summaryrefslogtreecommitdiff
path: root/lib/igt_audio.c
blob: 5b0860e6cc76a4ade1e09e804557c85e166a64c3 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
/*
 * Copyright © 2017 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 *
 * Authors:
 *  Paul Kocialkowski <paul.kocialkowski@linux.intel.com>
 */

#include "config.h"

#include <errno.h>
#include <fcntl.h>
#include <gsl/gsl_fft_real.h>
#include <math.h>
#include <unistd.h>

#include "igt_audio.h"
#include "igt_core.h"

#define FREQS_MAX 64

/**
 * SECTION:igt_audio
 * @short_description: Library for audio-related tests
 * @title: Audio
 * @include: igt_audio.h
 *
 * This library contains helpers for audio-related tests. More specifically,
 * it allows generating additions of sine signals as well as detecting them.
 */

struct audio_signal_freq {
	int freq;
	int channel;

	int16_t *period;
	size_t period_len;
	int offset;
};

struct audio_signal {
	int channels;
	int sampling_rate;

	struct audio_signal_freq freqs[FREQS_MAX];
	size_t freqs_count;
};

/**
 * audio_signal_init:
 * @channels: The number of channels to use for the signal
 * @sampling_rate: The sampling rate to use for the signal
 *
 * Allocate and initialize an audio signal structure with the given parameters.
 *
 * Returns: A newly-allocated audio signal structure
 */
struct audio_signal *audio_signal_init(int channels, int sampling_rate)
{
	struct audio_signal *signal;

	signal = malloc(sizeof(struct audio_signal));
	memset(signal, 0, sizeof(struct audio_signal));

	signal->sampling_rate = sampling_rate;
	signal->channels = channels;

	return signal;
}

/**
 * audio_signal_add_frequency:
 * @signal: The target signal structure
 * @frequency: The frequency to add to the signal
 * @channel: The channel to add this frequency to, or -1 to add it to all
 * channels
 *
 * Add a frequency to the signal.
 *
 * Returns: An integer equal to zero for success and negative for failure
 */
int audio_signal_add_frequency(struct audio_signal *signal, int frequency,
				int channel)
{
	size_t index = signal->freqs_count;
	struct audio_signal_freq *freq;

	igt_assert(index < FREQS_MAX);
	igt_assert(channel < signal->channels);

	/* Stay within the Nyquist–Shannon sampling theorem. */
	if (frequency > signal->sampling_rate / 2) {
		igt_debug("Skipping frequency %d: too high for a %d Hz "
			  "sampling rate\n", frequency, signal->sampling_rate);
		return -1;
	}

	/* Clip the frequency to an integer multiple of the sampling rate.
	 * This to be able to store a full period of it and use that for
	 * signal generation, instead of recurrent calls to sin().
	 */
	frequency = signal->sampling_rate / (signal->sampling_rate / frequency);

	igt_debug("Adding test frequency %d to channel %d\n",
		  frequency, channel);

	freq = &signal->freqs[index];
	memset(freq, 0, sizeof(*freq));
	freq->freq = frequency;
	freq->channel = channel;

	signal->freqs_count++;

	return 0;
}

/**
 * audio_signal_synthesize:
 * @signal: The target signal structure
 *
 * Synthesize the data tables for the audio signal, that can later be used
 * to fill audio buffers. The resources allocated by this function must be
 * freed with a call to audio_signal_clean when the signal is no longer used.
 */
void audio_signal_synthesize(struct audio_signal *signal)
{
	int16_t *period;
	double value;
	size_t period_len;
	int freq;
	int i, j;

	for (i = 0; i < signal->freqs_count; i++) {
		freq = signal->freqs[i].freq;
		period_len = signal->sampling_rate / freq;

		period = calloc(1, period_len * sizeof(int16_t));

		for (j = 0; j < period_len; j++) {
			value = 2.0 * M_PI * freq / signal->sampling_rate * j;
			value = sin(value) * INT16_MAX / signal->freqs_count;

			period[j] = (int16_t) value;
		}

		signal->freqs[i].period = period;
		signal->freqs[i].period_len = period_len;
	}
}

/**
 * audio_signal_fini:
 *
 * Release the signal.
 */
void audio_signal_fini(struct audio_signal *signal)
{
	audio_signal_reset(signal);
	free(signal);
}

