Close(vlc_object_t *p_this);
51 static int Demux(demux_t *p_demux);
52 static int Control(demux_t *, int, va_list);
54 /*****************************************************************************
56 *****************************************************************************/
59 set_shortname(N_("QTSound"))
60 set_description(N_("QuickTime Sound Capture"))
61 set_category(CAT_INPUT)
62 set_subcategory(SUBCAT_INPUT_ACCESS)
63 add_shortcut("qtsound")
64 set_capability("access_demux", 0)
65 set_callbacks(Open, Close)
69 /*****************************************************************************
71 *****************************************************************************/
72 @interface VLCDecompressedAudioOutput : QTCaptureDecompressedAudioOutput
75 AudioBuffer *currentAudioBuffer;
77 UInt32 numberOfSamples;
82 - (id)initWithDemux:(demux_t *)p_demux;
83 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
84 - (BOOL)checkCurrentAudioBuffer;
86 - (mtime_t)getCurrentPts;
87 - (void *)getCurrentAudioBufferData;
88 - (UInt32)getCurrentTotalDataSize;
89 - (UInt32)getNumberOfSamples;
93 @implementation VLCDecompressedAudioOutput : QTCaptureDecompressedAudioOutput
94 - (id)initWithDemux:(demux_t *)p_demux
96 if (self = [super init]) {
98 currentAudioBuffer = nil;
99 date_Init(&date, 44100, 1);
111 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection
113 AudioBufferList *tempAudioBufferList;
114 UInt32 totalDataSize = 0;
117 @synchronized (self) {
118 numberOfSamples = [sampleBuffer numberOfSamples];
119 date_Increment(&date,numberOfSamples);
120 currentPts = date_Get(&date);
122 tempAudioBufferList = [sampleBuffer audioBufferListWithOptions:0];
123 if (tempAudioBufferList->mNumberBuffers == 2) {
125 * Compute totalDataSize as sum of all data blocks in the
128 for (count = 0; count < tempAudioBufferList->mNumberBuffers; count++)
129 totalDataSize += tempAudioBufferList->mBuffers[count].mDataByteSize;
132 * Allocate storage for the interleaved audio data
134 rawAudioData = malloc(totalDataSize);
135 if (NULL == rawAudioData) {
136 msg_Err(p_qtsound, "Raw audiodata could not be allocated");
140 msg_Err(p_qtsound, "Too many or only one channel found: %i.",
141 tempAudioBufferList->mNumberBuffers);
146 * Interleave raw data (provided in two separate channels as
147 * F32L) with 2 samples per frame
151 const float *b1Ptr, *b2Ptr;
155 uPtr = (float *)rawAudioData,
156 b1Ptr = (const float *) tempAudioBufferList->mBuffers[0].mData,
157 b2Ptr = (const float *) tempAudioBufferList->mBuffers[1].mData;
158 i < numberOfSamples; i++) {
163 if (currentAudioBuffer == nil) {
164 currentAudioBuffer = (AudioBuffer *)malloc(sizeof(AudioBuffer));
165 if (NULL == currentAudioBuffer) {
166 msg_Err(p_qtsound, "AudioBuffer could not be allocated.");
170 currentAudioBuffer->mNumberChannels = 2;
171 currentAudioBuffer->mDataByteSize = totalDataSize;
172 currentAudioBuffer->mData = rawAudioData;
177 - (BOOL)checkCurrentAudioBuffer
179 return (currentAudioBuffer) ? 1 : 0;
184 FREENULL(rawAudioData);
187 - (mtime_t)getCurrentPts
189 /* FIXME: can this getter be minimized? */
192 if(!currentAudioBuffer || currentPts == previousPts)
195 @synchronized (self) {
196 pts = previousPts = currentPts;
199 return (currentAudioBuffer->mData) ? currentPts : 0;
202 - (void *)getCurrentAudioBufferData
204 return currentAudioBuffer->mData;
207 - (UInt32)getCurrentTotalDataSize
209 return currentAudioBuffer->mDataByteSize;
212 - (UInt32)getNumberOfSamples
214 return numberOfSamples;
219 /*****************************************************************************
221 *****************************************************************************/
224 QTCaptureSession * session;
225 QTCaptureDevice * audiodevice;
226 VLCDecompressedAudioOutput * audiooutput;
227 es_out_id_t *p_es_audio;
230 /*****************************************************************************
231 * Open: initialize interface
232 *****************************************************************************/
233 static int Open(vlc_object_t *p_this)
235 demux_t *p_demux = (demux_t*)p_this;
237 es_format_t audiofmt;
238 char *psz_uid = NULL;
241 NSString *qtk_curraudiodevice_uid;
242 NSAutoreleasePool *pool;
243 NSArray *myAudioDevices, *audioformat_array;
244 QTFormatDescription *audio_format;
245 QTCaptureDeviceInput *audioInput;
246 NSError *o_returnedAudioError;
248 if(p_demux->psz_location && *p_demux->psz_location)
249 psz_uid = p_demux->psz_location;
251 msg_Dbg(p_demux, "qtsound uid = %s", psz_uid);
252 qtk_curraudiodevice_uid = [[NSString alloc] initWithFormat:@"%s", psz_uid];
254 pool = [[NSAutoreleasePool alloc] init];
256 p_demux->p_sys = p_sys = calloc(1, sizeof(demux_sys_t));
260 msg_Dbg(p_demux, "qtsound : uid = %s", [qtk_curraudiodevice_uid UTF8String]);
261 myAudioDevices = [[[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeSound] arrayByAddingObjectsFromArray:[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeMuxed]] retain];
262 if([myAudioDevices count] == 0) {
263 dialog_FatalWait(p_demux, _("No Audio Input device found"),
264 _("Your Mac does not seem to be equipped with a suitable audio input device."
