diff --git a/doc/ffmpeg-resampler.texi b/doc/ffmpeg-resampler.texi index 8470c042da..863253aaa7 100644 --- a/doc/ffmpeg-resampler.texi +++ b/doc/ffmpeg-resampler.texi @@ -106,29 +106,54 @@ select triangular dither select triangular dither with high pass @end table +@item resampler +Set resampling engine. Default value is swr. + +Supported values: +@table @samp +@item swr +select the native SW Resampler; filter options precision and cheby are not +applicable in this case. +@item soxr +select the SoX Resampler (where available); compensation, and filter options +filter_size, phase_shift, filter_type & kaiser_beta, are not applicable in this +case. +@end table + @item filter_size -Set resampling filter size, default value is 16. +For swr only, set resampling filter size, default value is 16. @item phase_shift -Set resampling phase shift, default value is 10, must be included +For swr only, set resampling phase shift, default value is 10, must be included between 0 and 30. @item linear_interp Use Linear Interpolation if set to 1, default value is 0. @item cutoff -Set cutoff frequency ratio. Must be a float value between 0 and 1, -default value is 0.8. +Set cutoff frequency (swr: 6dB point; soxr: 0dB point) ratio; must be a float +value between 0 and 1. Default value is 0.8 with swr, and 0.91 with soxr +(which, with a sample-rate of 44100, preserves the entire audio band to 20kHz). + +@item precision +For soxr only, the precision in bits to which the resampled signal will be +calculated. The default value of 20 (which, with suitable dithering, is +appropriate for a destination bit-depth of 16) gives SoX's 'High Quality'; a +value of 28 gives SoX's 'Very High Quality'. + +@item cheby +For soxr only, selects passband rolloff none (Chebyshev) & higher-precision +approximation for 'irrational' ratios. Default value is 0. @item min_comp -Set the minimum difference between timestamps and audio data (in +For swr only, set the minimum difference between timestamps and audio data (in seconds) to trigger stretching/squeezing/filling or trimming of the data to make it match the timestamps. The default is that stretching/squeezing/filling and trimming is disabled (@option{min_comp} = @code{FLT_MAX}). @item min_hard_comp -Set the minimum difference between timestamps and audio data (in +For swr only, set the minimum difference between timestamps and audio data (in seconds) to trigger adding/dropping samples to make it match the timestamps. This option effectively is a threshold to select between hard (trim/fill) and soft (squeeze/stretch) compensation. Note that @@ -136,14 +161,14 @@ all compensation is by default disabled through @option{min_comp}. The default is 0.1. @item comp_duration -Set duration (in seconds) over which data is stretched/squeezed to -make it match the timestamps. Must be a non-negative double float -value, default value is 1.0. +For swr only, set duration (in seconds) over which data is stretched/squeezed +to make it match the timestamps. Must be a non-negative double float value, +default value is 1.0. @item max_soft_comp -Set maximum factor by which data is stretched/squeezed to make it -match the timestamps. Must be a non-negative double float value, -default value is 0. +For swr only, set maximum factor by which data is stretched/squeezed to make it +match the timestamps. Must be a non-negative double float value, default value +is 0. @item matrix_encoding Select matrixed stereo encoding. @@ -161,7 +186,7 @@ select Dolby Pro Logic II Default value is @code{none}. @item filter_type -Select resampling filter type. This only affects resampling +For swr only, select resampling filter type. This only affects resampling operations. It accepts the following values: @@ -175,8 +200,8 @@ select Kaiser Windowed Sinc @end table @item kaiser_beta -Set Kaiser Window Beta value. Must be an integer included between 2 -and 16, default value is 9. +For swr only, set Kaiser Window Beta value. Must be an integer included between +2 and 16, default value is 9. @end table diff --git a/libswresample/swresample.c b/libswresample/swresample.c index 5caab4046f..263acfa7d1 100644 --- a/libswresample/swresample.c +++ b/libswresample/swresample.c @@ -80,15 +80,17 @@ static const AVOption options[]={ {"triangular" , "select triangular dither" , 0 , AV_OPT_TYPE_CONST, {.i64=SWR_DITHER_TRIANGULAR }, INT_MIN, INT_MAX , PARAM, "dither_method"}, {"triangular_hp" , "select triangular dither with high pass" , 0 , AV_OPT_TYPE_CONST, {.i64=SWR_DITHER_TRIANGULAR_HIGHPASS }, INT_MIN, INT_MAX, PARAM, "dither_method"}, -{"filter_size" , "set resampling filter size" , OFFSET(filter_size) , AV_OPT_TYPE_INT , {.i64=16 }, 0 , INT_MAX , PARAM }, -{"phase_shift" , "set resampling phase shift" , OFFSET(phase_shift) , AV_OPT_TYPE_INT , {.i64=10 }, 0 , 30 , PARAM }, +{"filter_size" , "set swr resampling filter size", OFFSET(filter_size) , AV_OPT_TYPE_INT , {.i64=16 }, 0 , INT_MAX , PARAM }, +{"phase_shift" , "set swr resampling phase shift", OFFSET(phase_shift) , AV_OPT_TYPE_INT , {.i64=10 }, 0 , 30 , PARAM }, {"linear_interp" , "enable linear interpolation" , OFFSET(linear_interp) , AV_OPT_TYPE_INT , {.i64=0 }, 0 , 1 , PARAM }, {"cutoff" , "set cutoff frequency ratio" , OFFSET(cutoff) , AV_OPT_TYPE_DOUBLE,{.dbl=0. }, 0 , 1 , PARAM }, {"resampler" , "set resampling Engine" , OFFSET(engine) , AV_OPT_TYPE_INT , {.i64=0 }, 0 , SWR_ENGINE_NB-1, PARAM, "resampler"}, {"swr" , "select SW Resampler" , 0 , AV_OPT_TYPE_CONST, {.i64=SWR_ENGINE_SWR }, INT_MIN, INT_MAX , PARAM, "resampler"}, {"soxr" , "select SoX Resampler" , 0 , AV_OPT_TYPE_CONST, {.i64=SWR_ENGINE_SOXR }, INT_MIN, INT_MAX , PARAM, "resampler"}, -{"precision" , "set resampling precision" , OFFSET(precision) , AV_OPT_TYPE_DOUBLE,{.dbl=20.0 }, 15.0 , 33.0 , PARAM }, -{"cheby" , "enable Chebyshev passband" , OFFSET(cheby) , AV_OPT_TYPE_INT , {.i64=0 }, 0 , 1 , PARAM }, +{"precision" , "set soxr resampling precision (in bits)" + , OFFSET(precision) , AV_OPT_TYPE_DOUBLE,{.dbl=20.0 }, 15.0 , 33.0 , PARAM }, +{"cheby" , "enable soxr Chebyshev passband & higher-precision irrational ratio approximation" + , OFFSET(cheby) , AV_OPT_TYPE_INT , {.i64=0 }, 0 , 1 , PARAM }, {"min_comp" , "set minimum difference between timestamps and audio data (in seconds) below which no timestamp compensation of either kind is applied" , OFFSET(min_compensation),AV_OPT_TYPE_FLOAT ,{.dbl=FLT_MAX }, 0 , FLT_MAX , PARAM }, {"min_hard_comp" , "set minimum difference between timestamps and audio data (in seconds) to trigger padding/trimming the data." @@ -105,12 +107,12 @@ static const AVOption options[]={ { "dolby", "select Dolby", 0, AV_OPT_TYPE_CONST, { .i64 = AV_MATRIX_ENCODING_DOLBY }, INT_MIN, INT_MAX, PARAM, "matrix_encoding" }, { "dplii", "select Dolby Pro Logic II", 0, AV_OPT_TYPE_CONST, { .i64 = AV_MATRIX_ENCODING_DPLII }, INT_MIN, INT_MAX, PARAM, "matrix_encoding" }, -{ "filter_type" , "select filter type" , OFFSET(filter_type) , AV_OPT_TYPE_INT , { .i64 = SWR_FILTER_TYPE_KAISER }, SWR_FILTER_TYPE_CUBIC, SWR_FILTER_TYPE_KAISER, PARAM, "filter_type" }, +{ "filter_type" , "select swr filter type" , OFFSET(filter_type) , AV_OPT_TYPE_INT , { .i64 = SWR_FILTER_TYPE_KAISER }, SWR_FILTER_TYPE_CUBIC, SWR_FILTER_TYPE_KAISER, PARAM, "filter_type" }, { "cubic" , "select cubic" , 0 , AV_OPT_TYPE_CONST, { .i64 = SWR_FILTER_TYPE_CUBIC }, INT_MIN, INT_MAX, PARAM, "filter_type" }, { "blackman_nuttall", "select Blackman Nuttall Windowed Sinc", 0 , AV_OPT_TYPE_CONST, { .i64 = SWR_FILTER_TYPE_BLACKMAN_NUTTALL }, INT_MIN, INT_MAX, PARAM, "filter_type" }, { "kaiser" , "select Kaiser Windowed Sinc" , 0 , AV_OPT_TYPE_CONST, { .i64 = SWR_FILTER_TYPE_KAISER }, INT_MIN, INT_MAX, PARAM, "filter_type" }, -{ "kaiser_beta" , "set Kaiser Window Beta" , OFFSET(kaiser_beta) , AV_OPT_TYPE_INT , {.i64=9 }, 2 , 16 , PARAM }, +{ "kaiser_beta" , "set swr Kaiser Window Beta" , OFFSET(kaiser_beta) , AV_OPT_TYPE_INT , {.i64=9 }, 2 , 16 , PARAM }, {0} }; diff --git a/libswresample/swresample_internal.h b/libswresample/swresample_internal.h index d471517d9e..70a361ba9e 100644 --- a/libswresample/swresample_internal.h +++ b/libswresample/swresample_internal.h @@ -74,17 +74,17 @@ struct SwrContext { int filter_size; /**< length of each FIR filter in the resampling filterbank relative to the cutoff frequency */ int phase_shift; /**< log2 of the number of entries in the resampling polyphase filterbank */ int linear_interp; /**< if 1 then the resampling FIR filter will be linearly interpolated */ - double cutoff; /**< resampling cutoff frequency. 1.0 corresponds to half the output sample rate */ - enum SwrFilterType filter_type; /**< resampling filter type */ - int kaiser_beta; /**< beta value for Kaiser window (only applicable if filter_type == AV_FILTER_TYPE_KAISER) */ - double precision; /**< resampling precision (in bits) */ - int cheby; /**< if 1 then the resampling FIR filter will be configured for maximal passband flatness */ + double cutoff; /**< resampling cutoff frequency (swr: 6dB point; soxr: 0dB point). 1.0 corresponds to half the output sample rate */ + enum SwrFilterType filter_type; /**< swr resampling filter type */ + int kaiser_beta; /**< swr beta value for Kaiser window (only applicable if filter_type == AV_FILTER_TYPE_KAISER) */ + double precision; /**< soxr resampling precision (in bits) */ + int cheby; /**< soxr: if 1 then passband rolloff will be none (Chebyshev) & irrational ratio approximation precision will be higher */ - float min_compensation; ///< minimum below which no compensation will happen - float min_hard_compensation; ///< minimum below which no silence inject / sample drop will happen - float soft_compensation_duration; ///< duration over which soft compensation is applied - float max_soft_compensation; ///< maximum soft compensation in seconds over soft_compensation_duration - float async; ///< simple 1 parameter async, similar to ffmpegs -async + float min_compensation; ///< swr minimum below which no compensation will happen + float min_hard_compensation; ///< swr minimum below which no silence inject / sample drop will happen + float soft_compensation_duration; ///< swr duration over which soft compensation is applied + float max_soft_compensation; ///< swr maximum soft compensation in seconds over soft_compensation_duration + float async; ///< swr simple 1 parameter async, similar to ffmpegs -async int resample_first; ///< 1 if resampling must come first, 0 if rematrixing int rematrix; ///< flag to indicate if rematrixing is needed (basically if input and output layouts mismatch)