visualization.h File Reference

Go to the source code of this file.

Macros
#define	AUDIO_VISUALIZATION_SAMPLE_RATE   48000
#define	AUDIO_VISUALIZATION_COLOR_STANDARD   0
#define	AUDIO_VISUALIZATION_COLOR_DARKER   1
#define	AUDIO_VISUALIZATION_COLOR_BRIGHTER   2

Enumerations
enum	audio_visualization_source_t {   AUDIO_VISUALIZATION_SOURCE_MIC = 0 , AUDIO_VISUALIZATION_SOURCE_MEDIA , AUDIO_VISUALIZATION_SOURCE_REMOTE , AUDIO_VISUALIZATION_SOURCE_LOCAL_MIX ,   AUDIO_VISUALIZATION_SOURCE_MIX , AUDIO_VISUALIZATION_SOURCE_COUNT }

Functions
void	audio_visualization_submit (audio_visualization_source_t source, const float *samples, size_t count)
void	audio_visualization_read (audio_visualization_source_t source, float *samples, size_t count)
char *	audio_visualization_render_waveform (unsigned int width, unsigned int height, audio_visualization_source_t source, bool use_color, int color_mode, bool flip_x, bool flip_y)
char *	audio_visualization_render_fft (unsigned int width, unsigned int height, audio_visualization_source_t source, bool use_color, int color_mode, bool flip_x, bool flip_y)

Macro Definition Documentation

AUDIO_VISUALIZATION_COLOR_BRIGHTER

#define AUDIO_VISUALIZATION_COLOR_BRIGHTER   2

Definition at line 20 of file visualization.h.

AUDIO_VISUALIZATION_COLOR_DARKER

#define AUDIO_VISUALIZATION_COLOR_DARKER   1

Definition at line 19 of file visualization.h.

AUDIO_VISUALIZATION_COLOR_STANDARD

#define AUDIO_VISUALIZATION_COLOR_STANDARD   0

Definition at line 18 of file visualization.h.

AUDIO_VISUALIZATION_SAMPLE_RATE

#define AUDIO_VISUALIZATION_SAMPLE_RATE   48000

Definition at line 16 of file visualization.h.

Enumeration Type Documentation

audio_visualization_source_t

enum audio_visualization_source_t

Enumerator
AUDIO_VISUALIZATION_SOURCE_MIC
AUDIO_VISUALIZATION_SOURCE_MEDIA
AUDIO_VISUALIZATION_SOURCE_REMOTE
AUDIO_VISUALIZATION_SOURCE_LOCAL_MIX
AUDIO_VISUALIZATION_SOURCE_MIX
AUDIO_VISUALIZATION_SOURCE_COUNT

Definition at line 7 of file visualization.h.

             {
  AUDIO_VISUALIZATION_SOURCE_MIC = 0,
  AUDIO_VISUALIZATION_SOURCE_MEDIA,
  AUDIO_VISUALIZATION_SOURCE_REMOTE,
  AUDIO_VISUALIZATION_SOURCE_LOCAL_MIX,
  AUDIO_VISUALIZATION_SOURCE_MIX,
  AUDIO_VISUALIZATION_SOURCE_COUNT
} audio_visualization_source_t;

Function Documentation

audio_visualization_read()

void audio_visualization_read	(	audio_visualization_source_t	source,
		float *	samples,
		size_t	count
	)

Definition at line 115 of file visualization.c.

                                                                                                 {
  if (!samples)
    return;
  if (source < AUDIO_VISUALIZATION_SOURCE_MIC || source >= AUDIO_VISUALIZATION_SOURCE_COUNT) {
    memset(samples, 0, count * sizeof(float));
    return;
  }
  static_mutex_lock(&history_mutex);
  size_t available_count = count < VISUALIZATION_HISTORY ? count : VISUALIZATION_HISTORY;
  size_t output_offset = count - available_count;
  for (size_t i = 0; i < count; i++)
    samples[i] = 0.0f;
  size_t first_source = source;
  size_t last_source = source;
  if (source == AUDIO_VISUALIZATION_SOURCE_MIX) {
    first_source = AUDIO_VISUALIZATION_SOURCE_MIC;
    last_source = AUDIO_VISUALIZATION_SOURCE_REMOTE;
  } else if (source == AUDIO_VISUALIZATION_SOURCE_LOCAL_MIX) {
    first_source = AUDIO_VISUALIZATION_SOURCE_MIC;
    last_source = AUDIO_VISUALIZATION_SOURCE_MEDIA;
  }
  for (size_t current = first_source; current <= last_source; current++) {
    size_t start = (write_position[current] + VISUALIZATION_HISTORY - available_count) % VISUALIZATION_HISTORY;
    for (size_t i = 0; i < available_count; i++)
      samples[output_offset + i] += history[current][(start + i) % VISUALIZATION_HISTORY];
  }
  static_mutex_unlock(&history_mutex);
  for (size_t i = 0; i < count; i++)
    samples[i] = fmaxf(-1.0f, fminf(1.0f, samples[i]));
}

