avfilter/af_astats: also measure maximal difference between two consecutive samples
While here also mention bit depth in documentation. Signed-off-by: Paul B Mahol <onemda@gmail.com>
This commit is contained in:
		
							parent
							
								
									6074956fa1
								
							
						
					
					
						commit
						59a9998908
					
				| @ -952,23 +952,27 @@ Available keys for each channel are: | |||||||
| DC_offset | DC_offset | ||||||
| Min_level | Min_level | ||||||
| Max_level | Max_level | ||||||
|  | Max_difference | ||||||
| Peak_level | Peak_level | ||||||
| RMS_peak | RMS_peak | ||||||
| RMS_trough | RMS_trough | ||||||
| Crest_factor | Crest_factor | ||||||
| Flat_factor | Flat_factor | ||||||
| Peak_count | Peak_count | ||||||
|  | Bit_depth | ||||||
| 
 | 
 | ||||||
| and for Overall: | and for Overall: | ||||||
| DC_offset | DC_offset | ||||||
| Min_level | Min_level | ||||||
| Max_level | Max_level | ||||||
|  | Max_difference | ||||||
| Peak_level | Peak_level | ||||||
| RMS_level | RMS_level | ||||||
| RMS_peak | RMS_peak | ||||||
| RMS_trough | RMS_trough | ||||||
| Flat_factor | Flat_factor | ||||||
| Peak_count | Peak_count | ||||||
|  | Bit_depth | ||||||
| Number_of_samples | Number_of_samples | ||||||
| 
 | 
 | ||||||
| For example full key look like this @code{lavfi.astats.1.DC_offset} or | For example full key look like this @code{lavfi.astats.1.DC_offset} or | ||||||
| @ -976,6 +980,9 @@ this @code{lavfi.astats.Overall.Peak_count}. | |||||||
| 
 | 
 | ||||||
| For description what each key means read bellow. | For description what each key means read bellow. | ||||||
| 
 | 
 | ||||||
|  | @item reset | ||||||
|  | Set number of frame after which stats are going to be recalculated. | ||||||
|  | Default is disabled. | ||||||
| @end table | @end table | ||||||
| 
 | 
 | ||||||
| A description of each shown parameter follows: | A description of each shown parameter follows: | ||||||
| @ -990,6 +997,9 @@ Minimal sample level. | |||||||
| @item Max level | @item Max level | ||||||
| Maximal sample level. | Maximal sample level. | ||||||
| 
 | 
 | ||||||
|  | @item Max difference | ||||||
|  | Maximal difference between two consecutive samples. | ||||||
|  | 
 | ||||||
| @item Peak level dB | @item Peak level dB | ||||||
| @item RMS level dB | @item RMS level dB | ||||||
| Standard peak and RMS level measured in dBFS. | Standard peak and RMS level measured in dBFS. | ||||||
| @ -1008,6 +1018,9 @@ Flatness (i.e. consecutive samples with the same value) of the signal at its pea | |||||||
| @item Peak count | @item Peak count | ||||||
| Number of occasions (not the number of samples) that the signal attained either | Number of occasions (not the number of samples) that the signal attained either | ||||||
| @var{Min level} or @var{Max level}. | @var{Min level} or @var{Max level}. | ||||||
|  | 
 | ||||||
|  | @item Bit depth | ||||||
|  | Overall bit depth of audio. Number of bits used for each sample. | ||||||
