enum DisplayMode { COMBINED, SEPARATE, NB_MODES };
enum DataMode { D_MAGNITUDE, D_PHASE, NB_DMODES };
+enum FrequencyScale { F_LINEAR, F_LOG, NB_FSCALES };
enum DisplayScale { LINEAR, SQRT, CBRT, LOG, FOURTHRT, FIFTHRT, NB_SCALES };
enum ColorMode { CHANNEL, INTENSITY, RAINBOW, MORELAND, NEBULAE, FIRE, FIERY, FRUIT, COOL, MAGMA, GREEN, VIRIDIS, PLASMA, CIVIDIS, TERRAIN, NB_CLMODES };
enum SlideMode { REPLACE, SCROLL, FULLFRAME, RSCROLL, NB_SLIDES };
int mode; ///< channel display mode
int color_mode; ///< display color scheme
int scale;
+ int fscale;
float saturation; ///< color saturation multiplier
float rotation; ///< color rotation
int start, stop; ///< zoom mode
int single_pic;
int legend;
int start_x, start_y;
+ int (*plot_channel)(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs);
} ShowSpectrumContext;
#define OFFSET(x) offsetof(ShowSpectrumContext, x)
{ "log", "logarithmic", 0, AV_OPT_TYPE_CONST, {.i64=LOG}, 0, 0, FLAGS, "scale" },
{ "4thrt","4th root", 0, AV_OPT_TYPE_CONST, {.i64=FOURTHRT}, 0, 0, FLAGS, "scale" },
{ "5thrt","5th root", 0, AV_OPT_TYPE_CONST, {.i64=FIFTHRT}, 0, 0, FLAGS, "scale" },
+ { "fscale", "set frequency scale", OFFSET(fscale), AV_OPT_TYPE_INT, {.i64=F_LINEAR}, 0, NB_FSCALES-1, FLAGS, "fscale" },
+ { "lin", "linear", 0, AV_OPT_TYPE_CONST, {.i64=F_LINEAR}, 0, 0, FLAGS, "fscale" },
+ { "log", "logarithmic", 0, AV_OPT_TYPE_CONST, {.i64=F_LOG}, 0, 0, FLAGS, "fscale" },
{ "saturation", "color saturation multiplier", OFFSET(saturation), AV_OPT_TYPE_FLOAT, {.dbl = 1}, -10, 10, FLAGS },
{ "win_func", "set window function", OFFSET(win_func), AV_OPT_TYPE_INT, {.i64 = WFUNC_HANNING}, 0, NB_WFUNC-1, FLAGS, "win_func" },
{ "rect", "Rectangular", 0, AV_OPT_TYPE_CONST, {.i64=WFUNC_RECT}, 0, 0, FLAGS, "win_func" },
/* set input audio formats */
formats = ff_make_format_list(sample_fmts);
- if ((ret = ff_formats_ref(formats, &inlink->out_formats)) < 0)
+ if ((ret = ff_formats_ref(formats, &inlink->outcfg.formats)) < 0)
return ret;
layouts = ff_all_channel_layouts();
- if ((ret = ff_channel_layouts_ref(layouts, &inlink->out_channel_layouts)) < 0)
+ if ((ret = ff_channel_layouts_ref(layouts, &inlink->outcfg.channel_layouts)) < 0)
return ret;
formats = ff_all_samplerates();
- if ((ret = ff_formats_ref(formats, &inlink->out_samplerates)) < 0)
+ if ((ret = ff_formats_ref(formats, &inlink->outcfg.samplerates)) < 0)
return ret;
/* set output video format */
formats = ff_make_format_list(pix_fmts);
- if ((ret = ff_formats_ref(formats, &outlink->in_formats)) < 0)
+ if ((ret = ff_formats_ref(formats, &outlink->incfg.formats)) < 0)
return ret;
return 0;
}
if (s->stop) {
- double theta, phi, psi, a, b, S, c;
+ float theta, phi, psi, a, b, S, c;
FFTComplex *g = s->fft_data[ch];
FFTComplex *h = s->fft_scratch[ch];
int L = s->buf_size;
int N = s->win_size;
int M = s->win_size / 2;
- phi = 2.0 * M_PI * (s->stop - s->start) / (double)inlink->sample_rate / (M - 1);
- theta = 2.0 * M_PI * s->start / (double)inlink->sample_rate;
+ phi = 2.f * M_PI * (s->stop - s->start) / (float)inlink->sample_rate / (M - 1);
+ theta = 2.f * M_PI * s->start / (float)inlink->sample_rate;
for (int n = 0; n < M; n++) {
- h[n].re = cos(n * n / 2.0 * phi);
- h[n].im = sin(n * n / 2.0 * phi);
+ h[n].re = cosf(n * n / 2.f * phi);
+ h[n].im = sinf(n * n / 2.f * phi);
}
for (int n = M; n < L; n++) {
- h[n].re = 0.0;
- h[n].im = 0.0;
+ h[n].re = 0.f;
+ h[n].im = 0.f;
}
for (int n = L - N; n < L; n++) {
- h[n].re = cos((L - n) * (L - n) / 2.0 * phi);
- h[n].im = sin((L - n) * (L - n) / 2.0 * phi);
+ h[n].re = cosf((L - n) * (L - n) / 2.f * phi);
+ h[n].im = sinf((L - n) * (L - n) / 2.f * phi);
}
for (int n = 0; n < N; n++) {
}
for (int n = N; n < L; n++) {
- g[n].re = 0.;
- g[n].im = 0.;
+ g[n].re = 0.f;
+ g[n].im = 0.f;
}
for (int n = 0; n < N; n++) {
- psi = n * theta + n * n / 2.0 * phi;
- c = cos(psi);
- S = -sin(psi);
+ psi = n * theta + n * n / 2.f * phi;
+ c = cosf(psi);
+ S = -sinf(psi);
a = c * g[n].re - S * g[n].im;
b = S * g[n].re + c * g[n].im;
g[n].re = a;
av_fft_calc(s->ifft[ch], g);
for (int k = 0; k < M; k++) {
- psi = k * k / 2.0 * phi;
- c = cos(psi);
- S = -sin(psi);
+ psi = k * k / 2.f * phi;
+ c = cosf(psi);
+ S = -sinf(psi);
a = c * g[k].re - S * g[k].im;
b = S * g[k].re + c * g[k].im;
s->fft_data[ch][k].re = a;
if (s->color_mode == CHANNEL) {
if (s->nb_display_channels > 1) {
- *uf *= 0.5 * sin((2 * M_PI * ch) / s->nb_display_channels + M_PI * s->rotation);
- *vf *= 0.5 * cos((2 * M_PI * ch) / s->nb_display_channels + M_PI * s->rotation);
+ *uf *= 0.5f * sinf((2 * M_PI * ch) / s->nb_display_channels + M_PI * s->rotation);
+ *vf *= 0.5f * cosf((2 * M_PI * ch) / s->nb_display_channels + M_PI * s->rotation);
} else {
- *uf *= 0.5 * sin(M_PI * s->rotation);
- *vf *= 0.5 * cos(M_PI * s->rotation + M_PI_2);
+ *uf *= 0.5f * sinf(M_PI * s->rotation);
+ *vf *= 0.5f * cosf(M_PI * s->rotation + M_PI_2);
}
} else {
- *uf += *uf * sin(M_PI * s->rotation);
- *vf += *vf * cos(M_PI * s->rotation + M_PI_2);
+ *uf += *uf * sinf(M_PI * s->rotation);
+ *vf += *vf * cosf(M_PI * s->rotation + M_PI_2);
}
*uf *= s->saturation;
return units;
}
+static float log_scale(const float value, const float min, const float max)
+{
+ if (value < min)
+ return min;
+ if (value > max)
+ return max;
+
+ {
+ const float b = logf(max / min) / (max - min);
+ const float a = max / expf(max * b);
+
+ return expf(value * b) * a;
+ }
+}
+
+static float get_log_hz(const int bin, const int num_bins, const float sample_rate)
+{
+ const float max_freq = sample_rate / 2;
+ const float hz_per_bin = max_freq / num_bins;
+ const float freq = hz_per_bin * bin;
+ const float scaled_freq = log_scale(freq + 1, 21, max_freq) - 1;
+
+ return num_bins * scaled_freq / max_freq;
+}
+
+static float inv_log_scale(const float value, const float min, const float max)
+{
+ if (value < min)
+ return min;
+ if (value > max)
+ return max;
+
+ {
+ const float b = logf(max / min) / (max - min);
+ const float a = max / expf(max * b);
+
+ return logf(value / a) / b;
+ }
+}
+
+static float bin_pos(const int bin, const int num_bins, const float sample_rate)
+{
+ const float max_freq = sample_rate / 2;
+ const float hz_per_bin = max_freq / num_bins;
+ const float freq = hz_per_bin * bin;
+ const float scaled_freq = inv_log_scale(freq + 1, 21, max_freq) - 1;
+
+ return num_bins * scaled_freq / max_freq;
+}
+
static int draw_legend(AVFilterContext *ctx, int samples)
{
ShowSpectrumContext *s = ctx->priv;
inlink->channel_layout);
text = av_asprintf("%d Hz | %s", inlink->sample_rate, chlayout_str);
+ if (!text)
+ return AVERROR(ENOMEM);
drawtext(s->outpicref, 2, outlink->h - 10, "CREATED BY LIBAVFILTER", 0);
drawtext(s->outpicref, outlink->w - 2 - strlen(text) * 10, outlink->h - 10, text, 0);
+ av_freep(&text);
if (s->stop) {
- char *text = av_asprintf("Zoom: %d Hz - %d Hz", s->start, s->stop);
+ text = av_asprintf("Zoom: %d Hz - %d Hz", s->start, s->stop);
+ if (!text)
+ return AVERROR(ENOMEM);
drawtext(s->outpicref, outlink->w - 2 - strlen(text) * 10, 3, text, 0);
av_freep(&text);
}
- av_freep(&text);
-
dst = s->outpicref->data[0] + (s->start_y - 1) * s->outpicref->linesize[0] + s->start_x - 1;
for (x = 0; x < s->w + 1; x++)
dst[x] = 200;
}
for (y = 0; y < h; y += 40) {
float range = s->stop ? s->stop - s->start : inlink->sample_rate / 2;
- float hertz = s->start + y * range / (float)(1 << (int)ceil(log2(h)));
+ float bin = s->fscale == F_LINEAR ? y : get_log_hz(y, h, inlink->sample_rate);
+ float hertz = s->start + bin * range / (float)(1 << (int)ceil(log2(h)));
char *units;
if (hertz == 0)
for (x = 0; x < s->w && s->single_pic; x+=80) {
float seconds = x * spp / inlink->sample_rate;
char *units = get_time(ctx, seconds, x);
+ if (!units)
+ return AVERROR(ENOMEM);
drawtext(s->outpicref, s->start_x + x - 4 * strlen(units), s->h + s->start_y + 6, units, 0);
drawtext(s->outpicref, s->start_x + x - 4 * strlen(units), s->start_y - 12, units, 0);
}
for (x = 0; x < w - 79; x += 80) {
float range = s->stop ? s->stop - s->start : inlink->sample_rate / 2;
- float hertz = s->start + x * range / (float)(1 << (int)ceil(log2(w)));
+ float bin = s->fscale == F_LINEAR ? x : get_log_hz(x, w, inlink->sample_rate);
+ float hertz = s->start + bin * range / (float)(1 << (int)ceil(log2(w)));
char *units;
if (hertz == 0)
for (y = 0; y < s->h && s->single_pic; y+=40) {
float seconds = y * spp / inlink->sample_rate;
char *units = get_time(ctx, seconds, x);
+ if (!units)
+ return AVERROR(ENOMEM);
drawtext(s->outpicref, s->start_x - 8 * strlen(units) - 4, s->start_y + y - 4, units, 0);
av_free(units);
}
for (y = 0; ch == 0 && y < h; y += h / 10) {
- float value = 120.0 * log10(1. - y / (float)h);
+ float value = 120.f * log10f(1.f - y / (float)h);
char *text;
if (value < -120)
return 0;
}