References AUDIO_VISUALIZATION_SOURCE_COUNT, AUDIO_VISUALIZATION_SOURCE_LOCAL_MIX, AUDIO_VISUALIZATION_SOURCE_MEDIA, AUDIO_VISUALIZATION_SOURCE_MIC, AUDIO_VISUALIZATION_SOURCE_MIX, AUDIO_VISUALIZATION_SOURCE_REMOTE, and VISUALIZATION_HISTORY.

Referenced by audio_visualization_render_fft(), and audio_visualization_render_waveform().

audio_visualization_render_fft()

char * audio_visualization_render_fft	(	unsigned int	width,
		unsigned int	height,
		audio_visualization_source_t	source,
		bool	use_color,
		int	color_mode,
		bool	flip_x,
		bool	flip_y
	)

Definition at line 275 of file visualization.c.

                                                                                               {
  if (width < 16 || height < 6 || width > MAX_TERMINAL_WIDTH || height > MAX_TERMINAL_HEIGHT ||
      source < AUDIO_VISUALIZATION_SOURCE_MIC || source > AUDIO_VISUALIZATION_SOURCE_MIX)
    return NULL;
 
  float *samples = SAFE_MALLOC(VISUALIZATION_HISTORY * sizeof(float), float *);
  if (!samples)
    return NULL;
  audio_visualization_read(source, samples, VISUALIZATION_HISTORY);
  size_t cell_count = (size_t)width * height;
  char *cells = SAFE_MALLOC(cell_count, char *);
  unsigned char *colors = SAFE_MALLOC(cell_count * 3, unsigned char *);
  if (!cells || !colors) {
    SAFE_FREE(colors);
    SAFE_FREE(cells);
    SAFE_FREE(samples);
    return NULL;
  }
  memset(cells, ' ', cell_count);
  memset(colors, 0, cell_count * 3);
 
  /* Each column is a 21 ms Hann-windowed spectrum; adjacent columns overlap heavily. */
  float real[FFT_SIZE], imaginary[FFT_SIZE];
  static const char levels[] = " .:-=+*#%@";
  const unsigned int spectrum_height = height;
  for (unsigned int x = 0; x < width; x++) {
    size_t center = (size_t)x * (VISUALIZATION_HISTORY - 1) / (width - 1);
    size_t start = center > FFT_SIZE / 2 ? center - FFT_SIZE / 2 : 0;
    if (start + FFT_SIZE > VISUALIZATION_HISTORY)
      start = VISUALIZATION_HISTORY - FFT_SIZE;
    for (size_t i = 0; i < FFT_SIZE; i++) {
      float window = 0.5f - 0.5f * cosf(6.28318530718f * (float)i / (float)(FFT_SIZE - 1));
      real[i] = samples[start + i] * window;
      imaginary[i] = 0.0f;
    }
    visualization_fft(real, imaginary);
 
    for (unsigned int row = 0; row < spectrum_height; row++) {
      float top_fraction = (float)(spectrum_height - row - 1) / (float)spectrum_height;
      float bottom_fraction = (float)(spectrum_height - row) / (float)spectrum_height;
      float low_hz = 35.0f * powf(18000.0f / 35.0f, top_fraction);
      float high_hz = 35.0f * powf(18000.0f / 35.0f, bottom_fraction);
      size_t low_bin = (size_t)fmaxf(1.0f, floorf(low_hz * FFT_SIZE / AUDIO_VISUALIZATION_SAMPLE_RATE));
      size_t high_bin =
          (size_t)fmaxf((float)(low_bin + 1), ceilf(high_hz * FFT_SIZE / AUDIO_VISUALIZATION_SAMPLE_RATE));
      if (high_bin > FFT_SIZE / 2)
        high_bin = FFT_SIZE / 2;
      float power = 0.0f;
      size_t bins = 0;
      for (size_t bin = low_bin; bin < high_bin; bin++) {
        float magnitude = hypotf(real[bin], imaginary[bin]) / (FFT_SIZE * 0.25f);
        power += magnitude * magnitude;
        bins++;
      }
      float magnitude = bins ? sqrtf(power / (float)bins) : 0.0f;
      /* dB scaling leaves room for quiet speech while keeping loud music detailed. */
      float db = 20.0f * log10f(fmaxf(magnitude, 0.00001f));
      float intensity = fmaxf(0.0f, fminf(1.0f, (db + 58.0f) / 52.0f));
      int level = (int)lrintf(intensity * (float)(sizeof(levels) - 2));
      unsigned int source_x = flip_x ? width - x - 1 : x;
      unsigned int source_row = flip_y ? height - row - 1 : row;
      size_t cell = (size_t)source_row * width + source_x;
      cells[cell] = levels[level];
      float band_hz = sqrtf(low_hz * high_hz);
      visualization_color_for_level(band_hz, intensity, color_mode, &colors[cell * 3], &colors[cell * 3 + 1],
                                    &colors[cell * 3 + 2]);
    }
  }
 
  size_t capacity = cell_count * (use_color ? 24U : 1U) + (size_t)height + 1;
  char *frame = SAFE_MALLOC(capacity, char *);
  if (!frame) {
    SAFE_FREE(colors);
    SAFE_FREE(cells);
    SAFE_FREE(samples);
    return NULL;
  }
  size_t used = 0;
  for (unsigned int row = 0; row < height; row++) {
    for (unsigned int x = 0; x < width; x++) {
      size_t cell = (size_t)row * width + x;
      char ch = cells[cell];
      if (use_color && ch != ' ') {
        used += (size_t)snprintf(frame + used, capacity - used, "\033[38;2;%u;%u;%um%c\033[0m", colors[cell * 3],
                                 colors[cell * 3 + 1], colors[cell * 3 + 2], ch);
      } else {
        frame[used++] = ch;
      }
    }
    frame[used++] = '\n';
  }
  frame[used] = '\0';
  SAFE_FREE(colors);
  SAFE_FREE(cells);
  SAFE_FREE(samples);
  return frame;
}

References audio_visualization_read(), AUDIO_VISUALIZATION_SAMPLE_RATE, AUDIO_VISUALIZATION_SOURCE_MIX, FFT_SIZE, frame, MAX_TERMINAL_HEIGHT, MAX_TERMINAL_WIDTH, SAFE_FREE, SAFE_MALLOC, and VISUALIZATION_HISTORY.

Referenced by mirror_render_audio_visualization().

audio_visualization_render_waveform()

char * audio_visualization_render_waveform	(	unsigned int	width,
		unsigned int	height,
		audio_visualization_source_t	source,
		bool	use_color,
		int	color_mode,
		bool	flip_x,
		bool	flip_y
	)

Definition at line 146 of file visualization.c.

                                                                                                    {
  if (width < 8 || height < 4 || width > MAX_TERMINAL_WIDTH || height > MAX_TERMINAL_HEIGHT ||
      source < AUDIO_VISUALIZATION_SOURCE_MIC || source > AUDIO_VISUALIZATION_SOURCE_MIX)
    return NULL;
 
  /* Keep a fixed one-second view so new audio enters on the right smoothly. */
  const size_t sample_count = VISUALIZATION_HISTORY;
  float *samples = SAFE_MALLOC(sample_count * sizeof(float), float *);
  if (!samples)
    return NULL;
  audio_visualization_read(source, samples, sample_count);
 
  const unsigned int grid_height = height;
  const unsigned int center = grid_height / 2;
  const unsigned int half_height = grid_height - 1 - center;
  const size_t cell_count = (size_t)width * grid_height;
  char *cells = SAFE_MALLOC(cell_count, char *);
  if (!cells) {
    SAFE_FREE(samples);
    return NULL;
  }
  memset(cells, ' ', cell_count);
  unsigned char palette[MAX_TERMINAL_WIDTH][3] = {{0}};
 
  float positive_peak = 0.0f;
  float negative_peak = 0.0f;
  size_t x = 0;
 
  for (size_t i = 0; i < sample_count; i++) {
    float sample = samples[i];
    positive_peak = fmaxf(positive_peak, sample);
    negative_peak = fmaxf(negative_peak, -sample);
 
    size_t next_bucket = ((i + 1) * width) / sample_count;
    if (next_bucket != x || i + 1 == sample_count) {
      size_t begin = x * sample_count / width;
      size_t end = (x + 1) * sample_count / width;
      size_t bucket_center = begin + (end - begin) / 2;
      float centroid = waveform_spectral_centroid(samples, sample_count, bucket_center);
      /* A gentle square-root scale keeps speech visible without flattening loud peaks. */
      float level = sqrtf(fminf(fmaxf(positive_peak, negative_peak), 1.0f));
      visualization_color_for_level(centroid, level, color_mode, &palette[x][0], &palette[x][1], &palette[x][2]);
      unsigned int positive_amplitude = (unsigned int)lrintf(sqrtf(fminf(positive_peak, 1.0f)) * (float)half_height);
      unsigned int negative_amplitude = (unsigned int)lrintf(sqrtf(fminf(negative_peak, 1.0f)) * (float)half_height);
      if (positive_amplitude > half_height)
        positive_amplitude = half_height;
      if (negative_amplitude > half_height)
        negative_amplitude = half_height;
      unsigned int top = center - positive_amplitude;
      unsigned int bottom = center + negative_amplitude;
      for (unsigned int y = top; y <= bottom; y++) {
        char ch = (y == top || y == bottom) ? '+' : '#';
        cells[(size_t)y * width + x] = ch;
      }
 
      /* Advance to the next time slice in the left-to-right history. */
      positive_peak = 0.0f;
      negative_peak = 0.0f;
      x = next_bucket;
    }
  }
 
  size_t capacity = cell_count * (use_color ? 24U : 1U) + (size_t)height + 1;
  char *frame = SAFE_MALLOC(capacity, char *);
  if (!frame) {
    SAFE_FREE(cells);
    SAFE_FREE(samples);
    return NULL;
  }
  size_t used = 0;
  for (unsigned int y = 0; y < grid_height; y++) {
    for (unsigned int x = 0; x < width; x++) {
      unsigned int source_x = flip_x ? width - x - 1 : x;
      unsigned int source_y = flip_y ? grid_height - y - 1 : y;
      char ch = cells[(size_t)source_y * width + source_x];
      if (use_color && ch != ' ') {
        used += (size_t)snprintf(frame + used, capacity - used, "\033[38;2;%u;%u;%um%c\033[0m", palette[source_x][0],
                                 palette[source_x][1], palette[source_x][2], ch);
      } else {
        frame[used++] = ch;
      }
    }
    frame[used++] = '\n';
  }
  frame[used] = '\0';
  SAFE_FREE(cells);
  SAFE_FREE(samples);
  return frame;
}

References audio_visualization_read(), AUDIO_VISUALIZATION_SOURCE_MIX, frame, MAX_TERMINAL_HEIGHT, MAX_TERMINAL_WIDTH, SAFE_FREE, SAFE_MALLOC, and VISUALIZATION_HISTORY.

Referenced by mirror_render_audio_visualization().

audio_visualization_submit()

void audio_visualization_submit	(	audio_visualization_source_t	source,
		const float *	samples,
		size_t	count
	)

Definition at line 100 of file visualization.c.

                                                                                                         {
  if ((!GET_OPTION(waveform) && !GET_OPTION(fft)) || !samples || source < AUDIO_VISUALIZATION_SOURCE_MIC ||
      source > AUDIO_VISUALIZATION_SOURCE_REMOTE)
    return;
  static_mutex_lock(&history_mutex);
  size_t position = write_position[source];
  for (size_t i = 0; i < count; i++) {
    float sample = samples[i];
    history[source][position] = isfinite(sample) ? fmaxf(-1.0f, fminf(1.0f, sample)) : 0.0f;
    position = (position + 1) % VISUALIZATION_HISTORY;
  }
  write_position[source] = position;
  static_mutex_unlock(&history_mutex);
}

References AUDIO_VISUALIZATION_SOURCE_MIC, AUDIO_VISUALIZATION_SOURCE_REMOTE, GET_OPTION, and VISUALIZATION_HISTORY.

Referenced by audio_write_samples(), and mirror_submit_audio_samples().