/**
 * audio_signal_reset:
 * @signal: The target signal structure
 *
 * Free the resources allocated by audio_signal_synthesize and remove
 * the previously-added frequencies.
 */
void audio_signal_reset(struct audio_signal *signal)
{
	size_t i;

	for (i = 0; i < signal->freqs_count; i++) {
		free(signal->freqs[i].period);
	}

	signal->freqs_count = 0;
}

/**
 * audio_signal_fill:
 * @signal: The target signal structure
 * @buffer: The target buffer to fill
 * @samples: The number of samples to fill
 *
 * Fill the requested number of samples to the target buffer with the audio
 * signal data (in interleaved S16_LE format), at the requested sampling rate
 * and number of channels.
 */
void audio_signal_fill(struct audio_signal *signal, int16_t *buffer,
		       size_t buffer_len)
{
	int16_t *destination, *source;
	struct audio_signal_freq *freq;
	int total;
	int count;
	int i, j, k;

	memset(buffer, 0, sizeof(int16_t) * signal->channels * buffer_len);

	for (i = 0; i < signal->freqs_count; i++) {
		freq = &signal->freqs[i];
		total = 0;

		igt_assert(freq->period);

		while (total < buffer_len) {
			source = freq->period + freq->offset;
			destination = buffer + total * signal->channels;

			count = freq->period_len - freq->offset;
			if (count > buffer_len - total)
				count = buffer_len - total;

			freq->offset += count;
			freq->offset %= freq->period_len;

			for (j = 0; j < count; j++) {
				for (k = 0; k < signal->channels; k++) {
					if (freq->channel >= 0 &&
					    freq->channel != k)
						continue;
					destination[j * signal->channels + k] += source[j];
				}
			}

			total += count;
		}
	}
}

/**
 * Checks that frequencies specified in signal, and only those, are included
 * in the input data.
 *
 * sampling_rate is given in Hz. data_len is the number of elements in data.
 */
bool audio_signal_detect(struct audio_signal *signal, int sampling_rate,
			 int channel, double *data, size_t data_len)
{
	size_t bin_power_len = data_len / 2 + 1;
	double bin_power[bin_power_len];
	bool detected[FREQS_MAX];
	int ret, freq_accuracy, freq, local_max_freq;
	double max, local_max, threshold;
	size_t i, j;
	bool above, success;

	/* Allowed error in Hz due to FFT step */
	freq_accuracy = sampling_rate / data_len;
	igt_debug("Allowed freq. error: %d Hz\n", freq_accuracy);

	ret = gsl_fft_real_radix2_transform(data, 1, data_len);
	igt_assert(ret == 0);

	/* Compute the power received by every bin of the FFT, and record the
	 * maximum power received as a way to normalize all the others.
	 *
	 * For i < data_len / 2, the real part of the i-th term is stored at
	 * data[i] and its imaginary part is stored at data[data_len - i].
	 * i = 0 and i = data_len / 2 are special cases, they are purely real
	 * so their imaginary part isn't stored.
	 *
	 * The power is encoded as the magnitude of the complex number and the
	 * phase is encoded as its angle.
	 */
	max = 0;
	bin_power[0] = data[0];
	for (i = 1; i < bin_power_len - 1; i++) {
		bin_power[i] = hypot(data[i], data[data_len - i]);
		if (bin_power[i] > max)
			max = bin_power[i];
	}
	bin_power[bin_power_len - 1] = data[data_len / 2];

	for (i = 0; i < signal->freqs_count; i++)
		detected[i] = false;

	/* Do a linear search through the FFT bins' power to find the the local
	 * maximums that exceed half of the absolute maximum that we previously
	 * calculated.
	 *
	 * Since the frequencies might not be perfectly aligned with the bins of
	 * the FFT, we need to find the local maximum across some consecutive
	 * bins. Once the power returns under the power threshold, we compare
	 * the frequency of the bin that received the maximum power to the
	 * expected frequencies. If found, we mark this frequency as such,
	 * otherwise we warn that an unexpected frequency was found.
	 */
	threshold = max / 2;
	success = true;
	above = false;
	local_max = 0;
	local_max_freq = -1;
	for (i = 0; i < bin_power_len; i++) {
		freq = sampling_rate * i / data_len;

		if (bin_power[i] > threshold)
			above = true;

		if (!above) {
			continue;
		}

		/* If we were above the threshold and we're not anymore, it's
		 * time to decide whether the peak frequency is correct or
		 * invalid. */
		if (bin_power[i] < threshold) {
			for (j = 0; j < signal->freqs_count; j++) {
				if (signal->freqs[j].channel >= 0 &&
				    signal->freqs[j].channel != channel)
					continue;

				if (signal->freqs[j].freq >
				    local_max_freq - freq_accuracy &&
				    signal->freqs[j].freq <
				    local_max_freq + freq_accuracy) {
					detected[j] = true;
					igt_debug("Frequency %d detected\n",
						  local_max_freq);
					break;
				}
			}

			/* We haven't generated this frequency, but we detected
			 * it. */
			if (j == signal->freqs_count) {
				igt_debug("Detected additional frequency: %d\n",
					  local_max_freq);
				success = false;
			}

			above = false;
			local_max = 0;
			local_max_freq = -1;
		}

		if (bin_power[i] > local_max) {
			local_max = bin_power[i];
			local_max_freq = freq;
		}
	}