+static float get_value(AVFilterContext *ctx, int ch, int y)
+{
+ ShowSpectrumContext *s = ctx->priv;
+ float *magnitudes = s->magnitudes[ch];
+ float *phases = s->phases[ch];
+ float a;
+
+ switch (s->data) {
+ case D_MAGNITUDE:
+ /* get magnitude */
+ a = magnitudes[y];
+ break;
+ case D_PHASE:
+ /* get phase */
+ a = phases[y];
+ break;
+ default:
+ av_assert0(0);
+ }
+
+ /* apply scale */
+ switch (s->scale) {
+ case LINEAR:
+ a = av_clipf(a, 0, 1);
+ break;
+ case SQRT:
+ a = av_clipf(sqrtf(a), 0, 1);
+ break;
+ case CBRT:
+ a = av_clipf(cbrtf(a), 0, 1);
+ break;
+ case FOURTHRT:
+ a = av_clipf(sqrtf(sqrtf(a)), 0, 1);
+ break;
+ case FIFTHRT:
+ a = av_clipf(powf(a, 0.20), 0, 1);
+ break;
+ case LOG:
+ a = 1.f + log10f(av_clipf(a, 1e-6, 1)) / 6.f; // zero = -120dBFS
+ break;
+ default:
+ av_assert0(0);
+ }
+
+ return a;
+}
+
+static int plot_channel_lin(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
+{
+ ShowSpectrumContext *s = ctx->priv;
+ const int h = s->orientation == VERTICAL ? s->channel_height : s->channel_width;
+ const int ch = jobnr;
+ float yf, uf, vf;
+ int y;
+
+ /* decide color range */
+ color_range(s, ch, &yf, &uf, &vf);
+
+ /* draw the channel */
+ for (y = 0; y < h; y++) {
+ int row = (s->mode == COMBINED) ? y : ch * h + y;
+ float *out = &s->color_buffer[ch][3 * row];
+ float a = get_value(ctx, ch, y);
+
+ pick_color(s, yf, uf, vf, a, out);
+ }
+
+ return 0;
+}
+
+static int plot_channel_log(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
+{
+ ShowSpectrumContext *s = ctx->priv;
+ AVFilterLink *inlink = ctx->inputs[0];
+ const int h = s->orientation == VERTICAL ? s->channel_height : s->channel_width;
+ const int ch = jobnr;
+ float y, yf, uf, vf;
+ int yy = 0;
+
+ /* decide color range */
+ color_range(s, ch, &yf, &uf, &vf);
+
+ /* draw the channel */
+ for (y = 0; y < h && yy < h; yy++) {
+ float pos0 = bin_pos(yy+0, h, inlink->sample_rate);
+ float pos1 = bin_pos(yy+1, h, inlink->sample_rate);
+ float delta = pos1 - pos0;
+ float a0, a1;
+
+ a0 = get_value(ctx, ch, yy+0);
+ a1 = get_value(ctx, ch, FFMIN(yy+1, h-1));
+ for (float j = pos0; j < pos1 && y + j - pos0 < h; j++) {
+ float row = (s->mode == COMBINED) ? y + j - pos0 : ch * h + y + j - pos0;
+ float *out = &s->color_buffer[ch][3 * FFMIN(lrintf(row), h-1)];
+ float lerpfrac = (j - pos0) / delta;
+
+ pick_color(s, yf, uf, vf, lerpfrac * a1 + (1.f-lerpfrac) * a0, out);
+ }
+ y += delta;
+ }
+
+ return 0;
+}
+
static int config_output(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
int i, fft_bits, h, w;
float overlap;
+ switch (s->fscale) {
+ case F_LINEAR: s->plot_channel = plot_channel_lin; break;
+ case F_LOG: s->plot_channel = plot_channel_log; break;
+ default: return AVERROR_BUG;
+ }
+
s->stop = FFMIN(s->stop, inlink->sample_rate / 2);
if (s->stop && s->stop <= s->start) {
av_log(ctx, AV_LOG_ERROR, "Stop frequency should be greater than start.\n");
generate_window_func(s->window_func_lut, s->win_size, s->win_func, &overlap);
