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
43 #define QTKIT_VERSION_MAX_ALLOWED 70700
45 #import <QTKit/QTKit.h>
47 /*****************************************************************************
49 *****************************************************************************/
50 static int Open( vlc_object_t *p_this );
51 static void Close( vlc_object_t *p_this );
52 static int Demux( demux_t *p_demux );
53 static int Control( demux_t *, int, va_list );
55 /*****************************************************************************
57 *****************************************************************************/
60 set_shortname( N_("QTSound") )
61 set_description( N_("QuickTime Sound Capture") )
62 set_category( CAT_INPUT )
63 set_subcategory( SUBCAT_INPUT_ACCESS )
64 add_shortcut( "qtsound" )
65 set_capability( "access_demux", 0 )
66 set_callbacks( Open, Close )
70 /*****************************************************************************
72 *****************************************************************************/
73 @interface VLCDecompressedAudioOutput : QTCaptureDecompressedAudioOutput
76 AudioBuffer *currentAudioBuffer;
77 block_t *rawAudioData;
78 UInt32 numberOfSamples;
83 - (id)initWithDemux:(demux_t *)p_demux;
84 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
85 - (BOOL)checkCurrentAudioBuffer;
87 - (mtime_t)getCurrentPts;
88 - (void *)getCurrentAudioBufferData;
89 - (UInt32)getCurrentTotalDataSize;
90 - (UInt32)getNumberOfSamples;
94 @implementation VLCDecompressedAudioOutput : QTCaptureDecompressedAudioOutput
95 - (id)initWithDemux:(demux_t *)p_demux
97 if( self = [super init] )
100 currentAudioBuffer = nil;
101 date_Init(&date, 44100, 1);
113 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection
115 AudioBufferList *tempAudioBufferList;
116 UInt32 totalDataSize = 0;
121 numberOfSamples = [sampleBuffer numberOfSamples];
122 date_Increment(&date,numberOfSamples);
123 currentPts = date_Get(&date);
125 tempAudioBufferList = [sampleBuffer audioBufferListWithOptions:0];
126 if (tempAudioBufferList->mNumberBuffers == 2)
129 * Compute totalDataSize as sum of all data blocks in the
132 for ( count = 0; count < tempAudioBufferList->mNumberBuffers; count++ )
134 totalDataSize += tempAudioBufferList->mBuffers[count].mDataByteSize;
137 * Allocate storage for the interleaved audio data
139 rawAudioData = block_Alloc(totalDataSize * sizeof(float));
140 if (NULL == rawAudioData)
142 msg_Err( p_qtsound, "Raw audiodata could not be allocated" );
148 msg_Err( p_qtsound, "Too many or only one channel found." );
153 * Interleave raw data (provided in two separate channels as
154 * F32L) with 2 samples per frame
159 const float *b1Ptr, *b2Ptr;
163 uPtr = (float *)rawAudioData,
164 b1Ptr = (const float *) tempAudioBufferList->mBuffers[0].mData,
165 b2Ptr = (const float *) tempAudioBufferList->mBuffers[1].mData;
166 i < numberOfSamples; i++)
172 if (currentAudioBuffer == nil)
174 currentAudioBuffer = (AudioBuffer *)malloc(sizeof(AudioBuffer));
175 if (NULL == currentAudioBuffer)
177 msg_Err( p_qtsound, "AudioBuffer could not be allocated." );
181 currentAudioBuffer->mNumberChannels = 2;
182 currentAudioBuffer->mDataByteSize = totalDataSize;
183 currentAudioBuffer->mData = rawAudioData;
188 - (BOOL)checkCurrentAudioBuffer
190 return (currentAudioBuffer) ? 1 : 0;
203 - (mtime_t)getCurrentPts
205 /* FIXME: can this getter be minimized? */
208 if( !currentAudioBuffer || currentPts == previousPts )
215 pts = previousPts = currentPts;
218 return (currentAudioBuffer->mData) ? currentPts : 0;
221 - (void *)getCurrentAudioBufferData
223 return currentAudioBuffer->mData;
226 - (UInt32)getCurrentTotalDataSize
228 return currentAudioBuffer->mDataByteSize;
231 - (UInt32)getNumberOfSamples
233 return numberOfSamples;
238 /*****************************************************************************
240 *****************************************************************************/
243 QTCaptureSession * session;
244 QTCaptureDevice * audiodevice;
245 VLCDecompressedAudioOutput * audiooutput;
246 es_out_id_t *p_es_audio;
247 int i_audio_max_buffer_size;
250 /*****************************************************************************
251 * Open: initialize interface
252 *****************************************************************************/
253 static int Open( vlc_object_t *p_this )
255 demux_t *p_demux = (demux_t*)p_this;
257 es_format_t audiofmt;
258 char *psz_uid = NULL;
261 NSString *qtk_curraudiodevice_uid;
262 NSAutoreleasePool *pool;
263 NSArray *myAudioDevices, *audioformat_array;
264 QTFormatDescription *audio_format;
265 QTCaptureDeviceInput *audioInput;
266 NSError *o_returnedAudioError;
268 if( p_demux->psz_location && *p_demux->psz_location )
270 psz_uid = p_demux->psz_location;
272 msg_Dbg( p_demux, "qtsound uid = %s", psz_uid );
273 qtk_curraudiodevice_uid = [[NSString alloc] initWithFormat:@"%s", psz_uid];
275 pool = [[NSAutoreleasePool alloc] init];
277 p_demux->p_sys = p_sys = calloc( 1, sizeof( demux_sys_t ) );
281 msg_Dbg( p_demux, "qtsound : uid = %s", [qtk_curraudiodevice_uid UTF8String]);
282 myAudioDevices = [[[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeSound] arrayByAddingObjectsFromArray:[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeMuxed]] retain];
283 if([myAudioDevices count] == 0)
285 dialog_FatalWait( p_demux, _("No Audio Input device found"),
286 _("Your Mac does not seem to be equipped with a suitable audio input device."
287 "Please check your connectors and drivers.") );
288 msg_Err( p_demux, "Can't find any Audio device" );
293 for(iaudio = 0; iaudio < [myAudioDevices count]; iaudio++){
294 QTCaptureDevice *qtk_audioDevice;
295 qtk_audioDevice = [myAudioDevices objectAtIndex:iaudio];
296 msg_Dbg( p_demux, "qtsound audio %u/%lu localizedDisplayName: %s uniqueID: %s", iaudio, [myAudioDevices count], [[qtk_audioDevice localizedDisplayName] UTF8String], [[qtk_audioDevice uniqueID] UTF8String]);
297 if([[[qtk_audioDevice uniqueID]stringByTrimmingCharactersInSet:[NSCharacterSet whitespaceCharacterSet]] isEqualToString:qtk_curraudiodevice_uid]){
298 msg_Dbg( p_demux, "Device found" );
304 if(iaudio < [myAudioDevices count]){
305 p_sys->audiodevice = [myAudioDevices objectAtIndex:iaudio];
309 /* cannot find designated audio device, fall back to open default audio device */
310 msg_Dbg(p_demux, "Cannot find designated uid audio device as %s. Fall back to open default audio device.", [qtk_curraudiodevice_uid UTF8String]);
311 p_sys->audiodevice = [QTCaptureDevice defaultInputDeviceWithMediaType: QTMediaTypeSound];
313 if( !p_sys->audiodevice )
315 dialog_FatalWait( p_demux, _("No audio input device found"),
316 _("Your Mac does not seem to be equipped with a suitable audio input device."
317 "Please check your connectors and drivers.") );
318 msg_Err( p_demux, "Can't find any Audio device" );
323 if( ![p_sys->audiodevice open: &o_returnedAudioError] )
325 msg_Err( p_demux, "Unable to open the audio capture device (%ld)", [o_returnedAudioError code] );
329 if( [p_sys->audiodevice isInUseByAnotherApplication] == YES )
331 msg_Err( p_demux, "default audio capture device is exclusively in use by another application" );
334 audioInput = [[QTCaptureDeviceInput alloc] initWithDevice: p_sys->audiodevice];
337 msg_Err( p_demux, "can't create a valid audio capture input facility" );
340 msg_Dbg( p_demux, "created valid audio capture input facility" );
343 p_sys->audiooutput = [[VLCDecompressedAudioOutput alloc] initWithDemux:p_demux];
344 msg_Dbg ( p_demux, "initialized audio output" );
346 /* Get the formats */
348 FIXME: the format description gathered here does not seem to be the same
349 in comparison to the format description collected from the actual sampleBuffer.
350 This information needs to be updated some other place. For the time being this shall suffice.
352 The following verbose output is an example of what is read from the input device during the below block
353 [0x3042138] qtsound demux debug: Audio localized format summary: Linear PCM, 24 bit little-endian signed integer, 2 channels, 44100 Hz
354 [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
355 [0x3042138] qtsound demux debug: Flag float 0 bigEndian 0 signedInt 1 packed 0 alignedHigh 0 non interleaved 0 non mixable 0
356 canonical 0 nativeFloatPacked 0 nativeEndian 0
358 However when reading this information from the sampleBuffer during the delegate call from
359 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
360 the following data shows up
361 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
362 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
363 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
364 canonical 1 nativeFloatPacked 1 nativeEndian 0
368 little-endian signed integer vs. little-endian floating point
369 format flag 00000004 vs. 00000029
370 bytes per packet 8 vs. 4
372 non interleaved 0 vs. 1 -> this makes a major difference when filling our own buffer
374 nativeFloatPacked 0 vs. 1
376 One would assume we'd need to feed the (es_format_t)audiofmt with the data collected here.
377 This is not the case. Audio will be transmitted in artefacts, due to wrong information.
379 At the moment this data is set manually, however one should consider trying to set this data dynamically
381 audioformat_array = [p_sys->audiodevice formatDescriptions];
383 for( int k = 0; k < [audioformat_array count]; k++ )
385 audio_format = (QTFormatDescription *)[audioformat_array objectAtIndex: k];
387 msg_Dbg( p_demux, "Audio localized format summary: %s", [[audio_format localizedFormatSummary] UTF8String]);
388 msg_Dbg( p_demux, "Audio format description attributes: %s",[[[audio_format formatDescriptionAttributes] description] UTF8String]);
390 AudioStreamBasicDescription asbd = {0};
391 NSValue *asbdValue = [audio_format attributeForKey:QTFormatDescriptionAudioStreamBasicDescriptionAttribute];
392 [asbdValue getValue:&asbd];
394 char formatIDString[5];
395 UInt32 formatID = CFSwapInt32HostToBig (asbd.mFormatID);
396 bcopy (&formatID, formatIDString, 4);
397 formatIDString[4] = '\0';
399 /* kept for development purposes */
401 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",
405 asbd.mBytesPerPacket,
406 asbd.mFramesPerPacket,
408 asbd.mChannelsPerFrame,
409 asbd.mBitsPerChannel);
411 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",
412 (asbd.mFormatFlags & kAudioFormatFlagIsFloat) != 0,
413 (asbd.mFormatFlags & kAudioFormatFlagIsBigEndian) != 0,
414 (asbd.mFormatFlags & kAudioFormatFlagIsSignedInteger) != 0,
415 (asbd.mFormatFlags & kAudioFormatFlagIsPacked) != 0,
416 (asbd.mFormatFlags & kAudioFormatFlagIsAlignedHigh) != 0,
417 (asbd.mFormatFlags & kAudioFormatFlagIsNonInterleaved) != 0,
418 (asbd.mFormatFlags & kAudioFormatFlagIsNonMixable) != 0,
420 (asbd.mFormatFlags & kAudioFormatFlagsCanonical) != 0,
421 (asbd.mFormatFlags & kAudioFormatFlagsNativeFloatPacked) != 0,
422 (asbd.mFormatFlags & kAudioFormatFlagsNativeEndian) != 0
427 if( [audioformat_array count] )
428 audio_format = [audioformat_array objectAtIndex: 0];
431 /* Now we can init */
432 audiocodec = VLC_CODEC_FL32;
433 es_format_Init( &audiofmt, AUDIO_ES, audiocodec);
435 audiofmt.audio.i_format = audiocodec;
436 audiofmt.audio.i_rate = 44100;
438 * i_physical_channels Describes the channels configuration of the
439 * samples (ie. number of channels which are available in the
440 * buffer, and positions).
442 audiofmt.audio.i_physical_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
444 * i_original_channels Describes from which original channels,
445 * before downmixing, the buffer is derived.