	/* Check that all frequencies we generated have been detected. */
	for (i = 0; i < signal->freqs_count; i++) {
		if (signal->freqs[i].channel >= 0 &&
		    signal->freqs[i].channel != channel)
			continue;

		if (!detected[i]) {
			igt_debug("Missing frequency: %d\n",
				  signal->freqs[i].freq);
			success = false;
		}
	}

	return success;
}

/**
 * Extracts a single channel from a multi-channel S32_LE input buffer.
 */
size_t audio_extract_channel_s32_le(double *dst, size_t dst_cap,
				    int32_t *src, size_t src_len,
				    int n_channels, int channel)
{
	size_t dst_len, i;

	igt_assert(channel < n_channels);
	igt_assert(src_len % n_channels == 0);
	dst_len = src_len / n_channels;
	igt_assert(dst_len <= dst_cap);
	for (i = 0; i < dst_len; i++)
		dst[i] = (double) src[i * n_channels + channel];

	return dst_len;
}

#define RIFF_TAG "RIFF"
#define WAVE_TAG "WAVE"
#define FMT_TAG "fmt "
#define DATA_TAG "data"

static void
append_to_buffer(char *dst, size_t *i, const void *src, size_t src_size)
{
	memcpy(&dst[*i], src, src_size);
	*i += src_size;
}

/**
 * audio_create_wav_file_s32_le:
 * @qualifier: the basename of the file (the test name will be prepended, and
 * the file extension will be appended)
 * @sample_rate: the sample rate in Hz
 * @channels: the number of channels
 * @path: if non-NULL, will be set to a pointer to the new file path (the
 * caller is responsible for free-ing it)
 *
 * Creates a new WAV file.
 *
 * After calling this function, the caller is expected to write S32_LE PCM data
 * to the returned file descriptor.
 *
 * See http://www-mmsp.ece.mcgill.ca/Documents/AudioFormats/WAVE/WAVE.html for
 * a WAV file format specification.
 *
 * Returns: a file descriptor to the newly created file, or -1 on error.
 */
int audio_create_wav_file_s32_le(const char *qualifier, uint32_t sample_rate,
				 uint16_t channels, char **path)
{
	char _path[PATH_MAX];
	const char *test_name, *subtest_name;
	int fd;
	char header[44];
	size_t i = 0;
	uint32_t file_size, chunk_size, byte_rate;
	uint16_t format, block_align, bits_per_sample;

	test_name = igt_test_name();
	subtest_name = igt_subtest_name();

	igt_assert(igt_frame_dump_path);
	snprintf(_path, sizeof(_path), "%s/audio-%s-%s-%s.wav",
		 igt_frame_dump_path, test_name, subtest_name, qualifier);

	if (path)
		*path = strdup(_path);

	igt_debug("Dumping %s audio to %s\n", qualifier, _path);
	fd = open(_path, O_WRONLY | O_CREAT | O_TRUNC);
	if (fd < 0) {
		igt_warn("open failed: %s\n", strerror(errno));
		return -1;
	}

	/* File header */
	file_size = UINT32_MAX; /* unknown file size */
	append_to_buffer(header, &i, RIFF_TAG, strlen(RIFF_TAG));
	append_to_buffer(header, &i, &file_size, sizeof(file_size));
	append_to_buffer(header, &i, WAVE_TAG, strlen(WAVE_TAG));

	/* Format chunk */
	chunk_size = 16;
	format = 1; /* PCM */
	bits_per_sample = 32; /* S32_LE */
	byte_rate = sample_rate * channels * bits_per_sample / 8;
	block_align = channels * bits_per_sample / 8;
	append_to_buffer(header, &i, FMT_TAG, strlen(FMT_TAG));
	append_to_buffer(header, &i, &chunk_size, sizeof(chunk_size));
	append_to_buffer(header, &i, &format, sizeof(format));
	append_to_buffer(header, &i, &channels, sizeof(channels));
	append_to_buffer(header, &i, &sample_rate, sizeof(sample_rate));
	append_to_buffer(header, &i, &byte_rate, sizeof(byte_rate));
	append_to_buffer(header, &i, &block_align, sizeof(block_align));
	append_to_buffer(header, &i, &bits_per_sample, sizeof(bits_per_sample));

	/* Data chunk */
	chunk_size = UINT32_MAX; /* unknown chunk size */
	append_to_buffer(header, &i, DATA_TAG, strlen(DATA_TAG));
	append_to_buffer(header, &i, &chunk_size, sizeof(chunk_size));

	igt_assert(i == sizeof(header));

	if (write(fd, header, sizeof(header)) != sizeof(header)) {
		igt_warn("write failed: %s'n", strerror(errno));
		close(fd);
		return -1;
	}

	return fd;
}