if (s->overlap == 1)
s->overlap = overlap;
- s->hop_size = (1. - s->overlap) * s->win_size;
+ s->hop_size = (1.f - s->overlap) * s->win_size;
if (s->hop_size < 1) {
av_log(ctx, AV_LOG_ERROR, "overlap %f too big\n", s->overlap);
return AVERROR(EINVAL);
for (s->win_scale = 0, i = 0; i < s->win_size; i++) {
s->win_scale += s->window_func_lut[i] * s->window_func_lut[i];
}
- s->win_scale = 1. / sqrt(s->win_scale);
+ s->win_scale = 1.f / sqrtf(s->win_scale);
/* prepare the initial picref buffer (black frame) */
av_frame_free(&s->outpicref);
#define RE(y, ch) s->fft_data[ch][y].re
#define IM(y, ch) s->fft_data[ch][y].im
-#define MAGNITUDE(y, ch) hypot(RE(y, ch), IM(y, ch))
-#define PHASE(y, ch) atan2(IM(y, ch), RE(y, ch))
+#define MAGNITUDE(y, ch) hypotf(RE(y, ch), IM(y, ch))
+#define PHASE(y, ch) atan2f(IM(y, ch), RE(y, ch))
static int calc_channel_magnitudes(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
{
}
}
-static int plot_channel(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
-{
- ShowSpectrumContext *s = ctx->priv;
- const int h = s->orientation == VERTICAL ? s->channel_height : s->channel_width;
- const int ch = jobnr;
- float *magnitudes = s->magnitudes[ch];
- float *phases = s->phases[ch];
- float yf, uf, vf;
- int y;
-
- /* decide color range */
- color_range(s, ch, &yf, &uf, &vf);
-
- /* draw the channel */
- for (y = 0; y < h; y++) {
- int row = (s->mode == COMBINED) ? y : ch * h + y;
- float *out = &s->color_buffer[ch][3 * row];
- float a;
-
- switch (s->data) {
- case D_MAGNITUDE:
- /* get magnitude */
- a = magnitudes[y];
- break;
- case D_PHASE:
- /* get phase */
- a = phases[y];
- break;
- default:
- av_assert0(0);
- }
-
- /* apply scale */
- switch (s->scale) {
- case LINEAR:
- a = av_clipf(a, 0, 1);
- break;
- case SQRT:
- a = av_clipf(sqrt(a), 0, 1);
- break;
- case CBRT:
- a = av_clipf(cbrt(a), 0, 1);
- break;
- case FOURTHRT:
- a = av_clipf(sqrt(sqrt(a)), 0, 1);
- break;
- case FIFTHRT:
- a = av_clipf(pow(a, 0.20), 0, 1);
- break;
- case LOG:
- a = 1 + log10(av_clipd(a, 1e-6, 1)) / 6; // zero = -120dBFS
- break;
- default:
- av_assert0(0);
- }
-
- pick_color(s, yf, uf, vf, a, out);
- }
-
- return 0;
-}
-
static int plot_spectrum_column(AVFilterLink *inlink, AVFrame *insamples)
{
AVFilterContext *ctx = inlink->dst;
/* initialize buffer for combining to black */
clear_combine_buffer(s, z);
- ctx->internal->execute(ctx, plot_channel, NULL, NULL, s->nb_display_channels);
+ ctx->internal->execute(ctx, s->plot_channel, NULL, NULL, s->nb_display_channels);
for (y = 0; y < z * 3; y++) {
for (x = 0; x < s->nb_display_channels; x++) {
s->xpos = 0;
if (!s->single_pic && (s->sliding != FULLFRAME || s->xpos == 0)) {
if (s->old_pts < outpicref->pts) {
+ AVFrame *clone;
+
if (s->legend) {
char *units = get_time(ctx, insamples->pts /(float)inlink->sample_rate, x);
+ if (!units)
+ return AVERROR(ENOMEM);
if (s->orientation == VERTICAL) {
for (y = 0; y < 10; y++) {
av_free(units);
}
s->old_pts = outpicref->pts;
- ret = ff_filter_frame(outlink, av_frame_clone(s->outpicref));
+ clone = av_frame_clone(s->outpicref);
+ if (!clone)
+ return AVERROR(ENOMEM);
+ ret = ff_filter_frame(outlink, clone);
if (ret < 0)
return ret;
return 0;
}
}
- if (s->outpicref && av_audio_fifo_size(s->fifo) >= s->win_size) {
+ if (s->outpicref && (av_audio_fifo_size(s->fifo) >= s->win_size ||
+ ff_outlink_get_status(inlink))) {
AVFrame *fin = ff_get_audio_buffer(inlink, s->win_size);
if (!fin)
return AVERROR(ENOMEM);
av_frame_free(&fin);
av_audio_fifo_drain(s->fifo, s->hop_size);
- if (ret <= 0)
+ if (ret <= 0 && !ff_outlink_get_status(inlink))
return ret;
}
}
FF_FILTER_FORWARD_STATUS(inlink, outlink);
- if (ff_outlink_frame_wanted(outlink) && av_audio_fifo_size(s->fifo) < s->win_size) {
- ff_inlink_request_frame(inlink);
+ if (av_audio_fifo_size(s->fifo) >= s->win_size ||
+ ff_outlink_get_status(inlink) == AVERROR_EOF) {
+ ff_filter_set_ready(ctx, 10);
return 0;
}
- if (av_audio_fifo_size(s->fifo) >= s->win_size) {
- ff_filter_set_ready(ctx, 10);
+ if (ff_outlink_frame_wanted(outlink) && av_audio_fifo_size(s->fifo) < s->win_size &&
+ ff_outlink_get_status(inlink) != AVERROR_EOF) {
+ ff_inlink_request_frame(inlink);
return 0;
}
+
return FFERROR_NOT_READY;
}
{ NULL }
};
-AVFilter ff_avf_showspectrum = {
+const AVFilter ff_avf_showspectrum = {
.name = "showspectrum",
.description = NULL_IF_CONFIG_SMALL("Convert input audio to a spectrum video output."),
.uninit = uninit,
{ "log", "logarithmic", 0, AV_OPT_TYPE_CONST, {.i64=LOG}, 0, 0, FLAGS, "scale" },
{ "4thrt","4th root", 0, AV_OPT_TYPE_CONST, {.i64=FOURTHRT}, 0, 0, FLAGS, "scale" },
{ "5thrt","5th root", 0, AV_OPT_TYPE_CONST, {.i64=FIFTHRT}, 0, 0, FLAGS, "scale" },
+ { "fscale", "set frequency scale", OFFSET(fscale), AV_OPT_TYPE_INT, {.i64=F_LINEAR}, 0, NB_FSCALES-1, FLAGS, "fscale" },
+ { "lin", "linear", 0, AV_OPT_TYPE_CONST, {.i64=F_LINEAR}, 0, 0, FLAGS, "fscale" },
+ { "log", "logarithmic", 0, AV_OPT_TYPE_CONST, {.i64=F_LOG}, 0, 0, FLAGS, "fscale" },
{ "saturation", "color saturation multiplier", OFFSET(saturation), AV_OPT_TYPE_FLOAT, {.dbl = 1}, -10, 10, FLAGS },
{ "win_func", "set window function", OFFSET(win_func), AV_OPT_TYPE_INT, {.i64 = WFUNC_HANNING}, 0, NB_WFUNC-1, FLAGS, "win_func" },
{ "rect", "Rectangular", 0, AV_OPT_TYPE_CONST, {.i64=WFUNC_RECT}, 0, 0, FLAGS, "win_func" },
if (consumed >= spb) {
int h = s->orientation == VERTICAL ? s->h : s->w;
- scale_magnitudes(s, 1. / (consumed / spf));
+ scale_magnitudes(s, 1.f / (consumed / spf));
plot_spectrum_column(inlink, fin);
consumed = 0;
x++;
{ NULL }
};
-AVFilter ff_avf_showspectrumpic = {
+const AVFilter ff_avf_showspectrumpic = {
.name = "showspectrumpic",
.description = NULL_IF_CONFIG_SMALL("Convert input audio to a spectrum video output single picture."),
.uninit = uninit,
projects / ffmpeg / blobdiff