| @end table | @end table | ||||||
| 
 | 
 | ||||||
| @section astreamsync | @section astreamsync | ||||||
|  | |||||||
| @ -33,6 +33,7 @@ typedef struct ChannelStats { | |||||||
|     double min, max; |     double min, max; | ||||||
|     double min_run, max_run; |     double min_run, max_run; | ||||||
|     double min_runs, max_runs; |     double min_runs, max_runs; | ||||||
|  |     double max_diff; | ||||||
|     uint64_t mask; |     uint64_t mask; | ||||||
|     uint64_t min_count, max_count; |     uint64_t min_count, max_count; | ||||||
|     uint64_t nb_samples; |     uint64_t nb_samples; | ||||||
| @ -160,6 +161,7 @@ static inline void update_stat(AudioStatsContext *s, ChannelStats *p, double d) | |||||||
|     p->sigma_x += d; |     p->sigma_x += d; | ||||||
|     p->sigma_x2 += d * d; |     p->sigma_x2 += d * d; | ||||||
|     p->avg_sigma_x2 = p->avg_sigma_x2 * s->mult + (1.0 - s->mult) * d * d; |     p->avg_sigma_x2 = p->avg_sigma_x2 * s->mult + (1.0 - s->mult) * d * d; | ||||||
|  |     p->max_diff = FFMAX(p->max_diff, FFABS(d - p->last)); | ||||||
|     p->last = d; |     p->last = d; | ||||||
|     p->mask |= llrint(d * (1LLU<<63)); |     p->mask |= llrint(d * (1LLU<<63)); | ||||||
| 
 | 
 | ||||||
| @ -190,7 +192,7 @@ static void set_metadata(AudioStatsContext *s, AVDictionary **metadata) | |||||||
| { | { | ||||||
|     uint64_t mask = 0, min_count = 0, max_count = 0, nb_samples = 0; |     uint64_t mask = 0, min_count = 0, max_count = 0, nb_samples = 0; | ||||||
|     double min_runs = 0, max_runs = 0, |     double min_runs = 0, max_runs = 0, | ||||||
|            min = DBL_MAX, max = DBL_MIN, |            min = DBL_MAX, max = DBL_MIN, max_diff = 0, | ||||||
|            max_sigma_x = 0, |            max_sigma_x = 0, | ||||||
|            sigma_x = 0, |            sigma_x = 0, | ||||||
|            sigma_x2 = 0, |            sigma_x2 = 0, | ||||||
| @ -206,6 +208,7 @@ static void set_metadata(AudioStatsContext *s, AVDictionary **metadata) | |||||||
| 
 | 
 | ||||||
|         min = FFMIN(min, p->min); |         min = FFMIN(min, p->min); | ||||||
|         max = FFMAX(max, p->max); |         max = FFMAX(max, p->max); | ||||||
|  |         max_diff = FFMAX(max_diff, p->max_diff); | ||||||
|         min_sigma_x2 = FFMIN(min_sigma_x2, p->min_sigma_x2); |         min_sigma_x2 = FFMIN(min_sigma_x2, p->min_sigma_x2); | ||||||
|         max_sigma_x2 = FFMAX(max_sigma_x2, p->max_sigma_x2); |         max_sigma_x2 = FFMAX(max_sigma_x2, p->max_sigma_x2); | ||||||
|         sigma_x += p->sigma_x; |         sigma_x += p->sigma_x; | ||||||
| @ -222,6 +225,7 @@ static void set_metadata(AudioStatsContext *s, AVDictionary **metadata) | |||||||
|         set_meta(metadata, c + 1, "DC_offset", "%f", p->sigma_x / p->nb_samples); |         set_meta(metadata, c + 1, "DC_offset", "%f", p->sigma_x / p->nb_samples); | ||||||
|         set_meta(metadata, c + 1, "Min_level", "%f", p->min); |         set_meta(metadata, c + 1, "Min_level", "%f", p->min); | ||||||