447 audiofmt.audio.i_original_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
449 * Please note that it may be completely arbitrary - buffers are not
450 * obliged to contain a integral number of so-called "frames". It's
451 * just here for the division:
452 * buffer_size = i_nb_samples * i_bytes_per_frame / i_frame_length
454 audiofmt.audio.i_bitspersample = 32;
455 audiofmt.audio.i_channels = 2;
456 audiofmt.audio.i_blockalign = audiofmt.audio.i_channels * (audiofmt.audio.i_bitspersample / 8);
457 audiofmt.i_bitrate = audiofmt.audio.i_channels * audiofmt.audio.i_rate * audiofmt.audio.i_bitspersample;
458 p_sys->i_audio_max_buffer_size = audiofmt.i_bitrate;
460 p_sys->session = [[QTCaptureSession alloc] init];
462 success = [p_sys->session addInput:audioInput error: &o_returnedAudioError];
465 msg_Err( p_demux, "the audio capture device could not be added to capture session (%ld)", [o_returnedAudioError code] );
469 success = [p_sys->session addOutput:p_sys->audiooutput error: &o_returnedAudioError];
472 msg_Err( p_demux, "audio output could not be added to capture session (%ld)", [o_returnedAudioError code] );
476 [p_sys->session startRunning];
479 p_demux->pf_demux = Demux;
480 p_demux->pf_control = Control;
481 p_demux->info.i_update = 0;
482 p_demux->info.i_title = 0;
483 p_demux->info.i_seekpoint = 0;
485 msg_Dbg( p_demux, "New audio es %d channels %dHz",
486 audiofmt.audio.i_channels, audiofmt.audio.i_rate );
488 p_sys->p_es_audio = es_out_Add( p_demux->out, &audiofmt );
490 [audioInput release];
493 msg_Dbg( p_demux, "QTSound: We have an audio device ready!" );
497 [audioInput release];
505 /*****************************************************************************
506 * Close: destroy interface
507 *****************************************************************************/
508 static void Close( vlc_object_t *p_this )
510 NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
511 demux_t *p_demux = (demux_t*)p_this;
512 demux_sys_t *p_sys = p_demux->p_sys;
514 [p_sys->session performSelectorOnMainThread:@selector(stopRunning) withObject:nil waitUntilDone:NO];
515 [p_sys->audiooutput performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
516 [p_sys->session performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
523 /*****************************************************************************
525 *****************************************************************************/
526 static int Demux( demux_t *p_demux )
528 demux_sys_t *p_sys = p_demux->p_sys;
530 NSAutoreleasePool *pool;
532 p_blocka = block_New( p_demux, p_sys->i_audio_max_buffer_size );
536 msg_Err( p_demux, "cannot get audio block" );
540 @synchronized (p_sys->audiooutput)
542 if ( [p_sys->audiooutput checkCurrentAudioBuffer] )
544 p_blocka->i_buffer = p_blocka->i_size = [p_sys->audiooutput getCurrentTotalDataSize];
545 p_blocka->p_buffer = p_blocka->p_start = [p_sys->audiooutput getCurrentAudioBufferData];
546 p_blocka->i_nb_samples = [p_sys->audiooutput getNumberOfSamples];
547 p_blocka->i_pts = [p_sys->audiooutput getCurrentPts];
551 if( !p_blocka->i_pts )
553 // Nothing to transfer yet, just forget
554 block_Release( p_blocka );
561 es_out_Control( p_demux->out, ES_OUT_SET_PCR, p_blocka->i_pts );
562 es_out_Send( p_demux->out, p_sys->p_es_audio, p_blocka );
565 @synchronized (p_sys->audiooutput)
567 [p_sys->audiooutput freeAudioMem];
573 /*****************************************************************************
575 *****************************************************************************/
576 static int Control( demux_t *p_demux, int i_query, va_list args )
583 /* Special for access_demux */
584 case DEMUX_CAN_PAUSE:
586 case DEMUX_SET_PAUSE_STATE:
587 case DEMUX_CAN_CONTROL_PACE:
588 pb = (bool*)va_arg( args, bool * );
592 case DEMUX_GET_PTS_DELAY:
593 pi64 = (int64_t*)va_arg( args, int64_t * );
594 *pi64 = INT64_C(1000) * var_InheritInteger( p_demux, "live-caching" );