265 "Please check your connectors and drivers."));
266 msg_Err(p_demux, "Can't find any Audio device");
271 for (iaudio = 0; iaudio < [myAudioDevices count]; iaudio++) {
272 QTCaptureDevice *qtk_audioDevice;
273 qtk_audioDevice = [myAudioDevices objectAtIndex:iaudio];
274 msg_Dbg(p_demux, "qtsound audio %u/%lu localizedDisplayName: %s uniqueID: %s", iaudio, [myAudioDevices count], [[qtk_audioDevice localizedDisplayName] UTF8String], [[qtk_audioDevice uniqueID] UTF8String]);
275 if ([[[qtk_audioDevice uniqueID]stringByTrimmingCharactersInSet:[NSCharacterSet whitespaceCharacterSet]] isEqualToString:qtk_curraudiodevice_uid]) {
276 msg_Dbg(p_demux, "Device found");
282 if(iaudio < [myAudioDevices count])
283 p_sys->audiodevice = [myAudioDevices objectAtIndex:iaudio];
285 /* cannot find designated audio device, fall back to open default audio device */
286 msg_Dbg(p_demux, "Cannot find designated uid audio device as %s. Fall back to open default audio device.", [qtk_curraudiodevice_uid UTF8String]);
287 p_sys->audiodevice = [QTCaptureDevice defaultInputDeviceWithMediaType: QTMediaTypeSound];
289 if(!p_sys->audiodevice) {
290 dialog_FatalWait(p_demux, _("No audio input device found"),
291 _("Your Mac does not seem to be equipped with a suitable audio input device."
292 "Please check your connectors and drivers."));
293 msg_Err(p_demux, "Can't find any Audio device");
298 if(![p_sys->audiodevice open: &o_returnedAudioError]) {
299 msg_Err(p_demux, "Unable to open the audio capture device (%ld)", [o_returnedAudioError code]);
303 if([p_sys->audiodevice isInUseByAnotherApplication] == YES) {
304 msg_Err(p_demux, "default audio capture device is exclusively in use by another application");
307 audioInput = [[QTCaptureDeviceInput alloc] initWithDevice: p_sys->audiodevice];
309 msg_Err(p_demux, "can't create a valid audio capture input facility");
312 msg_Dbg(p_demux, "created valid audio capture input facility");
314 p_sys->audiooutput = [[VLCDecompressedAudioOutput alloc] initWithDemux:p_demux];
315 msg_Dbg (p_demux, "initialized audio output");
317 /* Get the formats */
319 FIXME: the format description gathered here does not seem to be the same
320 in comparison to the format description collected from the actual sampleBuffer.
321 This information needs to be updated some other place. For the time being this shall suffice.
323 The following verbose output is an example of what is read from the input device during the below block
324 [0x3042138] qtsound demux debug: Audio localized format summary: Linear PCM, 24 bit little-endian signed integer, 2 channels, 44100 Hz
325 [0x3042138] qtsound demux debug: Sample Rate: 44100; Format ID: lpcm; Format Flags: 00000004; Bytes per Packet: 8; Frames per Packet: 1; Bytes per Frame: 8; Channels per Frame: 2; Bits per Channel: 24
326 [0x3042138] qtsound demux debug: Flag float 0 bigEndian 0 signedInt 1 packed 0 alignedHigh 0 non interleaved 0 non mixable 0
327 canonical 0 nativeFloatPacked 0 nativeEndian 0
329 However when reading this information from the sampleBuffer during the delegate call from
330 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
331 the following data shows up
332 2011-09-23 22:06:03.077 VLC[23070:f103] Audio localized format summary: Linear PCM, 32 bit little-endian floating point, 2 channels, 44100 Hz
333 2011-09-23 22:06:03.078 VLC[23070:f103] Sample Rate: 44100; Format ID: lpcm; Format Flags: 00000029; Bytes per Packet: 4; Frames per Packet: 1; Bytes per Frame: 4; Channels per Frame: 2; Bits per Channel: 32
334 2011-09-23 22:06:03.078 VLC[23070:f103] Flag float 1 bigEndian 0 signedInt 0 packed 1 alignedHigh 0 non interleaved 1 non mixable 0
335 canonical 1 nativeFloatPacked 1 nativeEndian 0
339 little-endian signed integer vs. little-endian floating point
340 format flag 00000004 vs. 00000029
341 bytes per packet 8 vs. 4
343 non interleaved 0 vs. 1 -> this makes a major difference when filling our own buffer
345 nativeFloatPacked 0 vs. 1
347 One would assume we'd need to feed the (es_format_t)audiofmt with the data collected here.
348 This is not the case. Audio will be transmitted in artefacts, due to wrong information.
350 At the moment this data is set manually, however one should consider trying to set this data dynamically
352 audioformat_array = [p_sys->audiodevice formatDescriptions];
354 for(int k = 0; k < [audioformat_array count]; k++) {
355 audio_format = (QTFormatDescription *)[audioformat_array objectAtIndex:k];
357 msg_Dbg(p_demux, "Audio localized format summary: %s", [[audio_format localizedFormatSummary] UTF8String]);
358 msg_Dbg(p_demux, "Audio format description attributes: %s",[[[audio_format formatDescriptionAttributes] description] UTF8String]);
360 AudioStreamBasicDescription asbd = {0};
361 NSValue *asbdValue = [audio_format attributeForKey:QTFormatDescriptionAudioStreamBasicDescriptionAttribute];
362 [asbdValue getValue:&asbd];
364 char formatIDString[5];
365 UInt32 formatID = CFSwapInt32HostToBig (asbd.mFormatID);
366 bcopy (&formatID, formatIDString, 4);
367 formatIDString[4] = '\0';
369 /* kept for development purposes */
371 msg_Dbg(p_demux, "Sample Rate: %.0lf; Format ID: %s; Format Flags: %.8x; Bytes per Packet: %d; Frames per Packet: %d; Bytes per Frame: %d; Channels per Frame: %d; Bits per Channel: %d",
375 asbd.mBytesPerPacket,
376 asbd.mFramesPerPacket,
378 asbd.mChannelsPerFrame,
379 asbd.mBitsPerChannel);
381 msg_Dbg(p_demux, "Flag float %d bigEndian %d signedInt %d packed %d alignedHigh %d non interleaved %d non mixable %d\ncanonical %d nativeFloatPacked %d nativeEndian %d",
382 (asbd.mFormatFlags & kAudioFormatFlagIsFloat) != 0,
383 (asbd.mFormatFlags & kAudioFormatFlagIsBigEndian) != 0,
384 (asbd.mFormatFlags & kAudioFormatFlagIsSignedInteger) != 0,
385 (asbd.mFormatFlags & kAudioFormatFlagIsPacked) != 0,
386 (asbd.mFormatFlags & kAudioFormatFlagIsAlignedHigh) != 0,
387 (asbd.mFormatFlags & kAudioFormatFlagIsNonInterleaved) != 0,
388 (asbd.mFormatFlags & kAudioFormatFlagIsNonMixable) != 0,
390 (asbd.mFormatFlags & kAudioFormatFlagsCanonical) != 0,
391 (asbd.mFormatFlags & kAudioFormatFlagsNativeFloatPacked) != 0,
392 (asbd.mFormatFlags & kAudioFormatFlagsNativeEndian) != 0
397 if([audioformat_array count])
398 audio_format = [audioformat_array objectAtIndex:0];
402 /* Now we can init */
403 audiocodec = VLC_CODEC_FL32;
404 es_format_Init(&audiofmt, AUDIO_ES, audiocodec);
406 audiofmt.audio.i_format = audiocodec;
407 audiofmt.audio.i_rate = 44100;
409 * i_physical_channels Describes the channels configuration of the
410 * samples (ie. number of channels which are available in the
411 * buffer, and positions).
413 audiofmt.audio.i_physical_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
415 * i_original_channels Describes from which original channels,
416 * before downmixing, the buffer is derived.
418 audiofmt.audio.i_original_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
420 * Please note that it may be completely arbitrary - buffers are not
421 * obliged to contain a integral number of so-called "frames". It's
422 * just here for the division:
423 * buffer_size = i_nb_samples * i_bytes_per_frame / i_frame_length
425 audiofmt.audio.i_bitspersample = 32;
426 audiofmt.audio.i_channels = 2;
427 audiofmt.audio.i_blockalign = audiofmt.audio.i_channels * (audiofmt.audio.i_bitspersample / 8);
428 audiofmt.i_bitrate = audiofmt.audio.i_channels * audiofmt.audio.i_rate * audiofmt.audio.i_bitspersample;
430 p_sys->session = [[QTCaptureSession alloc] init];
432 success = [p_sys->session addInput:audioInput error: &o_returnedAudioError];
434 msg_Err(p_demux, "the audio capture device could not be added to capture session (%ld)", [o_returnedAudioError code]);
438 success = [p_sys->session addOutput:p_sys->audiooutput error: &o_returnedAudioError];
440 msg_Err(p_demux, "audio output could not be added to capture session (%ld)", [o_returnedAudioError code]);
444 [p_sys->session startRunning];
447 p_demux->pf_demux = Demux;
448 p_demux->pf_control = Control;
449 p_demux->info.i_update = 0;
450 p_demux->info.i_title = 0;
451 p_demux->info.i_seekpoint = 0;
453 msg_Dbg(p_demux, "New audio es %d channels %dHz",
454 audiofmt.audio.i_channels, audiofmt.audio.i_rate);
456 p_sys->p_es_audio = es_out_Add(p_demux->out, &audiofmt);
458 [audioInput release];
461 msg_Dbg(p_demux, "QTSound: We have an audio device ready!");
465 [audioInput release];
473 /*****************************************************************************
474 * Close: destroy interface
475 *****************************************************************************/
476 static void Close(vlc_object_t *p_this)
478 NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
479 demux_t *p_demux = (demux_t*)p_this;
480 demux_sys_t *p_sys = p_demux->p_sys;
482 [p_sys->session performSelectorOnMainThread:@selector(stopRunning) withObject:nil waitUntilDone:NO];
483 [p_sys->audiooutput performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
484 [p_sys->session performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
491 /*****************************************************************************
493 *****************************************************************************/
494 static int Demux(demux_t *p_demux)
496 demux_sys_t *p_sys = p_demux->p_sys;
497 block_t *p_blocka = nil;
498 NSAutoreleasePool *pool;
502 @synchronized (p_sys->audiooutput) {
503 if ([p_sys->audiooutput checkCurrentAudioBuffer]) {
504 unsigned i_buffer_size = [p_sys->audiooutput getCurrentTotalDataSize];
505 p_blocka = block_Alloc(i_buffer_size);
508 msg_Err(p_demux, "cannot get audio block");
512 memcpy(p_blocka->p_buffer, [p_sys->audiooutput getCurrentAudioBufferData], i_buffer_size);
513 p_blocka->i_nb_samples = [p_sys->audiooutput getNumberOfSamples];
514 p_blocka->i_pts = [p_sys->audiooutput getCurrentPts];
516 [p_sys->audiooutput freeAudioMem];
521 if (!p_blocka->i_pts) {
522 block_Release(p_blocka);
524 // Nothing to transfer yet, just forget
529 es_out_Control(p_demux->out, ES_OUT_SET_PCR, p_blocka->i_pts);
530 es_out_Send(p_demux->out, p_sys->p_es_audio, p_blocka);
536 /*****************************************************************************
538 *****************************************************************************/
539 static int Control(demux_t *p_demux, int i_query, va_list args)
545 /* Special for access_demux */
546 case DEMUX_CAN_PAUSE:
548 case DEMUX_SET_PAUSE_STATE:
549 case DEMUX_CAN_CONTROL_PACE:
550 pb = (bool*)va_arg(args, bool *);
554 case DEMUX_GET_PTS_DELAY:
555 pi64 = (int64_t*)va_arg(args, int64_t *);
556 *pi64 = INT64_C(1000) * var_InheritInteger(p_demux, "live-caching");