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 UInt32 numberOfSamples;
82 - (id)initWithDemux:(demux_t *)p_demux;
83 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
84 - (BOOL)checkCurrentAudioBuffer;
85 - (mtime_t)getCurrentPts;
86 - (void *)getCurrentAudioBufferData;
87 - (UInt32)getCurrentTotalDataSize;
88 - (UInt32)getNumberOfSamples;
92 @implementation VLCDecompressedAudioOutput : QTCaptureDecompressedAudioOutput
93 - (id)initWithDemux:(demux_t *)p_demux
95 if( self = [super init] )
98 currentAudioBuffer = nil;
99 date_Init(&date, 44100, 1);
110 free(currentAudioBuffer);
111 currentAudioBuffer = nil;
116 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection
118 AudioBufferList *tempAudioBufferList;
119 block_t *rawAudioData;
120 UInt32 totalDataSize = 0;
125 numberOfSamples = [sampleBuffer numberOfSamples];
126 date_Increment(&date,numberOfSamples);
127 currentPts = date_Get(&date);
129 tempAudioBufferList = [sampleBuffer audioBufferListWithOptions:0];
130 if (tempAudioBufferList->mNumberBuffers == 2)
133 * Compute totalDataSize as sum of all data blocks in the
136 for ( count = 0; count < tempAudioBufferList->mNumberBuffers; count++ )
138 totalDataSize += tempAudioBufferList->mBuffers[count].mDataByteSize;
141 * Allocate storage for the interleaved audio data
143 rawAudioData = block_Alloc(totalDataSize * sizeof(float));
144 if (NULL == rawAudioData)
146 msg_Err( p_qtsound, "Raw audiodata could not be allocated" );
152 msg_Err( p_qtsound, "Too many or only one channel found." );
157 * Interleave raw data (provided in two separate channels as
158 * F32L) with 2 samples per frame
163 const float *b1Ptr, *b2Ptr;
167 uPtr = (float *)rawAudioData,
168 b1Ptr = (const float *) tempAudioBufferList->mBuffers[0].mData,
169 b2Ptr = (const float *) tempAudioBufferList->mBuffers[1].mData;
170 i < numberOfSamples; i++)
180 if (currentAudioBuffer == nil)
182 currentAudioBuffer = (AudioBuffer *)malloc(sizeof(AudioBuffer));
183 if (NULL == currentAudioBuffer)
185 msg_Err( p_qtsound, "AudioBuffer could not be allocated." );
189 currentAudioBuffer->mNumberChannels = 2;
190 currentAudioBuffer->mDataByteSize = totalDataSize;
191 currentAudioBuffer->mData = rawAudioData;
197 - (BOOL)checkCurrentAudioBuffer
199 return (currentAudioBuffer) ? 1 : 0;
202 - (mtime_t)getCurrentPts
204 /* FIXME: can this getter be minimized? */
207 if( !currentAudioBuffer || currentPts == previousPts )
214 pts = previousPts = currentPts;
217 return (currentAudioBuffer->mData) ? currentPts : 0;
220 - (void *)getCurrentAudioBufferData
222 return currentAudioBuffer->mData;
225 - (UInt32)getCurrentTotalDataSize
227 return currentAudioBuffer->mDataByteSize;
230 - (UInt32)getNumberOfSamples
232 return numberOfSamples;
237 /*****************************************************************************
239 *****************************************************************************/
242 QTCaptureSession * session;
243 QTCaptureDevice * audiodevice;
244 VLCDecompressedAudioOutput * audiooutput;
245 es_out_id_t *p_es_audio;
246 int i_audio_max_buffer_size;
249 /*****************************************************************************
250 * Open: initialize interface
251 *****************************************************************************/
252 static int Open( vlc_object_t *p_this )
254 demux_t *p_demux = (demux_t*)p_this;
256 es_format_t audiofmt;
257 char *psz_uid = NULL;
260 NSString *qtk_curraudiodevice_uid;
261 NSAutoreleasePool *pool;
262 NSArray *myAudioDevices, *audioformat_array;
263 QTFormatDescription *audio_format;
264 QTCaptureDeviceInput *audioInput;
265 NSError *o_returnedAudioError;
267 if( p_demux->psz_location && *p_demux->psz_location )
269 psz_uid = p_demux->psz_location;
271 msg_Dbg( p_demux, "qtsound uid = %s", psz_uid );
272 qtk_curraudiodevice_uid = [[NSString alloc] initWithFormat:@"%s", psz_uid];
274 pool = [[NSAutoreleasePool alloc] init];
276 p_demux->p_sys = p_sys = calloc( 1, sizeof( demux_sys_t ) );
280 msg_Dbg( p_demux, "qtsound : uid = %s", [qtk_curraudiodevice_uid UTF8String]);
281 myAudioDevices = [[[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeSound] arrayByAddingObjectsFromArray:[QTCaptureDevice inputDevicesWithMediaType:QTMediaTypeMuxed]] retain];
282 if([myAudioDevices count] == 0)
284 dialog_FatalWait( p_demux, _("No Audio Input device found"),
285 _("Your Mac does not seem to be equipped with a suitable audio input device."
286 "Please check your connectors and drivers.") );
287 msg_Err( p_demux, "Can't find any Audio device" );
292 for(iaudio = 0; iaudio < [myAudioDevices count]; iaudio++){
293 QTCaptureDevice *qtk_audioDevice;
294 qtk_audioDevice = [myAudioDevices objectAtIndex:iaudio];
295 msg_Dbg( p_demux, "qtsound audio %u/%lu localizedDisplayName: %s uniqueID: %s", iaudio, [myAudioDevices count], [[qtk_audioDevice localizedDisplayName] UTF8String], [[qtk_audioDevice uniqueID] UTF8String]);
296 if([[[qtk_audioDevice localizedDisplayName]stringByTrimmingCharactersInSet:[NSCharacterSet whitespaceCharacterSet]] isEqualToString:qtk_curraudiodevice_uid]){
297 msg_Dbg( p_demux, "Device found" );
303 if(iaudio < [myAudioDevices count]){
304 p_sys->audiodevice = [myAudioDevices objectAtIndex:iaudio];
308 /* cannot find designated audio device, fall back to open default audio device */
309 msg_Dbg(p_demux, "Cannot find designated uid audio device as %s. Fall back to open default audio device.", [qtk_curraudiodevice_uid UTF8String]);
310 p_sys->audiodevice = [QTCaptureDevice defaultInputDeviceWithMediaType: QTMediaTypeSound];
312 if( !p_sys->audiodevice )
314 dialog_FatalWait( p_demux, _("No audio input device found"),
315 _("Your Mac does not seem to be equipped with a suitable audio input device."
316 "Please check your connectors and drivers.") );
317 msg_Err( p_demux, "Can't find any Audio device" );
322 if( ![p_sys->audiodevice open: &o_returnedAudioError] )
324 msg_Err( p_demux, "Unable to open the audio capture device (%ld)", [o_returnedAudioError code] );
328 if( [p_sys->audiodevice isInUseByAnotherApplication] == YES )
330 msg_Err( p_demux, "default audio capture device is exclusively in use by another application" );
333 audioInput = [[QTCaptureDeviceInput alloc] initWithDevice: p_sys->audiodevice];
336 msg_Err( p_demux, "can't create a valid audio capture input facility" );
339 msg_Dbg( p_demux, "created valid audio capture input facility" );
342 p_sys->audiooutput = [[VLCDecompressedAudioOutput alloc] initWithDemux:p_demux];
343 msg_Dbg ( p_demux, "initialized audio output" );
345 /* Get the formats */
347 FIXME: the format description gathered here does not seem to be the same
348 in comparison to the format description collected from the actual sampleBuffer.
349 This information needs to be updated some other place. For the time being this shall suffice.
351 The following verbose output is an example of what is read from the input device during the below block
352 [0x3042138] qtsound demux debug: Audio localized format summary: Linear PCM, 24 bit little-endian signed integer, 2 channels, 44100 Hz
353 [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
354 [0x3042138] qtsound demux debug: Flag float 0 bigEndian 0 signedInt 1 packed 0 alignedHigh 0 non interleaved 0 non mixable 0
355 canonical 0 nativeFloatPacked 0 nativeEndian 0
357 However when reading this information from the sampleBuffer during the delegate call from
358 - (void)outputAudioSampleBuffer:(QTSampleBuffer *)sampleBuffer fromConnection:(QTCaptureConnection *)connection;
359 the following data shows up
360 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
361 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
362 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
363 canonical 1 nativeFloatPacked 1 nativeEndian 0
367 little-endian signed integer vs. little-endian floating point
368 format flag 00000004 vs. 00000029
369 bytes per packet 8 vs. 4
371 non interleaved 0 vs. 1 -> this makes a major difference when filling our own buffer
373 nativeFloatPacked 0 vs. 1
375 One would assume we'd need to feed the (es_format_t)audiofmt with the data collected here.
376 This is not the case. Audio will be transmitted in artefacts, due to wrong information.
378 At the moment this data is set manually, however one should consider trying to set this data dynamically
380 audioformat_array = [p_sys->audiodevice formatDescriptions];
382 for( int k = 0; k < [audioformat_array count]; k++ )
384 audio_format = (QTFormatDescription *)[audioformat_array objectAtIndex: k];
386 msg_Dbg( p_demux, "Audio localized format summary: %s", [[audio_format localizedFormatSummary] UTF8String]);
387 msg_Dbg( p_demux, "Audio format description attributes: %s",[[[audio_format formatDescriptionAttributes] description] UTF8String]);
389 AudioStreamBasicDescription asbd = {0};
390 NSValue *asbdValue = [audio_format attributeForKey:QTFormatDescriptionAudioStreamBasicDescriptionAttribute];
391 [asbdValue getValue:&asbd];
393 char formatIDString[5];
394 UInt32 formatID = CFSwapInt32HostToBig (asbd.mFormatID);
395 bcopy (&formatID, formatIDString, 4);
396 formatIDString[4] = '\0';
398 /* kept for development purposes */
400 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",
404 asbd.mBytesPerPacket,
405 asbd.mFramesPerPacket,
407 asbd.mChannelsPerFrame,
408 asbd.mBitsPerChannel);
410 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",
411 (asbd.mFormatFlags & kAudioFormatFlagIsFloat) != 0,
412 (asbd.mFormatFlags & kAudioFormatFlagIsBigEndian) != 0,
413 (asbd.mFormatFlags & kAudioFormatFlagIsSignedInteger) != 0,
414 (asbd.mFormatFlags & kAudioFormatFlagIsPacked) != 0,
415 (asbd.mFormatFlags & kAudioFormatFlagIsAlignedHigh) != 0,
416 (asbd.mFormatFlags & kAudioFormatFlagIsNonInterleaved) != 0,
417 (asbd.mFormatFlags & kAudioFormatFlagIsNonMixable) != 0,
419 (asbd.mFormatFlags & kAudioFormatFlagsCanonical) != 0,
420 (asbd.mFormatFlags & kAudioFormatFlagsNativeFloatPacked) != 0,
421 (asbd.mFormatFlags & kAudioFormatFlagsNativeEndian) != 0
426 if( [audioformat_array count] )
427 audio_format = [audioformat_array objectAtIndex: 0];
430 /* Now we can init */
431 audiocodec = VLC_CODEC_FL32;
432 es_format_Init( &audiofmt, AUDIO_ES, audiocodec);
434 audiofmt.audio.i_format = audiocodec;
435 audiofmt.audio.i_rate = 44100;
437 * i_physical_channels Describes the channels configuration of the
438 * samples (ie. number of channels which are available in the
439 * buffer, and positions).
441 audiofmt.audio.i_physical_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
443 * i_original_channels Describes from which original channels,
444 * before downmixing, the buffer is derived.
446 audiofmt.audio.i_original_channels = AOUT_CHAN_RIGHT | AOUT_CHAN_LEFT;
448 * i_bytes_per_frame Optional - for A/52, SPDIF and DTS types:
449 * Bytes used by one compressed frame, depends on bitrate.
451 audiofmt.audio.i_bytes_per_frame = 4;
453 * Number of sampleframes contained in one compressed frame.
455 audiofmt.audio.i_frame_length = 1;
457 * Please note that it may be completely arbitrary - buffers are not
458 * obliged to contain a integral number of so-called "frames". It's
459 * just here for the division:
460 * buffer_size = i_nb_samples * i_bytes_per_frame / i_frame_length
462 audiofmt.audio.i_bitspersample = 32;
463 audiofmt.audio.i_channels = 2;
464 audiofmt.audio.i_blockalign = audiofmt.audio.i_channels * audiofmt.audio.i_bitspersample / 16;
465 audiofmt.i_bitrate = audiofmt.audio.i_channels * audiofmt.audio.i_rate * audiofmt.audio.i_bitspersample;
466 p_sys->i_audio_max_buffer_size = 4096;
468 p_sys->session = [[QTCaptureSession alloc] init];
470 success = [p_sys->session addInput:audioInput error: &o_returnedAudioError];
473 msg_Err( p_demux, "the audio capture device could not be added to capture session (%ld)", [o_returnedAudioError code] );
477 success = [p_sys->session addOutput:p_sys->audiooutput error: &o_returnedAudioError];
480 msg_Err( p_demux, "audio output could not be added to capture session (%ld)", [o_returnedAudioError code] );
484 [p_sys->session startRunning];
487 p_demux->pf_demux = Demux;
488 p_demux->pf_control = Control;
489 p_demux->info.i_update = 0;
490 p_demux->info.i_title = 0;
491 p_demux->info.i_seekpoint = 0;
493 msg_Dbg( p_demux, "New audio es %d channels %dHz",
494 audiofmt.audio.i_channels, audiofmt.audio.i_rate );
496 p_sys->p_es_audio = es_out_Add( p_demux->out, &audiofmt );
498 [audioInput release];
501 msg_Dbg( p_demux, "QTSound: We have an audio device ready!" );
505 [audioInput release];
513 /*****************************************************************************
514 * Close: destroy interface
515 *****************************************************************************/
516 static void Close( vlc_object_t *p_this )
518 NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
519 demux_t *p_demux = (demux_t*)p_this;
520 demux_sys_t *p_sys = p_demux->p_sys;
522 [p_sys->session performSelectorOnMainThread:@selector(stopRunning) withObject:nil waitUntilDone:NO];
523 [p_sys->audiooutput performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
524 [p_sys->session performSelectorOnMainThread:@selector(release) withObject:nil waitUntilDone:NO];
531 /*****************************************************************************
533 *****************************************************************************/
534 static int Demux( demux_t *p_demux )
536 demux_sys_t *p_sys = p_demux->p_sys;
538 NSAutoreleasePool *pool;
540 p_blocka = block_New( p_demux, p_sys->i_audio_max_buffer_size );
544 msg_Err( p_demux, "cannot get audio block" );
548 pool = [[NSAutoreleasePool alloc] init];
550 @synchronized (p_sys->audiooutput)
552 if ( [p_sys->audiooutput checkCurrentAudioBuffer] )
554 p_blocka->i_pts = [p_sys->audiooutput getCurrentPts];
555 p_blocka->p_buffer = [p_sys->audiooutput getCurrentAudioBufferData];
556 p_blocka->i_nb_samples = [p_sys->audiooutput getNumberOfSamples];
557 p_blocka->i_buffer = [p_sys->audiooutput getCurrentTotalDataSize];
561 if( !p_blocka->i_pts )
563 // Nothing to transfer yet, just forget
564 block_Release( p_blocka );
574 es_out_Control( p_demux->out, ES_OUT_SET_PCR, p_blocka->i_pts );
575 es_out_Send( p_demux->out, p_sys->p_es_audio, p_blocka );
581 /*****************************************************************************
583 *****************************************************************************/
584 static int Control( demux_t *p_demux, int i_query, va_list args )
591 /* Special for access_demux */
592 case DEMUX_CAN_PAUSE:
594 case DEMUX_SET_PAUSE_STATE:
595 case DEMUX_CAN_CONTROL_PACE:
596 pb = (bool*)va_arg( args, bool * );
600 case DEMUX_GET_PTS_DELAY:
601 pi64 = (int64_t*)va_arg( args, int64_t * );
602 *pi64 = INT64_C(1000) * var_InheritInteger( p_demux, "live-caching" );