|         set_meta(metadata, c + 1, "Max_level", "%f", p->max); |         set_meta(metadata, c + 1, "Max_level", "%f", p->max); | ||||||
|  |         set_meta(metadata, c + 1, "Max_difference", "%f", p->max_diff); | ||||||
|         set_meta(metadata, c + 1, "Peak_level", "%f", LINEAR_TO_DB(FFMAX(-p->min, p->max))); |         set_meta(metadata, c + 1, "Peak_level", "%f", LINEAR_TO_DB(FFMAX(-p->min, p->max))); | ||||||
|         set_meta(metadata, c + 1, "RMS_level", "%f", LINEAR_TO_DB(sqrt(p->sigma_x2 / p->nb_samples))); |         set_meta(metadata, c + 1, "RMS_level", "%f", LINEAR_TO_DB(sqrt(p->sigma_x2 / p->nb_samples))); | ||||||
|         set_meta(metadata, c + 1, "RMS_peak", "%f", LINEAR_TO_DB(sqrt(p->max_sigma_x2))); |         set_meta(metadata, c + 1, "RMS_peak", "%f", LINEAR_TO_DB(sqrt(p->max_sigma_x2))); | ||||||
| @ -235,6 +239,7 @@ static void set_metadata(AudioStatsContext *s, AVDictionary **metadata) | |||||||
|     set_meta(metadata, 0, "Overall.DC_offset", "%f", max_sigma_x / (nb_samples / s->nb_channels)); |     set_meta(metadata, 0, "Overall.DC_offset", "%f", max_sigma_x / (nb_samples / s->nb_channels)); | ||||||
|     set_meta(metadata, 0, "Overall.Min_level", "%f", min); |     set_meta(metadata, 0, "Overall.Min_level", "%f", min); | ||||||
|     set_meta(metadata, 0, "Overall.Max_level", "%f", max); |     set_meta(metadata, 0, "Overall.Max_level", "%f", max); | ||||||
|  |     set_meta(metadata, 0, "Overall.Max_difference", "%f", max_diff); | ||||||
|     set_meta(metadata, 0, "Overall.Peak_level", "%f", LINEAR_TO_DB(FFMAX(-min, max))); |     set_meta(metadata, 0, "Overall.Peak_level", "%f", LINEAR_TO_DB(FFMAX(-min, max))); | ||||||
|     set_meta(metadata, 0, "Overall.RMS_level", "%f", LINEAR_TO_DB(sqrt(sigma_x2 / nb_samples))); |     set_meta(metadata, 0, "Overall.RMS_level", "%f", LINEAR_TO_DB(sqrt(sigma_x2 / nb_samples))); | ||||||
|     set_meta(metadata, 0, "Overall.RMS_peak", "%f", LINEAR_TO_DB(sqrt(max_sigma_x2))); |     set_meta(metadata, 0, "Overall.RMS_peak", "%f", LINEAR_TO_DB(sqrt(max_sigma_x2))); | ||||||
| @ -292,7 +297,7 @@ static void print_stats(AVFilterContext *ctx) | |||||||
|     AudioStatsContext *s = ctx->priv; |     AudioStatsContext *s = ctx->priv; | ||||||
|     uint64_t mask = 0, min_count = 0, max_count = 0, nb_samples = 0; |     uint64_t mask = 0, min_count = 0, max_count = 0, nb_samples = 0; | ||||||
|     double min_runs = 0, max_runs = 0, |     double min_runs = 0, max_runs = 0, | ||||||
|            min = DBL_MAX, max = DBL_MIN, |            min = DBL_MAX, max = DBL_MIN, max_diff = 0, | ||||||
|            max_sigma_x = 0, |            max_sigma_x = 0, | ||||||
|            sigma_x = 0, |            sigma_x = 0, | ||||||
|            sigma_x2 = 0, |            sigma_x2 = 0, | ||||||
| @ -308,6 +313,7 @@ static void print_stats(AVFilterContext *ctx) | |||||||
| 
 | 
 | ||||||
|         min = FFMIN(min, p->min); |         min = FFMIN(min, p->min); | ||||||
|         max = FFMAX(max, p->max); |         max = FFMAX(max, p->max); | ||||||
|  |         max_diff = FFMAX(max_diff, p->max_diff); | ||||||
|         min_sigma_x2 = FFMIN(min_sigma_x2, p->min_sigma_x2); |         min_sigma_x2 = FFMIN(min_sigma_x2, p->min_sigma_x2); | ||||||
|         max_sigma_x2 = FFMAX(max_sigma_x2, p->max_sigma_x2); |         max_sigma_x2 = FFMAX(max_sigma_x2, p->max_sigma_x2); | ||||||
|         sigma_x += p->sigma_x; |         sigma_x += p->sigma_x; | ||||||
| @ -325,6 +331,7 @@ static void print_stats(AVFilterContext *ctx) | |||||||
|         av_log(ctx, AV_LOG_INFO, "DC offset: %f\n", p->sigma_x / p->nb_samples); |         av_log(ctx, AV_LOG_INFO, "DC offset: %f\n", p->sigma_x / p->nb_samples); | ||||||
|         av_log(ctx, AV_LOG_INFO, "Min level: %f\n", p->min); |         av_log(ctx, AV_LOG_INFO, "Min level: %f\n", p->min); | ||||||
|         av_log(ctx, AV_LOG_INFO, "Max level: %f\n", p->max); |         av_log(ctx, AV_LOG_INFO, "Max level: %f\n", p->max); | ||||||
|  |         av_log(ctx, AV_LOG_INFO, "Max difference: %f\n", p->max_diff); | ||||||
|         av_log(ctx, AV_LOG_INFO, "Peak level dB: %f\n", LINEAR_TO_DB(FFMAX(-p->min, p->max))); |         av_log(ctx, AV_LOG_INFO, "Peak level dB: %f\n", LINEAR_TO_DB(FFMAX(-p->min, p->max))); | ||||||
|         av_log(ctx, AV_LOG_INFO, "RMS level dB: %f\n", LINEAR_TO_DB(sqrt(p->sigma_x2 / p->nb_samples))); |         av_log(ctx, AV_LOG_INFO, "RMS level dB: %f\n", LINEAR_TO_DB(sqrt(p->sigma_x2 / p->nb_samples))); | ||||||
|         av_log(ctx, AV_LOG_INFO, "RMS peak dB: %f\n", LINEAR_TO_DB(sqrt(p->max_sigma_x2))); |         av_log(ctx, AV_LOG_INFO, "RMS peak dB: %f\n", LINEAR_TO_DB(sqrt(p->max_sigma_x2))); | ||||||
| @ -340,6 +347,7 @@ static void print_stats(AVFilterContext *ctx) | |||||||
|     av_log(ctx, AV_LOG_INFO, "DC offset: %f\n", max_sigma_x / (nb_samples / s->nb_channels)); |     av_log(ctx, AV_LOG_INFO, "DC offset: %f\n", max_sigma_x / (nb_samples / s->nb_channels)); | ||||||
|     av_log(ctx, AV_LOG_INFO, "Min level: %f\n", min); |     av_log(ctx, AV_LOG_INFO, "Min level: %f\n", min); | ||||||
|     av_log(ctx, AV_LOG_INFO, "Max level: %f\n", max); |     av_log(ctx, AV_LOG_INFO, "Max level: %f\n", max); | ||||||
|  |     av_log(ctx, AV_LOG_INFO, "Max difference: %f\n", max_diff); | ||||||
|     av_log(ctx, AV_LOG_INFO, "Peak level dB: %f\n", LINEAR_TO_DB(FFMAX(-min, max))); |     av_log(ctx, AV_LOG_INFO, "Peak level dB: %f\n", LINEAR_TO_DB(FFMAX(-min, max))); | ||||||
|     av_log(ctx, AV_LOG_INFO, "RMS level dB: %f\n", LINEAR_TO_DB(sqrt(sigma_x2 / nb_samples))); |     av_log(ctx, AV_LOG_INFO, "RMS level dB: %f\n", LINEAR_TO_DB(sqrt(sigma_x2 / nb_samples))); | ||||||
|     av_log(ctx, AV_LOG_INFO, "RMS peak dB: %f\n", LINEAR_TO_DB(sqrt(max_sigma_x2))); |     av_log(ctx, AV_LOG_INFO, "RMS peak dB: %f\n", LINEAR_TO_DB(sqrt(max_sigma_x2))); | ||||||
|  | |||||||
		Loading…
	
	
			
			x
			
			
		
	
		Reference in New Issue
	
	Block a user