avfilter: add spectrumsynth filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Changelog

@@ -52,6 +52,7 @@ version <next>:
 - automatic bitstream filtering
 - showspectrumpic filter
 - libstagefright support removed
+- spectrumsynth filter
 
 
 version 2.8:

configure

@@ -2903,6 +2903,8 @@ showspectrumpic_filter_deps="avcodec"
 showspectrumpic_filter_select="fft"
 sofalizer_filter_deps="netcdf avcodec"
 sofalizer_filter_select="fft"
+spectrumsynth_filter_deps="avcodec"
+spectrumsynth_filter_select="fft"
 spp_filter_deps="gpl avcodec"
 spp_filter_select="fft idctdsp fdctdsp me_cmp pixblockdsp"
 stereo3d_filter_deps="gpl"
@@ -6081,6 +6083,7 @@ enabled sofalizer_filter    && prepend avfilter_deps "avcodec"
 enabled showfreqs_filter    && prepend avfilter_deps "avcodec"
 enabled showspectrum_filter && prepend avfilter_deps "avcodec"
 enabled smartblur_filter    && prepend avfilter_deps "swscale"
+enabled spectrumsynth_filter && prepend avfilter_deps "avcodec"
 enabled subtitles_filter    && prepend avfilter_deps "avformat avcodec"
 enabled uspp_filter         && prepend avfilter_deps "avcodec"
 

doc/filters.texi

@@ -14578,6 +14578,7 @@ Default is @code{combined}.
 
 @end table
 
+@anchor{showspectrum}
 @section showspectrum
 
 Convert input audio to a video output, representing the audio frequency
@@ -15003,6 +15004,68 @@ ffmpeg -i audio.mp3 -filter_complex "showwavespic,colorchannelmixer=rr=66/255:gg
 @end example
 @end itemize
 
+@section spectrumsynth
+
+Sythesize audio from 2 input video spectrums, first input stream represents
+magnitude across time and second represents phase across time.
+The filter will transform from frequency domain as displayed in videos back
+to time domain as presented in audio output.
+
+This filter is primarly created for reversing processed @ref{showspectrum}
+filter outputs, but can synthesize sound from other spectrograms too.
+But in such case results are going to be poor if the phase data is not
+available, because in such cases phase data need to be recreated, usually
+its just recreated from random noise.
+For best results use gray only output (@code{channel} color mode in
+@ref{showspectrum} filter) and @code{log} scale for magnitude video and
+@code{lin} scale for phase video. To produce phase, for 2nd video, use
+@code{data} option. Inputs videos should generally use @code{fullframe}
+slide mode as that saves resources needed for decoding video.
+
+The filter accepts the following options:
+
+@table @option
+@item sample_rate
+Specify sample rate of output audio, the sample rate of audio from which
+spectrum was generated may differ.
+
+@item channels
+Set number of channels represented in input video spectrums.
+
+@item scale
+Set scale which was used when generating magnitude input spectrum.
+Can be @code{lin} or @code{log}. Default is @code{log}.
+
+@item slide
+Set slide which was used when generating inputs spectrums.
+Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
+Default is @code{fullframe}.
+
+@item win_func
+Set window function used for resynthesis.
+
+@item overlap
+Set window overlap. In range @code{[0, 1]}. Default is @code{1},
+which means optimal overlap for selected window function will be picked.
+
+@item orientation
+Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
+Default is @code{vertical}.
+@end table
+
+@subsection Examples
+
+@itemize
+@item
+First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
+then resynthesize videos back to audio with spectrumsynth:
+@example
+ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=log:overlap=0.875:color=channel:slide=fullframe:data=magnitude -an -c:v rawvideo magnitude.nut
+ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=lin:overlap=0.875:color=channel:slide=fullframe:data=phase -an -c:v rawvideo phase.nut
+ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_fun=hann:overlap=0.875:slide=fullframe output.flac
+@end example
+@end itemize
+
 @section split, asplit
 
 Split input into several identical outputs.

libavfilter/Makefile

@@ -290,6 +290,7 @@ OBJS-$(CONFIG_SHOWSPECTRUMPIC_FILTER)        += avf_showspectrum.o window_func.o
 OBJS-$(CONFIG_SHOWVOLUME_FILTER)             += avf_showvolume.o
 OBJS-$(CONFIG_SHOWWAVES_FILTER)              += avf_showwaves.o
 OBJS-$(CONFIG_SHOWWAVESPIC_FILTER)           += avf_showwaves.o
+OBJS-$(CONFIG_SPECTRUMSYNTH_FILTER)          += vaf_spectrumsynth.o window_func.o
 
 # multimedia sources
 OBJS-$(CONFIG_AMOVIE_FILTER)                 += src_movie.o

libavfilter/allfilters.c

@@ -310,6 +310,7 @@ void avfilter_register_all(void)
     REGISTER_FILTER(SHOWVOLUME,     showvolume,     avf);
     REGISTER_FILTER(SHOWWAVES,      showwaves,      avf);
     REGISTER_FILTER(SHOWWAVESPIC,   showwavespic,   avf);
+    REGISTER_FILTER(SPECTRUMSYNTH,  spectrumsynth,  vaf);
 
     /* multimedia sources */
     REGISTER_FILTER(AMOVIE,         amovie,         avsrc);

libavfilter/version.h

@@ -30,7 +30,7 @@
 #include "libavutil/version.h"
 
 #define LIBAVFILTER_VERSION_MAJOR   6
-#define LIBAVFILTER_VERSION_MINOR  23
+#define LIBAVFILTER_VERSION_MINOR  24
 #define LIBAVFILTER_VERSION_MICRO 100
 
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \