1FFMPEG-CODECS(1) FFMPEG-CODECS(1)
2
6 ffmpeg-codecs - FFmpeg codecs
7
9 This document describes the codecs (decoders and encoders) provided by
10 the libavcodec library.
11
13 libavcodec provides some generic global options, which can be set on
14 all the encoders and decoders. In addition each codec may support so-
15 called private options, which are specific for a given codec.
16 Sometimes, a global option may only affect a specific kind of codec,
18 and may be nonsensical or ignored by another, so you need to be aware
19 of the meaning of the specified options. Also some options are meant
20 only for decoding or encoding.
21 Options may be set by specifying -option value in the FFmpeg tools, or
23 by setting the value explicitly in the "AVCodecContext" options or
24 using the libavutil/opt.h API for programmatic use.
25 The list of supported options follow:
27 b integer (encoding,audio,video)
29 Set bitrate in bits/s. Default value is 200K.
30 ab integer (encoding,audio)
32 Set audio bitrate (in bits/s). Default value is 128K.
33 bt integer (encoding,video)
35 Set video bitrate tolerance (in bits/s). In 1-pass mode, bitrate
36 tolerance specifies how far ratecontrol is willing to deviate from
37 the target average bitrate value. This is not related to min/max
38 bitrate. Lowering tolerance too much has an adverse effect on
39 quality.
40 flags flags (decoding/encoding,audio,video,subtitles)
42 Set generic flags.
43 Possible values:
45 mv4 Use four motion vector by macroblock (mpeg4).
47 qpel
49 Use 1/4 pel motion compensation.
50 loop
52 Use loop filter.
53 qscale
55 Use fixed qscale.
56 pass1
58 Use internal 2pass ratecontrol in first pass mode.
59 pass2
61 Use internal 2pass ratecontrol in second pass mode.
62 gray
64 Only decode/encode grayscale.
65 psnr
67 Set error[?] variables during encoding.
68 truncated
70 Input bitstream might be randomly truncated.
71 drop_changed
73 Don't output frames whose parameters differ from first decoded
74 frame in stream. Error AVERROR_INPUT_CHANGED is returned when
75 a frame is dropped.
76 ildct
78 Use interlaced DCT.
79 low_delay
81 Force low delay.
82 global_header
84 Place global headers in extradata instead of every keyframe.
85 bitexact
87 Only write platform-, build- and time-independent data. (except
88 (I)DCT). This ensures that file and data checksums are
89 reproducible and match between platforms. Its primary use is
90 for regression testing.
91 aic Apply H263 advanced intra coding / mpeg4 ac prediction.
93 ilme
95 Apply interlaced motion estimation.
96 cgop
98 Use closed gop.
99 output_corrupt
101 Output even potentially corrupted frames.
102 time_base rational number
104 Set codec time base.
105 It is the fundamental unit of time (in seconds) in terms of which
107 frame timestamps are represented. For fixed-fps content, timebase
108 should be "1 / frame_rate" and timestamp increments should be
109 identically 1.
110 g integer (encoding,video)
112 Set the group of picture (GOP) size. Default value is 12.
113 ar integer (decoding/encoding,audio)
115 Set audio sampling rate (in Hz).
116 ac integer (decoding/encoding,audio)
118 Set number of audio channels.
119 cutoff integer (encoding,audio)
121 Set cutoff bandwidth. (Supported only by selected encoders, see
122 their respective documentation sections.)
123 frame_size integer (encoding,audio)
125 Set audio frame size.
126 Each submitted frame except the last must contain exactly
128 frame_size samples per channel. May be 0 when the codec has
129 CODEC_CAP_VARIABLE_FRAME_SIZE set, in that case the frame size is
130 not restricted. It is set by some decoders to indicate constant
131 frame size.
132 frame_number integer
134 Set the frame number.
135 delay integer
137 qcomp float (encoding,video)
138 Set video quantizer scale compression (VBR). It is used as a
139 constant in the ratecontrol equation. Recommended range for default
140 rc_eq: 0.0-1.0.
141 qblur float (encoding,video)
143 Set video quantizer scale blur (VBR).
144 qmin integer (encoding,video)
146 Set min video quantizer scale (VBR). Must be included between -1
147 and 69, default value is 2.
148 qmax integer (encoding,video)
150 Set max video quantizer scale (VBR). Must be included between -1
151 and 1024, default value is 31.
152 qdiff integer (encoding,video)
154 Set max difference between the quantizer scale (VBR).
155 bf integer (encoding,video)
157 Set max number of B frames between non-B-frames.
158 Must be an integer between -1 and 16. 0 means that B-frames are
160 disabled. If a value of -1 is used, it will choose an automatic
161 value depending on the encoder.
162 Default value is 0.
164 b_qfactor float (encoding,video)
166 Set qp factor between P and B frames.
167 codec_tag integer
169 bug flags (decoding,video)
170 Workaround not auto detected encoder bugs.
171 Possible values:
173 autodetect
175 xvid_ilace
176 Xvid interlacing bug (autodetected if fourcc==XVIX)
177 ump4
179 (autodetected if fourcc==UMP4)
180 no_padding
182 padding bug (autodetected)
183 amv
185 qpel_chroma
186 std_qpel
187 old standard qpel (autodetected per fourcc/version)
188 qpel_chroma2
190 direct_blocksize
191 direct-qpel-blocksize bug (autodetected per fourcc/version)
192 edge
194 edge padding bug (autodetected per fourcc/version)
195 hpel_chroma
197 dc_clip
198 ms Workaround various bugs in microsoft broken decoders.
199 trunc
201 trancated frames
202 strict integer (decoding/encoding,audio,video)
204 Specify how strictly to follow the standards.
205 Possible values:
207 very
209 strictly conform to an older more strict version of the spec or
210 reference software
211 strict
213 strictly conform to all the things in the spec no matter what
214 consequences
215 normal
217 unofficial
218 allow unofficial extensions
219 experimental
221 allow non standardized experimental things, experimental
222 (unfinished/work in progress/not well tested) decoders and
223 encoders. Note: experimental decoders can pose a security
224 risk, do not use this for decoding untrusted input.
225 b_qoffset float (encoding,video)
227 Set QP offset between P and B frames.
228 err_detect flags (decoding,audio,video)
230 Set error detection flags.
231 Possible values:
233 crccheck
235 verify embedded CRCs
236 bitstream
238 detect bitstream specification deviations
239 buffer
241 detect improper bitstream length
242 explode
244 abort decoding on minor error detection
245 ignore_err
247 ignore decoding errors, and continue decoding. This is useful
248 if you want to analyze the content of a video and thus want
249 everything to be decoded no matter what. This option will not
250 result in a video that is pleasing to watch in case of errors.
251 careful
253 consider things that violate the spec and have not been seen in
254 the wild as errors
255 compliant
257 consider all spec non compliancies as errors
258 aggressive
260 consider things that a sane encoder should not do as an error
261 has_b_frames integer
263 block_align integer
264 rc_override_count integer
265 maxrate integer (encoding,audio,video)
266 Set max bitrate tolerance (in bits/s). Requires bufsize to be set.
267 minrate integer (encoding,audio,video)
269 Set min bitrate tolerance (in bits/s). Most useful in setting up a
270 CBR encode. It is of little use elsewise.
271 bufsize integer (encoding,audio,video)
273 Set ratecontrol buffer size (in bits).
274 i_qfactor float (encoding,video)
276 Set QP factor between P and I frames.
277 i_qoffset float (encoding,video)
279 Set QP offset between P and I frames.
280 dct integer (encoding,video)
282 Set DCT algorithm.
283 Possible values:
285 auto
287 autoselect a good one (default)
288 fastint
290 fast integer
291 int accurate integer
293 mmx
295 altivec
296 faan
297 floating point AAN DCT
298 lumi_mask float (encoding,video)
300 Compress bright areas stronger than medium ones.
301 tcplx_mask float (encoding,video)
303 Set temporal complexity masking.
304 scplx_mask float (encoding,video)
306 Set spatial complexity masking.
307 p_mask float (encoding,video)
309 Set inter masking.
310 dark_mask float (encoding,video)
312 Compress dark areas stronger than medium ones.
313 idct integer (decoding/encoding,video)
315 Select IDCT implementation.
316 Possible values:
318 auto
320 int
321 simple
322 simplemmx
323 simpleauto
324 Automatically pick a IDCT compatible with the simple one
325 arm
327 altivec
328 sh4
329 simplearm
330 simplearmv5te
331 simplearmv6
332 simpleneon
333 xvid
334 faani
335 floating point AAN IDCT
336 slice_count integer
338 ec flags (decoding,video)
339 Set error concealment strategy.
340 Possible values:
342 guess_mvs
344 iterative motion vector (MV) search (slow)
345 deblock
347 use strong deblock filter for damaged MBs
348 favor_inter
350 favor predicting from the previous frame instead of the current
351 bits_per_coded_sample integer
353 aspect rational number (encoding,video)
354 Set sample aspect ratio.
355 sar rational number (encoding,video)
357 Set sample aspect ratio. Alias to aspect.
358 debug flags (decoding/encoding,audio,video,subtitles)
360 Print specific debug info.
361 Possible values:
363 pict
365 picture info
366 rc rate control
368 bitstream
370 mb_type
371 macroblock (MB) type
372 qp per-block quantization parameter (QP)
374 dct_coeff
376 green_metadata
377 display complexity metadata for the upcoming frame, GoP or for
378 a given duration.
379 skip
381 startcode
382 er error recognition
383 mmco
385 memory management control operations (H.264)
386 bugs
388 buffers
389 picture buffer allocations
390 thread_ops
392 threading operations
393 nomc
395 skip motion compensation
396 cmp integer (encoding,video)
398 Set full pel me compare function.
399 Possible values:
401 sad sum of absolute differences, fast (default)
403 sse sum of squared errors
405 satd
407 sum of absolute Hadamard transformed differences
408 dct sum of absolute DCT transformed differences
410 psnr
412 sum of squared quantization errors (avoid, low quality)
413 bit number of bits needed for the block
415 rd rate distortion optimal, slow
417 zero
419 0
420 vsad
422 sum of absolute vertical differences
423 vsse
425 sum of squared vertical differences
426 nsse
428 noise preserving sum of squared differences
429 w53 5/3 wavelet, only used in snow
431 w97 9/7 wavelet, only used in snow
433 dctmax
435 chroma
436 subcmp integer (encoding,video)
437 Set sub pel me compare function.
438 Possible values:
440 sad sum of absolute differences, fast (default)
442 sse sum of squared errors
444 satd
446 sum of absolute Hadamard transformed differences
447 dct sum of absolute DCT transformed differences
449 psnr
451 sum of squared quantization errors (avoid, low quality)
452 bit number of bits needed for the block
454 rd rate distortion optimal, slow
456 zero
458 0
459 vsad
461 sum of absolute vertical differences
462 vsse
464 sum of squared vertical differences
465 nsse
467 noise preserving sum of squared differences
468 w53 5/3 wavelet, only used in snow
470 w97 9/7 wavelet, only used in snow
472 dctmax
474 chroma
475 mbcmp integer (encoding,video)
476 Set macroblock compare function.
477 Possible values:
479 sad sum of absolute differences, fast (default)
481 sse sum of squared errors
483 satd
485 sum of absolute Hadamard transformed differences
486 dct sum of absolute DCT transformed differences
488 psnr
490 sum of squared quantization errors (avoid, low quality)
491 bit number of bits needed for the block
493 rd rate distortion optimal, slow
495 zero
497 0
498 vsad
500 sum of absolute vertical differences
501 vsse
503 sum of squared vertical differences
504 nsse
506 noise preserving sum of squared differences
507 w53 5/3 wavelet, only used in snow
509 w97 9/7 wavelet, only used in snow
511 dctmax
513 chroma
514 ildctcmp integer (encoding,video)
515 Set interlaced dct compare function.
516 Possible values:
518 sad sum of absolute differences, fast (default)
520 sse sum of squared errors
522 satd
524 sum of absolute Hadamard transformed differences
525 dct sum of absolute DCT transformed differences
527 psnr
529 sum of squared quantization errors (avoid, low quality)
530 bit number of bits needed for the block
532 rd rate distortion optimal, slow
534 zero
536 0
537 vsad
539 sum of absolute vertical differences
540 vsse
542 sum of squared vertical differences
543 nsse
545 noise preserving sum of squared differences
546 w53 5/3 wavelet, only used in snow
548 w97 9/7 wavelet, only used in snow
550 dctmax
552 chroma
553 dia_size integer (encoding,video)
554 Set diamond type & size for motion estimation.
555 (1024, INT_MAX)
557 full motion estimation(slowest)
558 (768, 1024]
560 umh motion estimation
561 (512, 768]
563 hex motion estimation
564 (256, 512]
566 l2s diamond motion estimation
567 [2,256]
569 var diamond motion estimation
570 (-1, 2)
572 small diamond motion estimation
573 -1 funny diamond motion estimation
575 (INT_MIN, -1)
577 sab diamond motion estimation
578 last_pred integer (encoding,video)
580 Set amount of motion predictors from the previous frame.
581 precmp integer (encoding,video)
583 Set pre motion estimation compare function.
584 Possible values:
586 sad sum of absolute differences, fast (default)
588 sse sum of squared errors
590 satd
592 sum of absolute Hadamard transformed differences
593 dct sum of absolute DCT transformed differences
595 psnr
597 sum of squared quantization errors (avoid, low quality)
598 bit number of bits needed for the block
600 rd rate distortion optimal, slow
602 zero
604 0
605 vsad
607 sum of absolute vertical differences
608 vsse
610 sum of squared vertical differences
611 nsse
613 noise preserving sum of squared differences
614 w53 5/3 wavelet, only used in snow
616 w97 9/7 wavelet, only used in snow
618 dctmax
620 chroma
621 pre_dia_size integer (encoding,video)
622 Set diamond type & size for motion estimation pre-pass.
623 subq integer (encoding,video)
625 Set sub pel motion estimation quality.
626 me_range integer (encoding,video)
628 Set limit motion vectors range (1023 for DivX player).
629 global_quality integer (encoding,audio,video)
631 slice_flags integer
632 mbd integer (encoding,video)
633 Set macroblock decision algorithm (high quality mode).
634 Possible values:
636 simple
638 use mbcmp (default)
639 bits
641 use fewest bits
642 rd use best rate distortion
644 rc_init_occupancy integer (encoding,video)
646 Set number of bits which should be loaded into the rc buffer before
647 decoding starts.
648 flags2 flags (decoding/encoding,audio,video,subtitles)
650 Possible values:
651 fast
653 Allow non spec compliant speedup tricks.
654 noout
656 Skip bitstream encoding.
657 ignorecrop
659 Ignore cropping information from sps.
660 local_header
662 Place global headers at every keyframe instead of in extradata.
663 chunks
665 Frame data might be split into multiple chunks.
666 showall
668 Show all frames before the first keyframe.
669 export_mvs
671 Export motion vectors into frame side-data (see
672 "AV_FRAME_DATA_MOTION_VECTORS") for codecs that support it. See
673 also doc/examples/export_mvs.c.
674 skip_manual
676 Do not skip samples and export skip information as frame side
677 data.
678 ass_ro_flush_noop
680 Do not reset ASS ReadOrder field on flush.
681 export_side_data flags (decoding/encoding,audio,video,subtitles)
683 Possible values:
684 mvs Export motion vectors into frame side-data (see
686 "AV_FRAME_DATA_MOTION_VECTORS") for codecs that support it. See
687 also doc/examples/export_mvs.c.
688 prft
690 Export encoder Producer Reference Time into packet side-data
691 (see "AV_PKT_DATA_PRFT") for codecs that support it.
692 venc_params
694 Export video encoding parameters through frame side data (see
695 "AV_FRAME_DATA_VIDEO_ENC_PARAMS") for codecs that support it.
696 At present, those are H.264 and VP9.
697 film_grain
699 Export film grain parameters through frame side data (see
700 "AV_FRAME_DATA_FILM_GRAIN_PARAMS"). Supported at present by
701 AV1 decoders.
702 threads integer (decoding/encoding,video)
704 Set the number of threads to be used, in case the selected codec
705 implementation supports multi-threading.
706 Possible values:
708 auto, 0
710 automatically select the number of threads to set
711 Default value is auto.
713 dc integer (encoding,video)
715 Set intra_dc_precision.
716 nssew integer (encoding,video)
718 Set nsse weight.
719 skip_top integer (decoding,video)
721 Set number of macroblock rows at the top which are skipped.
722 skip_bottom integer (decoding,video)
724 Set number of macroblock rows at the bottom which are skipped.
725 profile integer (encoding,audio,video)
727 Set encoder codec profile. Default value is unknown. Encoder
728 specific profiles are documented in the relevant encoder
729 documentation.
730 level integer (encoding,audio,video)
732 Possible values:
733 unknown
735 lowres integer (decoding,audio,video)
736 Decode at 1= 1/2, 2=1/4, 3=1/8 resolutions.
737 mblmin integer (encoding,video)
739 Set min macroblock lagrange factor (VBR).
740 mblmax integer (encoding,video)
742 Set max macroblock lagrange factor (VBR).
743 skip_loop_filter integer (decoding,video)
745 skip_idct integer (decoding,video)
746 skip_frame integer (decoding,video)
747 Make decoder discard processing depending on the frame type
748 selected by the option value.
749 skip_loop_filter skips frame loop filtering, skip_idct skips frame
751 IDCT/dequantization, skip_frame skips decoding.
752 Possible values:
754 none
756 Discard no frame.
757 default
759 Discard useless frames like 0-sized frames.
760 noref
762 Discard all non-reference frames.
763 bidir
765 Discard all bidirectional frames.
766 nokey
768 Discard all frames excepts keyframes.
769 nointra
771 Discard all frames except I frames.
772 all Discard all frames.
774 Default value is default.
776 bidir_refine integer (encoding,video)
778 Refine the two motion vectors used in bidirectional macroblocks.
779 keyint_min integer (encoding,video)
781 Set minimum interval between IDR-frames.
782 refs integer (encoding,video)
784 Set reference frames to consider for motion compensation.
785 trellis integer (encoding,audio,video)
787 Set rate-distortion optimal quantization.
788 mv0_threshold integer (encoding,video)
790 compression_level integer (encoding,audio,video)
791 bits_per_raw_sample integer
792 channel_layout integer (decoding/encoding,audio)
793 Possible values:
794 request_channel_layout integer (decoding,audio)
796 Possible values:
797 rc_max_vbv_use float (encoding,video)
799 rc_min_vbv_use float (encoding,video)
800 ticks_per_frame integer (decoding/encoding,audio,video)
801 color_primaries integer (decoding/encoding,video)
802 Possible values:
803 bt709
805 BT.709
806 bt470m
808 BT.470 M
809 bt470bg
811 BT.470 BG
812 smpte170m
814 SMPTE 170 M
815 smpte240m
817 SMPTE 240 M
818 film
820 Film
821 bt2020
823 BT.2020
824 smpte428
826 smpte428_1
827 SMPTE ST 428-1
828 smpte431
830 SMPTE 431-2
831 smpte432
833 SMPTE 432-1
834 jedec-p22
836 JEDEC P22
837 color_trc integer (decoding/encoding,video)
839 Possible values:
840 bt709
842 BT.709
843 gamma22
845 BT.470 M
846 gamma28
848 BT.470 BG
849 smpte170m
851 SMPTE 170 M
852 smpte240m
854 SMPTE 240 M
855 linear
857 Linear
858 log
860 log100
861 Log
862 log_sqrt
864 log316
865 Log square root
866 iec61966_2_4
868 iec61966-2-4
869 IEC 61966-2-4
870 bt1361
872 bt1361e
873 BT.1361
874 iec61966_2_1
876 iec61966-2-1
877 IEC 61966-2-1
878 bt2020_10
880 bt2020_10bit
881 BT.2020 - 10 bit
882 bt2020_12
884 bt2020_12bit
885 BT.2020 - 12 bit
886 smpte2084
888 SMPTE ST 2084
889 smpte428
891 smpte428_1
892 SMPTE ST 428-1
893 arib-std-b67
895 ARIB STD-B67
896 colorspace integer (decoding/encoding,video)
898 Possible values:
899 rgb RGB
901 bt709
903 BT.709
904 fcc FCC
906 bt470bg
908 BT.470 BG
909 smpte170m
911 SMPTE 170 M
912 smpte240m
914 SMPTE 240 M
915 ycocg
917 YCOCG
918 bt2020nc
920 bt2020_ncl
921 BT.2020 NCL
922 bt2020c
924 bt2020_cl
925 BT.2020 CL
926 smpte2085
928 SMPTE 2085
929 chroma-derived-nc
931 Chroma-derived NCL
932 chroma-derived-c
934 Chroma-derived CL
935 ictcp
937 ICtCp
938 color_range integer (decoding/encoding,video)
940 If used as input parameter, it serves as a hint to the decoder,
941 which color_range the input has. Possible values:
942 tv
944 mpeg
945 MPEG (219*2^(n-8))
946 pc
948 jpeg
949 JPEG (2^n-1)
950 chroma_sample_location integer (decoding/encoding,video)
952 Possible values:
953 left
955 center
956 topleft
957 top
958 bottomleft
959 bottom
960 log_level_offset integer
961 Set the log level offset.
962 slices integer (encoding,video)
964 Number of slices, used in parallelized encoding.
965 thread_type flags (decoding/encoding,video)
967 Select which multithreading methods to use.
968 Use of frame will increase decoding delay by one frame per thread,
970 so clients which cannot provide future frames should not use it.
971 Possible values:
973 slice
975 Decode more than one part of a single frame at once.
976 Multithreading using slices works only when the video was
978 encoded with slices.
979 frame
981 Decode more than one frame at once.
982 Default value is slice+frame.
984 audio_service_type integer (encoding,audio)
986 Set audio service type.
987 Possible values:
989 ma Main Audio Service
991 ef Effects
993 vi Visually Impaired
995 hi Hearing Impaired
997 di Dialogue
999 co Commentary
1001 em Emergency
1003 vo Voice Over
1005 ka Karaoke
1007 request_sample_fmt sample_fmt (decoding,audio)
1009 Set sample format audio decoders should prefer. Default value is
1010 "none".
1011 pkt_timebase rational number
1013 sub_charenc encoding (decoding,subtitles)
1014 Set the input subtitles character encoding.
1015 field_order field_order (video)
1017 Set/override the field order of the video. Possible values:
1018 progressive
1020 Progressive video
1021 tt Interlaced video, top field coded and displayed first
1023 bb Interlaced video, bottom field coded and displayed first
1025 tb Interlaced video, top coded first, bottom displayed first
1027 bt Interlaced video, bottom coded first, top displayed first
1029 skip_alpha bool (decoding,video)
1031 Set to 1 to disable processing alpha (transparency). This works
1032 like the gray flag in the flags option which skips chroma
1033 information instead of alpha. Default is 0.
1034 codec_whitelist list (input)
1036 "," separated list of allowed decoders. By default all are allowed.
1037 dump_separator string (input)
1039 Separator used to separate the fields printed on the command line
1040 about the Stream parameters. For example, to separate the fields
1041 with newlines and indentation:
1042 ffprobe -dump_separator "
1044 " -i ~/videos/matrixbench_mpeg2.mpg
1045 max_pixels integer (decoding/encoding,video)
1047 Maximum number of pixels per image. This value can be used to avoid
1048 out of memory failures due to large images.
1049 apply_cropping bool (decoding,video)
1051 Enable cropping if cropping parameters are multiples of the
1052 required alignment for the left and top parameters. If the
1053 alignment is not met the cropping will be partially applied to
1054 maintain alignment. Default is 1 (enabled). Note: The required
1055 alignment depends on if "AV_CODEC_FLAG_UNALIGNED" is set and the
1056 CPU. "AV_CODEC_FLAG_UNALIGNED" cannot be changed from the command
1057 line. Also hardware decoders will not apply left/top Cropping.
1058
1060 Decoders are configured elements in FFmpeg which allow the decoding of
1061 multimedia streams.
1062 When you configure your FFmpeg build, all the supported native decoders
1064 are enabled by default. Decoders requiring an external library must be
1065 enabled manually via the corresponding "--enable-lib" option. You can
1066 list all available decoders using the configure option
1067 "--list-decoders".
1068 You can disable all the decoders with the configure option
1070 "--disable-decoders" and selectively enable / disable single decoders
1071 with the options "--enable-decoder=DECODER" /
1072 "--disable-decoder=DECODER".
1073 The option "-decoders" of the ff* tools will display the list of
1075 enabled decoders.
1076
1078 A description of some of the currently available video decoders
1079 follows.
1080 av1
1082 AOMedia Video 1 (AV1) decoder.
1083 Options
1085 operating_point
1087 Select an operating point of a scalable AV1 bitstream (0 - 31).
1088 Default is 0.
1089 rawvideo
1091 Raw video decoder.
1092 This decoder decodes rawvideo streams.
1094 Options
1096 top top_field_first
1098 Specify the assumed field type of the input video.
1099 -1 the video is assumed to be progressive (default)
1101 0 bottom-field-first is assumed
1103 1 top-field-first is assumed
1105 libdav1d
1107 dav1d AV1 decoder.
1108 libdav1d allows libavcodec to decode the AOMedia Video 1 (AV1) codec.
1110 Requires the presence of the libdav1d headers and library during
1111 configuration. You need to explicitly configure the build with
1112 "--enable-libdav1d".
1113 Options
1115 The following options are supported by the libdav1d wrapper.
1117 framethreads
1119 Set amount of frame threads to use during decoding. The default
1120 value is 0 (autodetect). This option is deprecated for libdav1d >=
1121 1.0 and will be removed in the future. Use the global option
1122 "threads" instead.
1123 tilethreads
1125 Set amount of tile threads to use during decoding. The default
1126 value is 0 (autodetect). This option is deprecated for libdav1d >=
1127 1.0 and will be removed in the future. Use the global option
1128 "threads" instead.
1129 filmgrain
1131 Apply film grain to the decoded video if present in the bitstream.
1132 Defaults to the internal default of the library. This option is
1133 deprecated and will be removed in the future. See the global option
1134 "export_side_data" to export Film Grain parameters instead of
1135 applying it.
1136 oppoint
1138 Select an operating point of a scalable AV1 bitstream (0 - 31).
1139 Defaults to the internal default of the library.
1140 alllayers
1142 Output all spatial layers of a scalable AV1 bitstream. The default
1143 value is false.
1144 libdavs2
1146 AVS2-P2/IEEE1857.4 video decoder wrapper.
1147 This decoder allows libavcodec to decode AVS2 streams with davs2
1149 library.
1150 libuavs3d
1152 AVS3-P2/IEEE1857.10 video decoder.
1153 libuavs3d allows libavcodec to decode AVS3 streams. Requires the
1155 presence of the libuavs3d headers and library during configuration.
1156 You need to explicitly configure the build with "--enable-libuavs3d".
1157 Options
1159 The following option is supported by the libuavs3d wrapper.
1161 frame_threads
1163 Set amount of frame threads to use during decoding. The default
1164 value is 0 (autodetect).
1165
1167 A description of some of the currently available audio decoders
1168 follows.
1169 ac3
1171 AC-3 audio decoder.
1172 This decoder implements part of ATSC A/52:2010 and ETSI TS 102 366, as
1174 well as the undocumented RealAudio 3 (a.k.a. dnet).
1175 AC-3 Decoder Options
1177 -drc_scale value
1179 Dynamic Range Scale Factor. The factor to apply to dynamic range
1180 values from the AC-3 stream. This factor is applied exponentially.
1181 The default value is 1. There are 3 notable scale factor ranges:
1182 drc_scale == 0
1184 DRC disabled. Produces full range audio.
1185 0 < drc_scale <= 1
1187 DRC enabled. Applies a fraction of the stream DRC value.
1188 Audio reproduction is between full range and full compression.
1189 drc_scale > 1
1191 DRC enabled. Applies drc_scale asymmetrically. Loud sounds are
1192 fully compressed. Soft sounds are enhanced.
1193 flac
1195 FLAC audio decoder.
1196 This decoder aims to implement the complete FLAC specification from
1198 Xiph.
1199 FLAC Decoder options
1201 -use_buggy_lpc
1203 The lavc FLAC encoder used to produce buggy streams with high lpc
1204 values (like the default value). This option makes it possible to
1205 decode such streams correctly by using lavc's old buggy lpc logic
1206 for decoding.
1207 ffwavesynth
1209 Internal wave synthesizer.
1210 This decoder generates wave patterns according to predefined sequences.
1212 Its use is purely internal and the format of the data it accepts is not
1213 publicly documented.
1214 libcelt
1216 libcelt decoder wrapper.
1217 libcelt allows libavcodec to decode the Xiph CELT ultra-low delay audio
1219 codec. Requires the presence of the libcelt headers and library during
1220 configuration. You need to explicitly configure the build with
1221 "--enable-libcelt".
1222 libgsm
1224 libgsm decoder wrapper.
1225 libgsm allows libavcodec to decode the GSM full rate audio codec.
1227 Requires the presence of the libgsm headers and library during
1228 configuration. You need to explicitly configure the build with
1229 "--enable-libgsm".
1230 This decoder supports both the ordinary GSM and the Microsoft variant.
1232 libilbc
1234 libilbc decoder wrapper.
1235 libilbc allows libavcodec to decode the Internet Low Bitrate Codec
1237 (iLBC) audio codec. Requires the presence of the libilbc headers and
1238 library during configuration. You need to explicitly configure the
1239 build with "--enable-libilbc".
1240 Options
1242 The following option is supported by the libilbc wrapper.
1244 enhance
1246 Enable the enhancement of the decoded audio when set to 1. The
1247 default value is 0 (disabled).
1248 libopencore-amrnb
1250 libopencore-amrnb decoder wrapper.
1251 libopencore-amrnb allows libavcodec to decode the Adaptive Multi-Rate
1253 Narrowband audio codec. Using it requires the presence of the
1254 libopencore-amrnb headers and library during configuration. You need to
1255 explicitly configure the build with "--enable-libopencore-amrnb".
1256 An FFmpeg native decoder for AMR-NB exists, so users can decode AMR-NB
1258 without this library.
1259 libopencore-amrwb
1261 libopencore-amrwb decoder wrapper.
1262 libopencore-amrwb allows libavcodec to decode the Adaptive Multi-Rate
1264 Wideband audio codec. Using it requires the presence of the
1265 libopencore-amrwb headers and library during configuration. You need to
1266 explicitly configure the build with "--enable-libopencore-amrwb".
1267 An FFmpeg native decoder for AMR-WB exists, so users can decode AMR-WB
1269 without this library.
1270 libopus
1272 libopus decoder wrapper.
1273 libopus allows libavcodec to decode the Opus Interactive Audio Codec.
1275 Requires the presence of the libopus headers and library during
1276 configuration. You need to explicitly configure the build with
1277 "--enable-libopus".
1278 An FFmpeg native decoder for Opus exists, so users can decode Opus
1280 without this library.
1281
1283 libaribb24
1284 ARIB STD-B24 caption decoder.
1285 Implements profiles A and C of the ARIB STD-B24 standard.
1287 libaribb24 Decoder Options
1289 -aribb24-base-path path
1291 Sets the base path for the libaribb24 library. This is utilized for
1292 reading of configuration files (for custom unicode conversions),
1293 and for dumping of non-text symbols as images under that location.
1294 Unset by default.
1296 -aribb24-skip-ruby-text boolean
1298 Tells the decoder wrapper to skip text blocks that contain half-
1299 height ruby text.
1300 Enabled by default.
1302 dvbsub
1304 Options
1305 compute_clut
1307 -2 Compute clut once if no matching CLUT is in the stream.
1308 -1 Compute clut if no matching CLUT is in the stream.
1310 0 Never compute CLUT
1312 1 Always compute CLUT and override the one provided in the
1314 stream.
1315 dvb_substream
1317 Selects the dvb substream, or all substreams if -1 which is
1318 default.
1319 dvdsub
1321 This codec decodes the bitmap subtitles used in DVDs; the same
1322 subtitles can also be found in VobSub file pairs and in some Matroska
1323 files.
1324 Options
1326 palette
1328 Specify the global palette used by the bitmaps. When stored in
1329 VobSub, the palette is normally specified in the index file; in
1330 Matroska, the palette is stored in the codec extra-data in the same
1331 format as in VobSub. In DVDs, the palette is stored in the IFO
1332 file, and therefore not available when reading from dumped VOB
1333 files.
1334 The format for this option is a string containing 16 24-bits
1336 hexadecimal numbers (without 0x prefix) separated by commas, for
1337 example "0d00ee, ee450d, 101010, eaeaea, 0ce60b, ec14ed, ebff0b,
1338 0d617a, 7b7b7b, d1d1d1, 7b2a0e, 0d950c, 0f007b, cf0dec, cfa80c,
1339 7c127b".
1340 ifo_palette
1342 Specify the IFO file from which the global palette is obtained.
1343 (experimental)
1344 forced_subs_only
1346 Only decode subtitle entries marked as forced. Some titles have
1347 forced and non-forced subtitles in the same track. Setting this
1348 flag to 1 will only keep the forced subtitles. Default value is 0.
1349 libzvbi-teletext
1351 Libzvbi allows libavcodec to decode DVB teletext pages and DVB teletext
1352 subtitles. Requires the presence of the libzvbi headers and library
1353 during configuration. You need to explicitly configure the build with
1354 "--enable-libzvbi".
1355 Options
1357 txt_page
1359 List of teletext page numbers to decode. Pages that do not match
1360 the specified list are dropped. You may use the special "*" string
1361 to match all pages, or "subtitle" to match all subtitle pages.
1362 Default value is *.
1363 txt_default_region
1365 Set default character set used for decoding, a value between 0 and
1366 87 (see ETS 300 706, Section 15, Table 32). Default value is -1,
1367 which does not override the libzvbi default. This option is needed
1368 for some legacy level 1.0 transmissions which cannot signal the
1369 proper charset.
1370 txt_chop_top
1372 Discards the top teletext line. Default value is 1.
1373 txt_format
1375 Specifies the format of the decoded subtitles.
1376 bitmap
1378 The default format, you should use this for teletext pages,
1379 because certain graphics and colors cannot be expressed in
1380 simple text or even ASS.
1381 text
1383 Simple text based output without formatting.
1384 ass Formatted ASS output, subtitle pages and teletext pages are
1386 returned in different styles, subtitle pages are stripped down
1387 to text, but an effort is made to keep the text alignment and
1388 the formatting.
1389 txt_left
1391 X offset of generated bitmaps, default is 0.
1392 txt_top
1394 Y offset of generated bitmaps, default is 0.
1395 txt_chop_spaces
1397 Chops leading and trailing spaces and removes empty lines from the
1398 generated text. This option is useful for teletext based subtitles
1399 where empty spaces may be present at the start or at the end of the
1400 lines or empty lines may be present between the subtitle lines
1401 because of double-sized teletext characters. Default value is 1.
1402 txt_duration
1404 Sets the display duration of the decoded teletext pages or
1405 subtitles in milliseconds. Default value is -1 which means infinity
1406 or until the next subtitle event comes.
1407 txt_transparent
1409 Force transparent background of the generated teletext bitmaps.
1410 Default value is 0 which means an opaque background.
1411 txt_opacity
1413 Sets the opacity (0-255) of the teletext background. If
1414 txt_transparent is not set, it only affects characters between a
1415 start box and an end box, typically subtitles. Default value is 0
1416 if txt_transparent is set, 255 otherwise.
1417
1419 Encoders are configured elements in FFmpeg which allow the encoding of
1420 multimedia streams.
1421 When you configure your FFmpeg build, all the supported native encoders
1423 are enabled by default. Encoders requiring an external library must be
1424 enabled manually via the corresponding "--enable-lib" option. You can
1425 list all available encoders using the configure option
1426 "--list-encoders".
1427 You can disable all the encoders with the configure option
1429 "--disable-encoders" and selectively enable / disable single encoders
1430 with the options "--enable-encoder=ENCODER" /
1431 "--disable-encoder=ENCODER".
1432 The option "-encoders" of the ff* tools will display the list of
1434 enabled encoders.
1435
1437 A description of some of the currently available audio encoders
1438 follows.
1439 aac
1441 Advanced Audio Coding (AAC) encoder.
1442 This encoder is the default AAC encoder, natively implemented into
1444 FFmpeg.
1445 Options
1447 b Set bit rate in bits/s. Setting this automatically activates
1449 constant bit rate (CBR) mode. If this option is unspecified it is
1450 set to 128kbps.
1451 q Set quality for variable bit rate (VBR) mode. This option is valid
1453 only using the ffmpeg command-line tool. For library interface
1454 users, use global_quality.
1455 cutoff
1457 Set cutoff frequency. If unspecified will allow the encoder to
1458 dynamically adjust the cutoff to improve clarity on low bitrates.
1459 aac_coder
1461 Set AAC encoder coding method. Possible values:
1462 twoloop
1464 Two loop searching (TLS) method. This is the default method.
1465 This method first sets quantizers depending on band thresholds
1467 and then tries to find an optimal combination by adding or
1468 subtracting a specific value from all quantizers and adjusting
1469 some individual quantizer a little. Will tune itself based on
1470 whether aac_is, aac_ms and aac_pns are enabled.
1471 anmr
1473 Average noise to mask ratio (ANMR) trellis-based solution.
1474 This is an experimental coder which currently produces a lower
1476 quality, is more unstable and is slower than the default
1477 twoloop coder but has potential. Currently has no support for
1478 the aac_is or aac_pns options. Not currently recommended.
1479 fast
1481 Constant quantizer method.
1482 Uses a cheaper version of twoloop algorithm that doesn't try to
1484 do as many clever adjustments. Worse with low bitrates (less
1485 than 64kbps), but is better and much faster at higher bitrates.
1486 aac_ms
1488 Sets mid/side coding mode. The default value of "auto" will
1489 automatically use M/S with bands which will benefit from such
1490 coding. Can be forced for all bands using the value "enable", which
1491 is mainly useful for debugging or disabled using "disable".
1492 aac_is
1494 Sets intensity stereo coding tool usage. By default, it's enabled
1495 and will automatically toggle IS for similar pairs of stereo bands
1496 if it's beneficial. Can be disabled for debugging by setting the
1497 value to "disable".
1498 aac_pns
1500 Uses perceptual noise substitution to replace low entropy high
1501 frequency bands with imperceptible white noise during the decoding
1502 process. By default, it's enabled, but can be disabled for
1503 debugging purposes by using "disable".
1504 aac_tns
1506 Enables the use of a multitap FIR filter which spans through the
1507 high frequency bands to hide quantization noise during the encoding
1508 process and is reverted by the decoder. As well as decreasing
1509 unpleasant artifacts in the high range this also reduces the
1510 entropy in the high bands and allows for more bits to be used by
1511 the mid-low bands. By default it's enabled but can be disabled for
1512 debugging by setting the option to "disable".
1513 aac_ltp
1515 Enables the use of the long term prediction extension which
1516 increases coding efficiency in very low bandwidth situations such
1517 as encoding of voice or solo piano music by extending constant
1518 harmonic peaks in bands throughout frames. This option is implied
1519 by profile:a aac_low and is incompatible with aac_pred. Use in
1520 conjunction with -ar to decrease the samplerate.
1521 aac_pred
1523 Enables the use of a more traditional style of prediction where the
1524 spectral coefficients transmitted are replaced by the difference of
1525 the current coefficients minus the previous "predicted"
1526 coefficients. In theory and sometimes in practice this can improve
1527 quality for low to mid bitrate audio. This option implies the
1528 aac_main profile and is incompatible with aac_ltp.
1529 profile
1531 Sets the encoding profile, possible values:
1532 aac_low
1534 The default, AAC "Low-complexity" profile. Is the most
1535 compatible and produces decent quality.
1536 mpeg2_aac_low
1538 Equivalent to "-profile:a aac_low -aac_pns 0". PNS was
1539 introduced with the MPEG4 specifications.
1540 aac_ltp
1542 Long term prediction profile, is enabled by and will enable the
1543 aac_ltp option. Introduced in MPEG4.
1544 aac_main
1546 Main-type prediction profile, is enabled by and will enable the
1547 aac_pred option. Introduced in MPEG2.
1548 If this option is unspecified it is set to aac_low.
1550 ac3 and ac3_fixed
1552 AC-3 audio encoders.
1553 These encoders implement part of ATSC A/52:2010 and ETSI TS 102 366, as
1555 well as the undocumented RealAudio 3 (a.k.a. dnet).
1556 The ac3 encoder uses floating-point math, while the ac3_fixed encoder
1558 only uses fixed-point integer math. This does not mean that one is
1559 always faster, just that one or the other may be better suited to a
1560 particular system. The ac3_fixed encoder is not the default codec for
1561 any of the output formats, so it must be specified explicitly using the
1562 option "-acodec ac3_fixed" in order to use it.
1563 AC-3 Metadata
1565 The AC-3 metadata options are used to set parameters that describe the
1567 audio, but in most cases do not affect the audio encoding itself. Some
1568 of the options do directly affect or influence the decoding and
1569 playback of the resulting bitstream, while others are just for
1570 informational purposes. A few of the options will add bits to the
1571 output stream that could otherwise be used for audio data, and will
1572 thus affect the quality of the output. Those will be indicated
1573 accordingly with a note in the option list below.
1574 These parameters are described in detail in several publicly-available
1576 documents.
1577 *<<http://www.atsc.org/cms/standards/a_52-2010.pdf>>
1579 *<<http://www.atsc.org/cms/standards/a_54a_with_corr_1.pdf>>
1580 *<<http://www.dolby.com/uploadedFiles/zz-_Shared_Assets/English_PDFs/Professional/18_Metadata.Guide.pdf>>
1581 *<<http://www.dolby.com/uploadedFiles/zz-_Shared_Assets/English_PDFs/Professional/46_DDEncodingGuidelines.pdf>>
1582 Metadata Control Options
1584 -per_frame_metadata boolean
1586 Allow Per-Frame Metadata. Specifies if the encoder should check for
1587 changing metadata for each frame.
1588 0 The metadata values set at initialization will be used for
1590 every frame in the stream. (default)
1591 1 Metadata values can be changed before encoding each frame.
1593 Downmix Levels
1595 -center_mixlev level
1597 Center Mix Level. The amount of gain the decoder should apply to
1598 the center channel when downmixing to stereo. This field will only
1599 be written to the bitstream if a center channel is present. The
1600 value is specified as a scale factor. There are 3 valid values:
1601 0.707
1603 Apply -3dB gain
1604 0.595
1606 Apply -4.5dB gain (default)
1607 0.500
1609 Apply -6dB gain
1610 -surround_mixlev level
1612 Surround Mix Level. The amount of gain the decoder should apply to
1613 the surround channel(s) when downmixing to stereo. This field will
1614 only be written to the bitstream if one or more surround channels
1615 are present. The value is specified as a scale factor. There are 3
1616 valid values:
1617 0.707
1619 Apply -3dB gain
1620 0.500
1622 Apply -6dB gain (default)
1623 0.000
1625 Silence Surround Channel(s)
1626 Audio Production Information
1628 Audio Production Information is optional information describing the
1630 mixing environment. Either none or both of the fields are written to
1631 the bitstream.
1632 -mixing_level number
1634 Mixing Level. Specifies peak sound pressure level (SPL) in the
1635 production environment when the mix was mastered. Valid values are
1636 80 to 111, or -1 for unknown or not indicated. The default value is
1637 -1, but that value cannot be used if the Audio Production
1638 Information is written to the bitstream. Therefore, if the
1639 "room_type" option is not the default value, the "mixing_level"
1640 option must not be -1.
1641 -room_type type
1643 Room Type. Describes the equalization used during the final mixing
1644 session at the studio or on the dubbing stage. A large room is a
1645 dubbing stage with the industry standard X-curve equalization; a
1646 small room has flat equalization. This field will not be written
1647 to the bitstream if both the "mixing_level" option and the
1648 "room_type" option have the default values.
1649 0
1651 notindicated
1652 Not Indicated (default)
1653 1
1655 large
1656 Large Room
1657 2
1659 small
1660 Small Room
1661 Other Metadata Options
1663 -copyright boolean
1665 Copyright Indicator. Specifies whether a copyright exists for this
1666 audio.
1667 0
1669 off No Copyright Exists (default)
1670 1
1672 on Copyright Exists
1673 -dialnorm value
1675 Dialogue Normalization. Indicates how far the average dialogue
1676 level of the program is below digital 100% full scale (0 dBFS).
1677 This parameter determines a level shift during audio reproduction
1678 that sets the average volume of the dialogue to a preset level. The
1679 goal is to match volume level between program sources. A value of
1680 -31dB will result in no volume level change, relative to the source
1681 volume, during audio reproduction. Valid values are whole numbers
1682 in the range -31 to -1, with -31 being the default.
1683 -dsur_mode mode
1685 Dolby Surround Mode. Specifies whether the stereo signal uses Dolby
1686 Surround (Pro Logic). This field will only be written to the
1687 bitstream if the audio stream is stereo. Using this option does NOT
1688 mean the encoder will actually apply Dolby Surround processing.
1689 0
1691 notindicated
1692 Not Indicated (default)
1693 1
1695 off Not Dolby Surround Encoded
1696 2
1698 on Dolby Surround Encoded
1699 -original boolean
1701 Original Bit Stream Indicator. Specifies whether this audio is from
1702 the original source and not a copy.
1703 0
1705 off Not Original Source
1706 1
1708 on Original Source (default)
1709 Extended Bitstream Information
1711 The extended bitstream options are part of the Alternate Bit Stream
1713 Syntax as specified in Annex D of the A/52:2010 standard. It is grouped
1714 into 2 parts. If any one parameter in a group is specified, all values
1715 in that group will be written to the bitstream. Default values are
1716 used for those that are written but have not been specified. If the
1717 mixing levels are written, the decoder will use these values instead of
1718 the ones specified in the "center_mixlev" and "surround_mixlev" options
1719 if it supports the Alternate Bit Stream Syntax.
1720 Extended Bitstream Information - Part 1
1722 -dmix_mode mode
1724 Preferred Stereo Downmix Mode. Allows the user to select either
1725 Lt/Rt (Dolby Surround) or Lo/Ro (normal stereo) as the preferred
1726 stereo downmix mode.
1727 0
1729 notindicated
1730 Not Indicated (default)
1731 1
1733 ltrt
1734 Lt/Rt Downmix Preferred
1735 2
1737 loro
1738 Lo/Ro Downmix Preferred
1739 -ltrt_cmixlev level
1741 Lt/Rt Center Mix Level. The amount of gain the decoder should apply
1742 to the center channel when downmixing to stereo in Lt/Rt mode.
1743 1.414
1745 Apply +3dB gain
1746 1.189
1748 Apply +1.5dB gain
1749 1.000
1751 Apply 0dB gain
1752 0.841
1754 Apply -1.5dB gain
1755 0.707
1757 Apply -3.0dB gain
1758 0.595
1760 Apply -4.5dB gain (default)
1761 0.500
1763 Apply -6.0dB gain
1764 0.000
1766 Silence Center Channel
1767 -ltrt_surmixlev level
1769 Lt/Rt Surround Mix Level. The amount of gain the decoder should
1770 apply to the surround channel(s) when downmixing to stereo in Lt/Rt
1771 mode.
1772 0.841
1774 Apply -1.5dB gain
1775 0.707
1777 Apply -3.0dB gain
1778 0.595
1780 Apply -4.5dB gain
1781 0.500
1783 Apply -6.0dB gain (default)
1784 0.000
1786 Silence Surround Channel(s)
1787 -loro_cmixlev level
1789 Lo/Ro Center Mix Level. The amount of gain the decoder should apply
1790 to the center channel when downmixing to stereo in Lo/Ro mode.
1791 1.414
1793 Apply +3dB gain
1794 1.189
1796 Apply +1.5dB gain
1797 1.000
1799 Apply 0dB gain
1800 0.841
1802 Apply -1.5dB gain
1803 0.707
1805 Apply -3.0dB gain
1806 0.595
1808 Apply -4.5dB gain (default)
1809 0.500
1811 Apply -6.0dB gain
1812 0.000
1814 Silence Center Channel
1815 -loro_surmixlev level
1817 Lo/Ro Surround Mix Level. The amount of gain the decoder should
1818 apply to the surround channel(s) when downmixing to stereo in Lo/Ro
1819 mode.
1820 0.841
1822 Apply -1.5dB gain
1823 0.707
1825 Apply -3.0dB gain
1826 0.595
1828 Apply -4.5dB gain
1829 0.500
1831 Apply -6.0dB gain (default)
1832 0.000
1834 Silence Surround Channel(s)
1835 Extended Bitstream Information - Part 2
1837 -dsurex_mode mode
1839 Dolby Surround EX Mode. Indicates whether the stream uses Dolby
1840 Surround EX (7.1 matrixed to 5.1). Using this option does NOT mean
1841 the encoder will actually apply Dolby Surround EX processing.
1842 0
1844 notindicated
1845 Not Indicated (default)
1846 1
1848 on Dolby Surround EX Off
1849 2
1851 off Dolby Surround EX On
1852 -dheadphone_mode mode
1854 Dolby Headphone Mode. Indicates whether the stream uses Dolby
1855 Headphone encoding (multi-channel matrixed to 2.0 for use with
1856 headphones). Using this option does NOT mean the encoder will
1857 actually apply Dolby Headphone processing.
1858 0
1860 notindicated
1861 Not Indicated (default)
1862 1
1864 on Dolby Headphone Off
1865 2
1867 off Dolby Headphone On
1868 -ad_conv_type type
1870 A/D Converter Type. Indicates whether the audio has passed through
1871 HDCD A/D conversion.
1872 0
1874 standard
1875 Standard A/D Converter (default)
1876 1
1878 hdcd
1879 HDCD A/D Converter
1880 Other AC-3 Encoding Options
1882 -stereo_rematrixing boolean
1884 Stereo Rematrixing. Enables/Disables use of rematrixing for stereo
1885 input. This is an optional AC-3 feature that increases quality by
1886 selectively encoding the left/right channels as mid/side. This
1887 option is enabled by default, and it is highly recommended that it
1888 be left as enabled except for testing purposes.
1889 cutoff frequency
1891 Set lowpass cutoff frequency. If unspecified, the encoder selects a
1892 default determined by various other encoding parameters.
1893 Floating-Point-Only AC-3 Encoding Options
1895 These options are only valid for the floating-point encoder and do not
1897 exist for the fixed-point encoder due to the corresponding features not
1898 being implemented in fixed-point.
1899 -channel_coupling boolean
1901 Enables/Disables use of channel coupling, which is an optional AC-3
1902 feature that increases quality by combining high frequency
1903 information from multiple channels into a single channel. The per-
1904 channel high frequency information is sent with less accuracy in
1905 both the frequency and time domains. This allows more bits to be
1906 used for lower frequencies while preserving enough information to
1907 reconstruct the high frequencies. This option is enabled by default
1908 for the floating-point encoder and should generally be left as
1909 enabled except for testing purposes or to increase encoding speed.
1910 -1
1912 auto
1913 Selected by Encoder (default)
1914 0
1916 off Disable Channel Coupling
1917 1
1919 on Enable Channel Coupling
1920 -cpl_start_band number
1922 Coupling Start Band. Sets the channel coupling start band, from 1
1923 to 15. If a value higher than the bandwidth is used, it will be
1924 reduced to 1 less than the coupling end band. If auto is used, the
1925 start band will be determined by the encoder based on the bit rate,
1926 sample rate, and channel layout. This option has no effect if
1927 channel coupling is disabled.
1928 -1
1930 auto
1931 Selected by Encoder (default)
1932 flac
1934 FLAC (Free Lossless Audio Codec) Encoder
1935 Options
1937 The following options are supported by FFmpeg's flac encoder.
1939 compression_level
1941 Sets the compression level, which chooses defaults for many other
1942 options if they are not set explicitly. Valid values are from 0 to
1943 12, 5 is the default.
1944 frame_size
1946 Sets the size of the frames in samples per channel.
1947 lpc_coeff_precision
1949 Sets the LPC coefficient precision, valid values are from 1 to 15,
1950 15 is the default.
1951 lpc_type
1953 Sets the first stage LPC algorithm
1954 none
1956 LPC is not used
1957 fixed
1959 fixed LPC coefficients
1960 levinson
1962 cholesky
1963 lpc_passes
1964 Number of passes to use for Cholesky factorization during LPC
1965 analysis
1966 min_partition_order
1968 The minimum partition order
1969 max_partition_order
1971 The maximum partition order
1972 prediction_order_method
1974 estimation
1975 2level
1976 4level
1977 8level
1978 search
1979 Bruteforce search
1980 log
1982 ch_mode
1983 Channel mode
1984 auto
1986 The mode is chosen automatically for each frame
1987 indep
1989 Channels are independently coded
1990 left_side
1992 right_side
1993 mid_side
1994 exact_rice_parameters
1995 Chooses if rice parameters are calculated exactly or approximately.
1996 if set to 1 then they are chosen exactly, which slows the code down
1997 slightly and improves compression slightly.
1998 multi_dim_quant
2000 Multi Dimensional Quantization. If set to 1 then a 2nd stage LPC
2001 algorithm is applied after the first stage to finetune the
2002 coefficients. This is quite slow and slightly improves compression.
2003 opus
2005 Opus encoder.
2006 This is a native FFmpeg encoder for the Opus format. Currently its in
2008 development and only implements the CELT part of the codec. Its quality
2009 is usually worse and at best is equal to the libopus encoder.
2010 Options
2012 b Set bit rate in bits/s. If unspecified it uses the number of
2014 channels and the layout to make a good guess.
2015 opus_delay
2017 Sets the maximum delay in milliseconds. Lower delays than 20ms will
2018 very quickly decrease quality.
2019 libfdk_aac
2021 libfdk-aac AAC (Advanced Audio Coding) encoder wrapper.
2022 The libfdk-aac library is based on the Fraunhofer FDK AAC code from the
2024 Android project.
2025 Requires the presence of the libfdk-aac headers and library during
2027 configuration. You need to explicitly configure the build with
2028 "--enable-libfdk-aac". The library is also incompatible with GPL, so if
2029 you allow the use of GPL, you should configure with "--enable-gpl
2030 --enable-nonfree --enable-libfdk-aac".
2031 This encoder has support for the AAC-HE profiles.
2033 VBR encoding, enabled through the vbr or flags +qscale options, is
2035 experimental and only works with some combinations of parameters.
2036 Support for encoding 7.1 audio is only available with libfdk-aac 0.1.3
2038 or higher.
2039 For more information see the fdk-aac project at
2041 <http://sourceforge.net/p/opencore-amr/fdk-aac/>.
2042 Options
2044 The following options are mapped on the shared FFmpeg codec options.
2046 b Set bit rate in bits/s. If the bitrate is not explicitly specified,
2048 it is automatically set to a suitable value depending on the
2049 selected profile.
2050 In case VBR mode is enabled the option is ignored.
2052 ar Set audio sampling rate (in Hz).
2054 channels
2056 Set the number of audio channels.
2057 flags +qscale
2059 Enable fixed quality, VBR (Variable Bit Rate) mode. Note that VBR
2060 is implicitly enabled when the vbr value is positive.
2061 cutoff
2063 Set cutoff frequency. If not specified (or explicitly set to 0) it
2064 will use a value automatically computed by the library. Default
2065 value is 0.
2066 profile
2068 Set audio profile.
2069 The following profiles are recognized:
2071 aac_low
2073 Low Complexity AAC (LC)
2074 aac_he
2076 High Efficiency AAC (HE-AAC)
2077 aac_he_v2
2079 High Efficiency AAC version 2 (HE-AACv2)
2080 aac_ld
2082 Low Delay AAC (LD)
2083 aac_eld
2085 Enhanced Low Delay AAC (ELD)
2086 If not specified it is set to aac_low.
2088 The following are private options of the libfdk_aac encoder.
2090 afterburner
2092 Enable afterburner feature if set to 1, disabled if set to 0. This
2093 improves the quality but also the required processing power.
2094 Default value is 1.
2096 eld_sbr
2098 Enable SBR (Spectral Band Replication) for ELD if set to 1,
2099 disabled if set to 0.
2100 Default value is 0.
2102 eld_v2
2104 Enable ELDv2 (LD-MPS extension for ELD stereo signals) for ELDv2 if
2105 set to 1, disabled if set to 0.
2106 Note that option is available when fdk-aac version
2108 (AACENCODER_LIB_VL0.AACENCODER_LIB_VL1.AACENCODER_LIB_VL2) >
2109 (4.0.0).
2110 Default value is 0.
2112 signaling
2114 Set SBR/PS signaling style.
2115 It can assume one of the following values:
2117 default
2119 choose signaling implicitly (explicit hierarchical by default,
2120 implicit if global header is disabled)
2121 implicit
2123 implicit backwards compatible signaling
2124 explicit_sbr
2126 explicit SBR, implicit PS signaling
2127 explicit_hierarchical
2129 explicit hierarchical signaling
2130 Default value is default.
2132 latm
2134 Output LATM/LOAS encapsulated data if set to 1, disabled if set to
2135 0.
2136 Default value is 0.
2138 header_period
2140 Set StreamMuxConfig and PCE repetition period (in frames) for
2141 sending in-band configuration buffers within LATM/LOAS transport
2142 layer.
2143 Must be a 16-bits non-negative integer.
2145 Default value is 0.
2147 vbr Set VBR mode, from 1 to 5. 1 is lowest quality (though still pretty
2149 good) and 5 is highest quality. A value of 0 will disable VBR, and
2150 CBR (Constant Bit Rate) is enabled.
2151 Currently only the aac_low profile supports VBR encoding.
2153 VBR modes 1-5 correspond to roughly the following average bit
2155 rates:
2156 1 32 kbps/channel
2158 2 40 kbps/channel
2160 3 48-56 kbps/channel
2162 4 64 kbps/channel
2164 5 about 80-96 kbps/channel
2166 Default value is 0.
2168 Examples
2170 • Use ffmpeg to convert an audio file to VBR AAC in an M4A (MP4)
2172 container:
2173 ffmpeg -i input.wav -codec:a libfdk_aac -vbr 3 output.m4a
2175 • Use ffmpeg to convert an audio file to CBR 64k kbps AAC, using the
2177 High-Efficiency AAC profile:
2178 ffmpeg -i input.wav -c:a libfdk_aac -profile:a aac_he -b:a 64k output.m4a
2180 libmp3lame
2182 LAME (Lame Ain't an MP3 Encoder) MP3 encoder wrapper.
2183 Requires the presence of the libmp3lame headers and library during
2185 configuration. You need to explicitly configure the build with
2186 "--enable-libmp3lame".
2187 See libshine for a fixed-point MP3 encoder, although with a lower
2189 quality.
2190 Options
2192 The following options are supported by the libmp3lame wrapper. The
2194 lame-equivalent of the options are listed in parentheses.
2195 b (-b)
2197 Set bitrate expressed in bits/s for CBR or ABR. LAME "bitrate" is
2198 expressed in kilobits/s.
2199 q (-V)
2201 Set constant quality setting for VBR. This option is valid only
2202 using the ffmpeg command-line tool. For library interface users,
2203 use global_quality.
2204 compression_level (-q)
2206 Set algorithm quality. Valid arguments are integers in the 0-9
2207 range, with 0 meaning highest quality but slowest, and 9 meaning
2208 fastest while producing the worst quality.
2209 cutoff (--lowpass)
2211 Set lowpass cutoff frequency. If unspecified, the encoder
2212 dynamically adjusts the cutoff.
2213 reservoir
2215 Enable use of bit reservoir when set to 1. Default value is 1. LAME
2216 has this enabled by default, but can be overridden by use --nores
2217 option.
2218 joint_stereo (-m j)
2220 Enable the encoder to use (on a frame by frame basis) either L/R
2221 stereo or mid/side stereo. Default value is 1.
2222 abr (--abr)
2224 Enable the encoder to use ABR when set to 1. The lame --abr sets
2225 the target bitrate, while this options only tells FFmpeg to use ABR
2226 still relies on b to set bitrate.
2227 libopencore-amrnb
2229 OpenCORE Adaptive Multi-Rate Narrowband encoder.
2230 Requires the presence of the libopencore-amrnb headers and library
2232 during configuration. You need to explicitly configure the build with
2233 "--enable-libopencore-amrnb --enable-version3".
2234 This is a mono-only encoder. Officially it only supports 8000Hz sample
2236 rate, but you can override it by setting strict to unofficial or lower.
2237 Options
2239 b Set bitrate in bits per second. Only the following bitrates are
2241 supported, otherwise libavcodec will round to the nearest valid
2242 bitrate.
2243 4750
2245 5150
2246 5900
2247 6700
2248 7400
2249 7950
2250 10200
2251 12200
2252 dtx Allow discontinuous transmission (generate comfort noise) when set
2253 to 1. The default value is 0 (disabled).
2254 libopus
2256 libopus Opus Interactive Audio Codec encoder wrapper.
2257 Requires the presence of the libopus headers and library during
2259 configuration. You need to explicitly configure the build with
2260 "--enable-libopus".
2261 Option Mapping
2263 Most libopus options are modelled after the opusenc utility from opus-
2265 tools. The following is an option mapping chart describing options
2266 supported by the libopus wrapper, and their opusenc-equivalent in
2267 parentheses.
2268 b (bitrate)
2270 Set the bit rate in bits/s. FFmpeg's b option is expressed in
2271 bits/s, while opusenc's bitrate in kilobits/s.
2272 vbr (vbr, hard-cbr, and cvbr)
2274 Set VBR mode. The FFmpeg vbr option has the following valid
2275 arguments, with the opusenc equivalent options in parentheses:
2276 off (hard-cbr)
2278 Use constant bit rate encoding.
2279 on (vbr)
2281 Use variable bit rate encoding (the default).
2282 constrained (cvbr)
2284 Use constrained variable bit rate encoding.
2285 compression_level (comp)
2287 Set encoding algorithm complexity. Valid options are integers in
2288 the 0-10 range. 0 gives the fastest encodes but lower quality,
2289 while 10 gives the highest quality but slowest encoding. The
2290 default is 10.
2291 frame_duration (framesize)
2293 Set maximum frame size, or duration of a frame in milliseconds. The
2294 argument must be exactly the following: 2.5, 5, 10, 20, 40, 60.
2295 Smaller frame sizes achieve lower latency but less quality at a
2296 given bitrate. Sizes greater than 20ms are only interesting at
2297 fairly low bitrates. The default is 20ms.
2298 packet_loss (expect-loss)
2300 Set expected packet loss percentage. The default is 0.
2301 fec (n/a)
2303 Enable inband forward error correction. packet_loss must be non-
2304 zero to take advantage - frequency of FEC 'side-data' is
2305 proportional to expected packet loss. Default is disabled.
2306 application (N.A.)
2308 Set intended application type. Valid options are listed below:
2309 voip
2311 Favor improved speech intelligibility.
2312 audio
2314 Favor faithfulness to the input (the default).
2315 lowdelay
2317 Restrict to only the lowest delay modes.
2318 cutoff (N.A.)
2320 Set cutoff bandwidth in Hz. The argument must be exactly one of the
2321 following: 4000, 6000, 8000, 12000, or 20000, corresponding to
2322 narrowband, mediumband, wideband, super wideband, and fullband
2323 respectively. The default is 0 (cutoff disabled).
2324 mapping_family (mapping_family)
2326 Set channel mapping family to be used by the encoder. The default
2327 value of -1 uses mapping family 0 for mono and stereo inputs, and
2328 mapping family 1 otherwise. The default also disables the surround
2329 masking and LFE bandwidth optimzations in libopus, and requires
2330 that the input contains 8 channels or fewer.
2331 Other values include 0 for mono and stereo, 1 for surround sound
2333 with masking and LFE bandwidth optimizations, and 255 for
2334 independent streams with an unspecified channel layout.
2335 apply_phase_inv (N.A.) (requires libopus >= 1.2)
2337 If set to 0, disables the use of phase inversion for intensity
2338 stereo, improving the quality of mono downmixes, but slightly
2339 reducing normal stereo quality. The default is 1 (phase inversion
2340 enabled).
2341 libshine
2343 Shine Fixed-Point MP3 encoder wrapper.
2344 Shine is a fixed-point MP3 encoder. It has a far better performance on
2346 platforms without an FPU, e.g. armel CPUs, and some phones and tablets.
2347 However, as it is more targeted on performance than quality, it is not
2348 on par with LAME and other production-grade encoders quality-wise.
2349 Also, according to the project's homepage, this encoder may not be free
2350 of bugs as the code was written a long time ago and the project was
2351 dead for at least 5 years.
2352 This encoder only supports stereo and mono input. This is also CBR-
2354 only.
2355 The original project (last updated in early 2007) is at
2357 <http://sourceforge.net/projects/libshine-fxp/>. We only support the
2358 updated fork by the Savonet/Liquidsoap project at
2359 <https://github.com/savonet/shine>.
2360 Requires the presence of the libshine headers and library during
2362 configuration. You need to explicitly configure the build with
2363 "--enable-libshine".
2364 See also libmp3lame.
2366 Options
2368 The following options are supported by the libshine wrapper. The
2370 shineenc-equivalent of the options are listed in parentheses.
2371 b (-b)
2373 Set bitrate expressed in bits/s for CBR. shineenc -b option is
2374 expressed in kilobits/s.
2375 libtwolame
2377 TwoLAME MP2 encoder wrapper.
2378 Requires the presence of the libtwolame headers and library during
2380 configuration. You need to explicitly configure the build with
2381 "--enable-libtwolame".
2382 Options
2384 The following options are supported by the libtwolame wrapper. The
2386 twolame-equivalent options follow the FFmpeg ones and are in
2387 parentheses.
2388 b (-b)
2390 Set bitrate expressed in bits/s for CBR. twolame b option is
2391 expressed in kilobits/s. Default value is 128k.
2392 q (-V)
2394 Set quality for experimental VBR support. Maximum value range is
2395 from -50 to 50, useful range is from -10 to 10. The higher the
2396 value, the better the quality. This option is valid only using the
2397 ffmpeg command-line tool. For library interface users, use
2398 global_quality.
2399 mode (--mode)
2401 Set the mode of the resulting audio. Possible values:
2402 auto
2404 Choose mode automatically based on the input. This is the
2405 default.
2406 stereo
2408 Stereo
2409 joint_stereo
2411 Joint stereo
2412 dual_channel
2414 Dual channel
2415 mono
2417 Mono
2418 psymodel (--psyc-mode)
2420 Set psychoacoustic model to use in encoding. The argument must be
2421 an integer between -1 and 4, inclusive. The higher the value, the
2422 better the quality. The default value is 3.
2423 energy_levels (--energy)
2425 Enable energy levels extensions when set to 1. The default value is
2426 0 (disabled).
2427 error_protection (--protect)
2429 Enable CRC error protection when set to 1. The default value is 0
2430 (disabled).
2431 copyright (--copyright)
2433 Set MPEG audio copyright flag when set to 1. The default value is 0
2434 (disabled).
2435 original (--original)
2437 Set MPEG audio original flag when set to 1. The default value is 0
2438 (disabled).
2439 libvo-amrwbenc
2441 VisualOn Adaptive Multi-Rate Wideband encoder.
2442 Requires the presence of the libvo-amrwbenc headers and library during
2444 configuration. You need to explicitly configure the build with
2445 "--enable-libvo-amrwbenc --enable-version3".
2446 This is a mono-only encoder. Officially it only supports 16000Hz sample
2448 rate, but you can override it by setting strict to unofficial or lower.
2449 Options
2451 b Set bitrate in bits/s. Only the following bitrates are supported,
2453 otherwise libavcodec will round to the nearest valid bitrate.
2454 6600
2456 8850
2457 12650
2458 14250
2459 15850
2460 18250
2461 19850
2462 23050
2463 23850
2464 dtx Allow discontinuous transmission (generate comfort noise) when set
2465 to 1. The default value is 0 (disabled).
2466 libvorbis
2468 libvorbis encoder wrapper.
2469 Requires the presence of the libvorbisenc headers and library during
2471 configuration. You need to explicitly configure the build with
2472 "--enable-libvorbis".
2473 Options
2475 The following options are supported by the libvorbis wrapper. The
2477 oggenc-equivalent of the options are listed in parentheses.
2478 To get a more accurate and extensive documentation of the libvorbis
2480 options, consult the libvorbisenc's and oggenc's documentations. See
2481 <http://xiph.org/vorbis/>, <http://wiki.xiph.org/Vorbis-tools>, and
2482 oggenc(1).
2483 b (-b)
2485 Set bitrate expressed in bits/s for ABR. oggenc -b is expressed in
2486 kilobits/s.
2487 q (-q)
2489 Set constant quality setting for VBR. The value should be a float
2490 number in the range of -1.0 to 10.0. The higher the value, the
2491 better the quality. The default value is 3.0.
2492 This option is valid only using the ffmpeg command-line tool. For
2494 library interface users, use global_quality.
2495 cutoff (--advanced-encode-option lowpass_frequency=N)
2497 Set cutoff bandwidth in Hz, a value of 0 disables cutoff. oggenc's
2498 related option is expressed in kHz. The default value is 0 (cutoff
2499 disabled).
2500 minrate (-m)
2502 Set minimum bitrate expressed in bits/s. oggenc -m is expressed in
2503 kilobits/s.
2504 maxrate (-M)
2506 Set maximum bitrate expressed in bits/s. oggenc -M is expressed in
2507 kilobits/s. This only has effect on ABR mode.
2508 iblock (--advanced-encode-option impulse_noisetune=N)
2510 Set noise floor bias for impulse blocks. The value is a float
2511 number from -15.0 to 0.0. A negative bias instructs the encoder to
2512 pay special attention to the crispness of transients in the encoded
2513 audio. The tradeoff for better transient response is a higher
2514 bitrate.
2515 mjpeg
2517 Motion JPEG encoder.
2518 Options
2520 huffman
2522 Set the huffman encoding strategy. Possible values:
2523 default
2525 Use the default huffman tables. This is the default strategy.
2526 optimal
2528 Compute and use optimal huffman tables.
2529 wavpack
2531 WavPack lossless audio encoder.
2532 Options
2534 The equivalent options for wavpack command line utility are listed in
2536 parentheses.
2537 Shared options
2539 The following shared options are effective for this encoder. Only
2541 special notes about this particular encoder will be documented here.
2542 For the general meaning of the options, see the Codec Options chapter.
2543 frame_size (--blocksize)
2545 For this encoder, the range for this option is between 128 and
2546 131072. Default is automatically decided based on sample rate and
2547 number of channel.
2548 For the complete formula of calculating default, see
2550 libavcodec/wavpackenc.c.
2551 compression_level (-f, -h, -hh, and -x)
2553 Private options
2555 joint_stereo (-j)
2557 Set whether to enable joint stereo. Valid values are:
2558 on (1)
2560 Force mid/side audio encoding.
2561 off (0)
2563 Force left/right audio encoding.
2564 auto
2566 Let the encoder decide automatically.
2567 optimize_mono
2569 Set whether to enable optimization for mono. This option is only
2570 effective for non-mono streams. Available values:
2571 on enabled
2573 off disabled
2575
2577 A description of some of the currently available video encoders
2578 follows.
2579 a64_multi, a64_multi5
2581 A64 / Commodore 64 multicolor charset encoder. "a64_multi5" is extended
2582 with 5th color (colram).
2583 GIF
2585 GIF image/animation encoder.
2586 Options
2588 gifflags integer
2590 Sets the flags used for GIF encoding.
2591 offsetting
2593 Enables picture offsetting.
2594 Default is enabled.
2596 transdiff
2598 Enables transparency detection between frames.
2599 Default is enabled.
2601 gifimage integer
2603 Enables encoding one full GIF image per frame, rather than an
2604 animated GIF.
2605 Default value is 0.
2607 global_palette integer
2609 Writes a palette to the global GIF header where feasible.
2610 If disabled, every frame will always have a palette written, even
2612 if there is a global palette supplied.
2613 Default value is 1.
2615 Hap
2617 Vidvox Hap video encoder.
2618 Options
2620 format integer
2622 Specifies the Hap format to encode.
2623 hap
2625 hap_alpha
2626 hap_q
2627 Default value is hap.
2629 chunks integer
2631 Specifies the number of chunks to split frames into, between 1 and
2632 64. This permits multithreaded decoding of large frames,
2633 potentially at the cost of data-rate. The encoder may modify this
2634 value to divide frames evenly.
2635 Default value is 1.
2637 compressor integer
2639 Specifies the second-stage compressor to use. If set to none,
2640 chunks will be limited to 1, as chunked uncompressed frames offer
2641 no benefit.
2642 none
2644 snappy
2645 Default value is snappy.
2647 jpeg2000
2649 The native jpeg 2000 encoder is lossy by default, the "-q:v" option can
2650 be used to set the encoding quality. Lossless encoding can be selected
2651 with "-pred 1".
2652 Options
2654 format integer
2656 Can be set to either "j2k" or "jp2" (the default) that makes it
2657 possible to store non-rgb pix_fmts.
2658 tile_width integer
2660 Sets tile width. Range is 1 to 1073741824. Default is 256.
2661 tile_height integer
2663 Sets tile height. Range is 1 to 1073741824. Default is 256.
2664 pred integer
2666 Allows setting the discrete wavelet transform (DWT) type
2667 dwt97int (Lossy)
2669 dwt53 (Lossless)
2670 Default is "dwt97int"
2672 sop boolean
2674 Enable this to add SOP marker at the start of each packet. Disabled
2675 by default.
2676 eph boolean
2678 Enable this to add EPH marker at the end of each packet header.
2679 Disabled by default.
2680 prog integer
2682 Sets the progression order to be used by the encoder. Possible
2683 values are:
2684 lrcp
2686 rlcp
2687 rpcl
2688 pcrl
2689 cprl
2690 Set to "lrcp" by default.
2692 layer_rates string
2694 By default, when this option is not used, compression is done using
2695 the quality metric. This option allows for compression using
2696 compression ratio. The compression ratio for each level could be
2697 specified. The compression ratio of a layer "l" species the what
2698 ratio of total file size is contained in the first "l" layers.
2699 Example usage:
2701 ffmpeg -i input.bmp -c:v jpeg2000 -layer_rates "100,10,1" output.j2k
2703 This would compress the image to contain 3 layers, where the data
2705 contained in the first layer would be compressed by 1000 times,
2706 compressed by 100 in the first two layers, and shall contain all
2707 data while using all 3 layers.
2708 librav1e
2710 rav1e AV1 encoder wrapper.
2711 Requires the presence of the rav1e headers and library during
2713 configuration. You need to explicitly configure the build with
2714 "--enable-librav1e".
2715 Options
2717 qmax
2719 Sets the maximum quantizer to use when using bitrate mode.
2720 qmin
2722 Sets the minimum quantizer to use when using bitrate mode.
2723 qp Uses quantizer mode to encode at the given quantizer (0-255).
2725 speed
2727 Selects the speed preset (0-10) to encode with.
2728 tiles
2730 Selects how many tiles to encode with.
2731 tile-rows
2733 Selects how many rows of tiles to encode with.
2734 tile-columns
2736 Selects how many columns of tiles to encode with.
2737 rav1e-params
2739 Set rav1e options using a list of key=value pairs separated by ":".
2740 See rav1e --help for a list of options.
2741 For example to specify librav1e encoding options with
2743 -rav1e-params:
2744 ffmpeg -i input -c:v librav1e -b:v 500K -rav1e-params speed=5:low_latency=true output.mp4
2746 libaom-av1
2748 libaom AV1 encoder wrapper.
2749 Requires the presence of the libaom headers and library during
2751 configuration. You need to explicitly configure the build with
2752 "--enable-libaom".
2753 Options
2755 The wrapper supports the following standard libavcodec options:
2757 b Set bitrate target in bits/second. By default this will use
2759 variable-bitrate mode. If maxrate and minrate are also set to the
2760 same value then it will use constant-bitrate mode, otherwise if crf
2761 is set as well then it will use constrained-quality mode.
2762 g keyint_min
2764 Set key frame placement. The GOP size sets the maximum distance
2765 between key frames; if zero the output stream will be intra-only.
2766 The minimum distance is ignored unless it is the same as the GOP
2767 size, in which case key frames will always appear at a fixed
2768 interval. Not set by default, so without this option the library
2769 has completely free choice about where to place key frames.
2770 qmin qmax
2772 Set minimum/maximum quantisation values. Valid range is from 0 to
2773 63 (warning: this does not match the quantiser values actually used
2774 by AV1 - divide by four to map real quantiser values to this
2775 range). Defaults to min/max (no constraint).
2776 minrate maxrate bufsize rc_init_occupancy
2778 Set rate control buffering parameters. Not used if not set -
2779 defaults to unconstrained variable bitrate.
2780 threads
2782 Set the number of threads to use while encoding. This may require
2783 the tiles or row-mt options to also be set to actually use the
2784 specified number of threads fully. Defaults to the number of
2785 hardware threads supported by the host machine.
2786 profile
2788 Set the encoding profile. Defaults to using the profile which
2789 matches the bit depth and chroma subsampling of the input.
2790 The wrapper also has some specific options:
2792 cpu-used
2794 Set the quality/encoding speed tradeoff. Valid range is from 0 to
2795 8, higher numbers indicating greater speed and lower quality. The
2796 default value is 1, which will be slow and high quality.
2797 auto-alt-ref
2799 Enable use of alternate reference frames. Defaults to the internal
2800 default of the library.
2801 arnr-max-frames (frames)
2803 Set altref noise reduction max frame count. Default is -1.
2804 arnr-strength (strength)
2806 Set altref noise reduction filter strength. Range is -1 to 6.
2807 Default is -1.
2808 aq-mode (aq-mode)
2810 Set adaptive quantization mode. Possible values:
2811 none (0)
2813 Disabled.
2814 variance (1)
2816 Variance-based.
2817 complexity (2)
2819 Complexity-based.
2820 cyclic (3)
2822 Cyclic refresh.
2823 tune (tune)
2825 Set the distortion metric the encoder is tuned with. Default is
2826 "psnr".
2827 psnr (0)
2829 ssim (1)
2830 lag-in-frames
2831 Set the maximum number of frames which the encoder may keep in
2832 flight at any one time for lookahead purposes. Defaults to the
2833 internal default of the library.
2834 error-resilience
2836 Enable error resilience features:
2837 default
2839 Improve resilience against losses of whole frames.
2840 Not enabled by default.
2842 crf Set the quality/size tradeoff for constant-quality (no bitrate
2844 target) and constrained-quality (with maximum bitrate target)
2845 modes. Valid range is 0 to 63, higher numbers indicating lower
2846 quality and smaller output size. Only used if set; by default only
2847 the bitrate target is used.
2848 static-thresh
2850 Set a change threshold on blocks below which they will be skipped
2851 by the encoder. Defined in arbitrary units as a nonnegative
2852 integer, defaulting to zero (no blocks are skipped).
2853 drop-threshold
2855 Set a threshold for dropping frames when close to rate control
2856 bounds. Defined as a percentage of the target buffer - when the
2857 rate control buffer falls below this percentage, frames will be
2858 dropped until it has refilled above the threshold. Defaults to
2859 zero (no frames are dropped).
2860 denoise-noise-level (level)
2862 Amount of noise to be removed for grain synthesis. Grain synthesis
2863 is disabled if this option is not set or set to 0.
2864 denoise-block-size (pixels)
2866 Block size used for denoising for grain synthesis. If not set, AV1
2867 codec uses the default value of 32.
2868 undershoot-pct (pct)
2870 Set datarate undershoot (min) percentage of the target bitrate.
2871 Range is -1 to 100. Default is -1.
2872 overshoot-pct (pct)
2874 Set datarate overshoot (max) percentage of the target bitrate.
2875 Range is -1 to 1000. Default is -1.
2876 minsection-pct (pct)
2878 Minimum percentage variation of the GOP bitrate from the target
2879 bitrate. If minsection-pct is not set, the libaomenc wrapper
2880 computes it as follows: "(minrate * 100 / bitrate)". Range is -1
2881 to 100. Default is -1 (unset).
2882 maxsection-pct (pct)
2884 Maximum percentage variation of the GOP bitrate from the target
2885 bitrate. If maxsection-pct is not set, the libaomenc wrapper
2886 computes it as follows: "(maxrate * 100 / bitrate)". Range is -1
2887 to 5000. Default is -1 (unset).
2888 frame-parallel (boolean)
2890 Enable frame parallel decodability features. Default is true.
2891 tiles
2893 Set the number of tiles to encode the input video with, as columns
2894 x rows. Larger numbers allow greater parallelism in both encoding
2895 and decoding, but may decrease coding efficiency. Defaults to the
2896 minimum number of tiles required by the size of the input video
2897 (this is 1x1 (that is, a single tile) for sizes up to and including
2898 4K).
2899 tile-columns tile-rows
2901 Set the number of tiles as log2 of the number of tile rows and
2902 columns. Provided for compatibility with libvpx/VP9.
2903 row-mt (Requires libaom >= 1.0.0-759-g90a15f4f2)
2905 Enable row based multi-threading. Disabled by default.
2906 enable-cdef (boolean)
2908 Enable Constrained Directional Enhancement Filter. The libaom-av1
2909 encoder enables CDEF by default.
2910 enable-restoration (boolean)
2912 Enable Loop Restoration Filter. Default is true for libaom-av1.
2913 enable-global-motion (boolean)
2915 Enable the use of global motion for block prediction. Default is
2916 true.
2917 enable-intrabc (boolean)
2919 Enable block copy mode for intra block prediction. This mode is
2920 useful for screen content. Default is true.
2921 enable-rect-partitions (boolean) (Requires libaom >= v2.0.0)
2923 Enable rectangular partitions. Default is true.
2924 enable-1to4-partitions (boolean) (Requires libaom >= v2.0.0)
2926 Enable 1:4/4:1 partitions. Default is true.
2927 enable-ab-partitions (boolean) (Requires libaom >= v2.0.0)
2929 Enable AB shape partitions. Default is true.
2930 enable-angle-delta (boolean) (Requires libaom >= v2.0.0)
2932 Enable angle delta intra prediction. Default is true.
2933 enable-cfl-intra (boolean) (Requires libaom >= v2.0.0)
2935 Enable chroma predicted from luma intra prediction. Default is
2936 true.
2937 enable-filter-intra (boolean) (Requires libaom >= v2.0.0)
2939 Enable filter intra predictor. Default is true.
2940 enable-intra-edge-filter (boolean) (Requires libaom >= v2.0.0)
2942 Enable intra edge filter. Default is true.
2943 enable-smooth-intra (boolean) (Requires libaom >= v2.0.0)
2945 Enable smooth intra prediction mode. Default is true.
2946 enable-paeth-intra (boolean) (Requires libaom >= v2.0.0)
2948 Enable paeth predictor in intra prediction. Default is true.
2949 enable-palette (boolean) (Requires libaom >= v2.0.0)
2951 Enable palette prediction mode. Default is true.
2952 enable-flip-idtx (boolean) (Requires libaom >= v2.0.0)
2954 Enable extended transform type, including FLIPADST_DCT,
2955 DCT_FLIPADST, FLIPADST_FLIPADST, ADST_FLIPADST, FLIPADST_ADST,
2956 IDTX, V_DCT, H_DCT, V_ADST, H_ADST, V_FLIPADST, H_FLIPADST. Default
2957 is true.
2958 enable-tx64 (boolean) (Requires libaom >= v2.0.0)
2960 Enable 64-pt transform. Default is true.
2961 reduced-tx-type-set (boolean) (Requires libaom >= v2.0.0)
2963 Use reduced set of transform types. Default is false.
2964 use-intra-dct-only (boolean) (Requires libaom >= v2.0.0)
2966 Use DCT only for INTRA modes. Default is false.
2967 use-inter-dct-only (boolean) (Requires libaom >= v2.0.0)
2969 Use DCT only for INTER modes. Default is false.
2970 use-intra-default-tx-only (boolean) (Requires libaom >= v2.0.0)
2972 Use Default-transform only for INTRA modes. Default is false.
2973 enable-ref-frame-mvs (boolean) (Requires libaom >= v2.0.0)
2975 Enable temporal mv prediction. Default is true.
2976 enable-reduced-reference-set (boolean) (Requires libaom >= v2.0.0)
2978 Use reduced set of single and compound references. Default is
2979 false.
2980 enable-obmc (boolean) (Requires libaom >= v2.0.0)
2982 Enable obmc. Default is true.
2983 enable-dual-filter (boolean) (Requires libaom >= v2.0.0)
2985 Enable dual filter. Default is true.
2986 enable-diff-wtd-comp (boolean) (Requires libaom >= v2.0.0)
2988 Enable difference-weighted compound. Default is true.
2989 enable-dist-wtd-comp (boolean) (Requires libaom >= v2.0.0)
2991 Enable distance-weighted compound. Default is true.
2992 enable-onesided-comp (boolean) (Requires libaom >= v2.0.0)
2994 Enable one sided compound. Default is true.
2995 enable-interinter-wedge (boolean) (Requires libaom >= v2.0.0)
2997 Enable interinter wedge compound. Default is true.
2998 enable-interintra-wedge (boolean) (Requires libaom >= v2.0.0)
3000 Enable interintra wedge compound. Default is true.
3001 enable-masked-comp (boolean) (Requires libaom >= v2.0.0)
3003 Enable masked compound. Default is true.
3004 enable-interintra-comp (boolean) (Requires libaom >= v2.0.0)
3006 Enable interintra compound. Default is true.
3007 enable-smooth-interintra (boolean) (Requires libaom >= v2.0.0)
3009 Enable smooth interintra mode. Default is true.
3010 aom-params
3012 Set libaom options using a list of key=value pairs separated by
3013 ":". For a list of supported options, see aomenc --help under the
3014 section "AV1 Specific Options".
3015 For example to specify libaom encoding options with -aom-params:
3017 ffmpeg -i input -c:v libaom-av1 -b:v 500K -aom-params tune=psnr:enable-tpl-model=1 output.mp4
3019 libsvtav1
3021 SVT-AV1 encoder wrapper.
3022 Requires the presence of the SVT-AV1 headers and library during
3024 configuration. You need to explicitly configure the build with
3025 "--enable-libsvtav1".
3026 Options
3028 profile
3030 Set the encoding profile.
3031 main
3033 high
3034 professional
3035 level
3036 Set the operating point level. For example: '4.0'
3037 hielevel
3039 Set the Hierarchical prediction levels.
3040 3level
3042 4level
3043 This is the default.
3044 tier
3046 Set the operating point tier.
3047 main
3049 This is the default.
3050 high
3052 rc Set the rate control mode to use.
3053 Possible modes:
3055 cqp Constant quantizer: use fixed values of qindex (dependent on
3057 the frame type) throughout the stream. This mode is the
3058 default.
3059 vbr Variable bitrate: use a target bitrate for the whole stream.
3061 cvbr
3063 Constrained variable bitrate: use a target bitrate for each
3064 GOP.
3065 qmax
3067 Set the maximum quantizer to use when using a bitrate mode.
3068 qmin
3070 Set the minimum quantizer to use when using a bitrate mode.
3071 qp Set the quantizer used in cqp rate control mode (0-63).
3073 sc_detection
3075 Enable scene change detection.
3076 la_depth
3078 Set number of frames to look ahead (0-120).
3079 preset
3081 Set the quality-speed tradeoff, in the range 0 to 8. Higher values
3082 are faster but lower quality. Defaults to 8 (highest speed).
3083 tile_rows
3085 Set log2 of the number of rows of tiles to use (0-6).
3086 tile_columns
3088 Set log2 of the number of columns of tiles to use (0-4).
3089 libkvazaar
3091 Kvazaar H.265/HEVC encoder.
3092 Requires the presence of the libkvazaar headers and library during
3094 configuration. You need to explicitly configure the build with
3095 --enable-libkvazaar.
3096 Options
3098 b Set target video bitrate in bit/s and enable rate control.
3100 kvazaar-params
3102 Set kvazaar parameters as a list of name=value pairs separated by
3103 commas (,). See kvazaar documentation for a list of options.
3104 libopenh264
3106 Cisco libopenh264 H.264/MPEG-4 AVC encoder wrapper.
3107 This encoder requires the presence of the libopenh264 headers and
3109 library during configuration. You need to explicitly configure the
3110 build with "--enable-libopenh264". The library is detected using pkg-
3111 config.
3112 For more information about the library see <http://www.openh264.org>.
3114 Options
3116 The following FFmpeg global options affect the configurations of the
3118 libopenh264 encoder.
3119 b Set the bitrate (as a number of bits per second).
3121 g Set the GOP size.
3123 maxrate
3125 Set the max bitrate (as a number of bits per second).
3126 flags +global_header
3128 Set global header in the bitstream.
3129 slices
3131 Set the number of slices, used in parallelized encoding. Default
3132 value is 0. This is only used when slice_mode is set to fixed.
3133 slice_mode
3135 Set slice mode. Can assume one of the following possible values:
3136 fixed
3138 a fixed number of slices
3139 rowmb
3141 one slice per row of macroblocks
3142 auto
3144 automatic number of slices according to number of threads
3145 dyn dynamic slicing
3147 Default value is auto.
3149 loopfilter
3151 Enable loop filter, if set to 1 (automatically enabled). To disable
3152 set a value of 0.
3153 profile
3155 Set profile restrictions. If set to the value of main enable CABAC
3156 (set the "SEncParamExt.iEntropyCodingModeFlag" flag to 1).
3157 max_nal_size
3159 Set maximum NAL size in bytes.
3160 allow_skip_frames
3162 Allow skipping frames to hit the target bitrate if set to 1.
3163 libtheora
3165 libtheora Theora encoder wrapper.
3166 Requires the presence of the libtheora headers and library during
3168 configuration. You need to explicitly configure the build with
3169 "--enable-libtheora".
3170 For more information about the libtheora project see
3172 <http://www.theora.org/>.
3173 Options
3175 The following global options are mapped to internal libtheora options
3177 which affect the quality and the bitrate of the encoded stream.
3178 b Set the video bitrate in bit/s for CBR (Constant Bit Rate) mode.
3180 In case VBR (Variable Bit Rate) mode is enabled this option is
3181 ignored.
3182 flags
3184 Used to enable constant quality mode (VBR) encoding through the
3185 qscale flag, and to enable the "pass1" and "pass2" modes.
3186 g Set the GOP size.
3188 global_quality
3190 Set the global quality as an integer in lambda units.
3191 Only relevant when VBR mode is enabled with "flags +qscale". The
3193 value is converted to QP units by dividing it by "FF_QP2LAMBDA",
3194 clipped in the [0 - 10] range, and then multiplied by 6.3 to get a
3195 value in the native libtheora range [0-63]. A higher value
3196 corresponds to a higher quality.
3197 q Enable VBR mode when set to a non-negative value, and set constant
3199 quality value as a double floating point value in QP units.
3200 The value is clipped in the [0-10] range, and then multiplied by
3202 6.3 to get a value in the native libtheora range [0-63].
3203 This option is valid only using the ffmpeg command-line tool. For
3205 library interface users, use global_quality.
3206 Examples
3208 • Set maximum constant quality (VBR) encoding with ffmpeg:
3210 ffmpeg -i INPUT -codec:v libtheora -q:v 10 OUTPUT.ogg
3212 • Use ffmpeg to convert a CBR 1000 kbps Theora video stream:
3214 ffmpeg -i INPUT -codec:v libtheora -b:v 1000k OUTPUT.ogg
3216 libvpx
3218 VP8/VP9 format supported through libvpx.
3219 Requires the presence of the libvpx headers and library during
3221 configuration. You need to explicitly configure the build with
3222 "--enable-libvpx".
3223 Options
3225 The following options are supported by the libvpx wrapper. The
3227 vpxenc-equivalent options or values are listed in parentheses for easy
3228 migration.
3229 To reduce the duplication of documentation, only the private options
3231 and some others requiring special attention are documented here. For
3232 the documentation of the undocumented generic options, see the Codec
3233 Options chapter.
3234 To get more documentation of the libvpx options, invoke the command
3236 ffmpeg -h encoder=libvpx, ffmpeg -h encoder=libvpx-vp9 or vpxenc
3237 --help. Further information is available in the libvpx API
3238 documentation.
3239 b (target-bitrate)
3241 Set bitrate in bits/s. Note that FFmpeg's b option is expressed in
3242 bits/s, while vpxenc's target-bitrate is in kilobits/s.
3243 g (kf-max-dist)
3245 keyint_min (kf-min-dist)
3246 qmin (min-q)
3247 qmax (max-q)
3248 bufsize (buf-sz, buf-optimal-sz)
3249 Set ratecontrol buffer size (in bits). Note vpxenc's options are
3250 specified in milliseconds, the libvpx wrapper converts this value
3251 as follows: "buf-sz = bufsize * 1000 / bitrate", "buf-optimal-sz =
3252 bufsize * 1000 / bitrate * 5 / 6".
3253 rc_init_occupancy (buf-initial-sz)
3255 Set number of bits which should be loaded into the rc buffer before
3256 decoding starts. Note vpxenc's option is specified in milliseconds,
3257 the libvpx wrapper converts this value as follows:
3258 "rc_init_occupancy * 1000 / bitrate".
3259 undershoot-pct
3261 Set datarate undershoot (min) percentage of the target bitrate.
3262 overshoot-pct
3264 Set datarate overshoot (max) percentage of the target bitrate.
3265 skip_threshold (drop-frame)
3267 qcomp (bias-pct)
3268 maxrate (maxsection-pct)
3269 Set GOP max bitrate in bits/s. Note vpxenc's option is specified as
3270 a percentage of the target bitrate, the libvpx wrapper converts
3271 this value as follows: "(maxrate * 100 / bitrate)".
3272 minrate (minsection-pct)
3274 Set GOP min bitrate in bits/s. Note vpxenc's option is specified as
3275 a percentage of the target bitrate, the libvpx wrapper converts
3276 this value as follows: "(minrate * 100 / bitrate)".
3277 minrate, maxrate, b end-usage=cbr
3279 "(minrate == maxrate == bitrate)".
3280 crf (end-usage=cq, cq-level)
3282 tune (tune)
3283 psnr (psnr)
3284 ssim (ssim)
3285 quality, deadline (deadline)
3286 best
3287 Use best quality deadline. Poorly named and quite slow, this
3288 option should be avoided as it may give worse quality output
3289 than good.
3290 good
3292 Use good quality deadline. This is a good trade-off between
3293 speed and quality when used with the cpu-used option.
3294 realtime
3296 Use realtime quality deadline.
3297 speed, cpu-used (cpu-used)
3299 Set quality/speed ratio modifier. Higher values speed up the encode
3300 at the cost of quality.
3301 nr (noise-sensitivity)
3303 static-thresh
3304 Set a change threshold on blocks below which they will be skipped
3305 by the encoder.
3306 slices (token-parts)
3308 Note that FFmpeg's slices option gives the total number of
3309 partitions, while vpxenc's token-parts is given as
3310 "log2(partitions)".
3311 max-intra-rate
3313 Set maximum I-frame bitrate as a percentage of the target bitrate.
3314 A value of 0 means unlimited.
3315 force_key_frames
3317 "VPX_EFLAG_FORCE_KF"
3318 Alternate reference frame related
3320 auto-alt-ref
3321 Enable use of alternate reference frames (2-pass only). Values
3322 greater than 1 enable multi-layer alternate reference frames
3323 (VP9 only).
3324 arnr-maxframes
3326 Set altref noise reduction max frame count.
3327 arnr-type
3329 Set altref noise reduction filter type: backward, forward,
3330 centered.
3331 arnr-strength
3333 Set altref noise reduction filter strength.
3334 rc-lookahead, lag-in-frames (lag-in-frames)
3336 Set number of frames to look ahead for frametype and
3337 ratecontrol.
3338 error-resilient
3340 Enable error resiliency features.
3341 sharpness integer
3343 Increase sharpness at the expense of lower PSNR. The valid range
3344 is [0, 7].
3345 ts-parameters
3347 Sets the temporal scalability configuration using a :-separated
3348 list of key=value pairs. For example, to specify temporal
3349 scalability parameters with "ffmpeg":
3350 ffmpeg -i INPUT -c:v libvpx -ts-parameters ts_number_layers=3:\
3352 ts_target_bitrate=250,500,1000:ts_rate_decimator=4,2,1:\
3353 ts_periodicity=4:ts_layer_id=0,2,1,2:ts_layering_mode=3 OUTPUT
3354 Below is a brief explanation of each of the parameters, please
3356 refer to "struct vpx_codec_enc_cfg" in "vpx/vpx_encoder.h" for more
3357 details.
3358 ts_number_layers
3360 Number of temporal coding layers.
3361 ts_target_bitrate
3363 Target bitrate for each temporal layer (in kbps). (bitrate
3364 should be inclusive of the lower temporal layer).
3365 ts_rate_decimator
3367 Frame rate decimation factor for each temporal layer.
3368 ts_periodicity
3370 Length of the sequence defining frame temporal layer
3371 membership.
3372 ts_layer_id
3374 Template defining the membership of frames to temporal layers.
3375 ts_layering_mode
3377 (optional) Selecting the temporal structure from a set of pre-
3378 defined temporal layering modes. Currently supports the
3379 following options.
3380 0 No temporal layering flags are provided internally, relies
3382 on flags being passed in using "metadata" field in
3383 "AVFrame" with following keys.
3384 vp8-flags
3386 Sets the flags passed into the encoder to indicate the
3387 referencing scheme for the current frame. Refer to
3388 function "vpx_codec_encode" in "vpx/vpx_encoder.h" for
3389 more details.
3390 temporal_id
3392 Explicitly sets the temporal id of the current frame to
3393 encode.
3394 2 Two temporal layers. 0-1...
3396 3 Three temporal layers. 0-2-1-2...; with single reference
3398 frame.
3399 4 Same as option "3", except there is a dependency between
3401 the two temporal layer 2 frames within the temporal period.
3402 VP9-specific options
3404 lossless
3405 Enable lossless mode.
3406 tile-columns
3408 Set number of tile columns to use. Note this is given as
3409 "log2(tile_columns)". For example, 8 tile columns would be
3410 requested by setting the tile-columns option to 3.
3411 tile-rows
3413 Set number of tile rows to use. Note this is given as
3414 "log2(tile_rows)". For example, 4 tile rows would be requested
3415 by setting the tile-rows option to 2.
3416 frame-parallel
3418 Enable frame parallel decodability features.
3419 aq-mode
3421 Set adaptive quantization mode (0: off (default), 1: variance
3422 2: complexity, 3: cyclic refresh, 4: equator360).
3423 colorspace color-space
3425 Set input color space. The VP9 bitstream supports signaling the
3426 following colorspaces:
3427 rgb sRGB
3429 bt709 bt709
3430 unspecified unknown
3431 bt470bg bt601
3432 smpte170m smpte170
3433 smpte240m smpte240
3434 bt2020_ncl bt2020
3435 row-mt boolean
3436 Enable row based multi-threading.
3437 tune-content
3439 Set content type: default (0), screen (1), film (2).
3440 corpus-complexity
3442 Corpus VBR mode is a variant of standard VBR where the
3443 complexity distribution midpoint is passed in rather than
3444 calculated for a specific clip or chunk.
3445 The valid range is [0, 10000]. 0 (default) uses standard VBR.
3447 enable-tpl boolean
3449 Enable temporal dependency model.
3450 ref-frame-config
3452 Using per-frame metadata, set members of the structure
3453 "vpx_svc_ref_frame_config_t" in "vpx/vp8cx.h" to fine-control
3454 referencing schemes and frame buffer management. Use a
3455 :-separated list of key=value pairs. For example,
3456 av_dict_set(&av_frame->metadata, "ref-frame-config", \
3458 "rfc_update_buffer_slot=7:rfc_lst_fb_idx=0:rfc_gld_fb_idx=1:rfc_alt_fb_idx=2:rfc_reference_last=0:rfc_reference_golden=0:rfc_reference_alt_ref=0");
3459 rfc_update_buffer_slot
3461 Indicates the buffer slot number to update
3462 rfc_update_last
3464 Indicates whether to update the LAST frame
3465 rfc_update_golden
3467 Indicates whether to update GOLDEN frame
3468 rfc_update_alt_ref
3470 Indicates whether to update ALT_REF frame
3471 rfc_lst_fb_idx
3473 LAST frame buffer index
3474 rfc_gld_fb_idx
3476 GOLDEN frame buffer index
3477 rfc_alt_fb_idx
3479 ALT_REF frame buffer index
3480 rfc_reference_last
3482 Indicates whether to reference LAST frame
3483 rfc_reference_golden
3485 Indicates whether to reference GOLDEN frame
3486 rfc_reference_alt_ref
3488 Indicates whether to reference ALT_REF frame
3489 rfc_reference_duration
3491 Indicates frame duration
3492 For more information about libvpx see: <http://www.webmproject.org/>
3494 libwebp
3496 libwebp WebP Image encoder wrapper
3497 libwebp is Google's official encoder for WebP images. It can encode in
3499 either lossy or lossless mode. Lossy images are essentially a wrapper
3500 around a VP8 frame. Lossless images are a separate codec developed by
3501 Google.
3502 Pixel Format
3504 Currently, libwebp only supports YUV420 for lossy and RGB for lossless
3506 due to limitations of the format and libwebp. Alpha is supported for
3507 either mode. Because of API limitations, if RGB is passed in when
3508 encoding lossy or YUV is passed in for encoding lossless, the pixel
3509 format will automatically be converted using functions from libwebp.
3510 This is not ideal and is done only for convenience.
3511 Options
3513 -lossless boolean
3515 Enables/Disables use of lossless mode. Default is 0.
3516 -compression_level integer
3518 For lossy, this is a quality/speed tradeoff. Higher values give
3519 better quality for a given size at the cost of increased encoding
3520 time. For lossless, this is a size/speed tradeoff. Higher values
3521 give smaller size at the cost of increased encoding time. More
3522 specifically, it controls the number of extra algorithms and
3523 compression tools used, and varies the combination of these tools.
3524 This maps to the method option in libwebp. The valid range is 0 to
3525 6. Default is 4.
3526 -qscale float
3528 For lossy encoding, this controls image quality, 0 to 100. For
3529 lossless encoding, this controls the effort and time spent at
3530 compressing more. The default value is 75. Note that for usage via
3531 libavcodec, this option is called global_quality and must be
3532 multiplied by FF_QP2LAMBDA.
3533 -preset type
3535 Configuration preset. This does some automatic settings based on
3536 the general type of the image.
3537 none
3539 Do not use a preset.
3540 default
3542 Use the encoder default.
3543 picture
3545 Digital picture, like portrait, inner shot
3546 photo
3548 Outdoor photograph, with natural lighting
3549 drawing
3551 Hand or line drawing, with high-contrast details
3552 icon
3554 Small-sized colorful images
3555 text
3557 Text-like
3558 libx264, libx264rgb
3560 x264 H.264/MPEG-4 AVC encoder wrapper.
3561 This encoder requires the presence of the libx264 headers and library
3563 during configuration. You need to explicitly configure the build with
3564 "--enable-libx264".
3565 libx264 supports an impressive number of features, including 8x8 and
3567 4x4 adaptive spatial transform, adaptive B-frame placement, CAVLC/CABAC
3568 entropy coding, interlacing (MBAFF), lossless mode, psy optimizations
3569 for detail retention (adaptive quantization, psy-RD, psy-trellis).
3570 Many libx264 encoder options are mapped to FFmpeg global codec options,
3572 while unique encoder options are provided through private options.
3573 Additionally the x264opts and x264-params private options allows one to
3574 pass a list of key=value tuples as accepted by the libx264
3575 "x264_param_parse" function.
3576 The x264 project website is at
3578 <http://www.videolan.org/developers/x264.html>.
3579 The libx264rgb encoder is the same as libx264, except it accepts packed
3581 RGB pixel formats as input instead of YUV.
3582 Supported Pixel Formats
3584 x264 supports 8- to 10-bit color spaces. The exact bit depth is
3586 controlled at x264's configure time.
3587 Options
3589 The following options are supported by the libx264 wrapper. The
3591 x264-equivalent options or values are listed in parentheses for easy
3592 migration.
3593 To reduce the duplication of documentation, only the private options
3595 and some others requiring special attention are documented here. For
3596 the documentation of the undocumented generic options, see the Codec
3597 Options chapter.
3598 To get a more accurate and extensive documentation of the libx264
3600 options, invoke the command x264 --fullhelp or consult the libx264
3601 documentation.
3602 b (bitrate)
3604 Set bitrate in bits/s. Note that FFmpeg's b option is expressed in
3605 bits/s, while x264's bitrate is in kilobits/s.
3606 bf (bframes)
3608 g (keyint)
3609 qmin (qpmin)
3610 Minimum quantizer scale.
3611 qmax (qpmax)
3613 Maximum quantizer scale.
3614 qdiff (qpstep)
3616 Maximum difference between quantizer scales.
3617 qblur (qblur)
3619 Quantizer curve blur
3620 qcomp (qcomp)
3622 Quantizer curve compression factor
3623 refs (ref)
3625 Number of reference frames each P-frame can use. The range is from
3626 0-16.
3627 sc_threshold (scenecut)
3629 Sets the threshold for the scene change detection.
3630 trellis (trellis)
3632 Performs Trellis quantization to increase efficiency. Enabled by
3633 default.
3634 nr (nr)
3636 me_range (merange)
3637 Maximum range of the motion search in pixels.
3638 me_method (me)
3640 Set motion estimation method. Possible values in the decreasing
3641 order of speed:
3642 dia (dia)
3644 epzs (dia)
3645 Diamond search with radius 1 (fastest). epzs is an alias for
3646 dia.
3647 hex (hex)
3649 Hexagonal search with radius 2.
3650 umh (umh)
3652 Uneven multi-hexagon search.
3653 esa (esa)
3655 Exhaustive search.
3656 tesa (tesa)
3658 Hadamard exhaustive search (slowest).
3659 forced-idr
3661 Normally, when forcing a I-frame type, the encoder can select any
3662 type of I-frame. This option forces it to choose an IDR-frame.
3663 subq (subme)
3665 Sub-pixel motion estimation method.
3666 b_strategy (b-adapt)
3668 Adaptive B-frame placement decision algorithm. Use only on first-
3669 pass.
3670 keyint_min (min-keyint)
3672 Minimum GOP size.
3673 coder
3675 Set entropy encoder. Possible values:
3676 ac Enable CABAC.
3678 vlc Enable CAVLC and disable CABAC. It generates the same effect as
3680 x264's --no-cabac option.
3681 cmp Set full pixel motion estimation comparison algorithm. Possible
3683 values:
3684 chroma
3686 Enable chroma in motion estimation.
3687 sad Ignore chroma in motion estimation. It generates the same
3689 effect as x264's --no-chroma-me option.
3690 threads (threads)
3692 Number of encoding threads.
3693 thread_type
3695 Set multithreading technique. Possible values:
3696 slice
3698 Slice-based multithreading. It generates the same effect as
3699 x264's --sliced-threads option.
3700 frame
3702 Frame-based multithreading.
3703 flags
3705 Set encoding flags. It can be used to disable closed GOP and enable
3706 open GOP by setting it to "-cgop". The result is similar to the
3707 behavior of x264's --open-gop option.
3708 rc_init_occupancy (vbv-init)
3710 preset (preset)
3711 Set the encoding preset.
3712 tune (tune)
3714 Set tuning of the encoding params.
3715 profile (profile)
3717 Set profile restrictions.
3718 fastfirstpass
3720 Enable fast settings when encoding first pass, when set to 1. When
3721 set to 0, it has the same effect of x264's --slow-firstpass option.
3722 crf (crf)
3724 Set the quality for constant quality mode.
3725 crf_max (crf-max)
3727 In CRF mode, prevents VBV from lowering quality beyond this point.
3728 qp (qp)
3730 Set constant quantization rate control method parameter.
3731 aq-mode (aq-mode)
3733 Set AQ method. Possible values:
3734 none (0)
3736 Disabled.
3737 variance (1)
3739 Variance AQ (complexity mask).
3740 autovariance (2)
3742 Auto-variance AQ (experimental).
3743 aq-strength (aq-strength)
3745 Set AQ strength, reduce blocking and blurring in flat and textured
3746 areas.
3747 psy Use psychovisual optimizations when set to 1. When set to 0, it has
3749 the same effect as x264's --no-psy option.
3750 psy-rd (psy-rd)
3752 Set strength of psychovisual optimization, in psy-rd:psy-trellis
3753 format.
3754 rc-lookahead (rc-lookahead)
3756 Set number of frames to look ahead for frametype and ratecontrol.
3757 weightb
3759 Enable weighted prediction for B-frames when set to 1. When set to
3760 0, it has the same effect as x264's --no-weightb option.
3761 weightp (weightp)
3763 Set weighted prediction method for P-frames. Possible values:
3764 none (0)
3766 Disabled
3767 simple (1)
3769 Enable only weighted refs
3770 smart (2)
3772 Enable both weighted refs and duplicates
3773 ssim (ssim)
3775 Enable calculation and printing SSIM stats after the encoding.
3776 intra-refresh (intra-refresh)
3778 Enable the use of Periodic Intra Refresh instead of IDR frames when
3779 set to 1.
3780 avcintra-class (class)
3782 Configure the encoder to generate AVC-Intra. Valid values are
3783 50,100 and 200
3784 bluray-compat (bluray-compat)
3786 Configure the encoder to be compatible with the bluray standard.
3787 It is a shorthand for setting "bluray-compat=1 force-cfr=1".
3788 b-bias (b-bias)
3790 Set the influence on how often B-frames are used.
3791 b-pyramid (b-pyramid)
3793 Set method for keeping of some B-frames as references. Possible
3794 values:
3795 none (none)
3797 Disabled.
3798 strict (strict)
3800 Strictly hierarchical pyramid.
3801 normal (normal)
3803 Non-strict (not Blu-ray compatible).
3804 mixed-refs
3806 Enable the use of one reference per partition, as opposed to one
3807 reference per macroblock when set to 1. When set to 0, it has the
3808 same effect as x264's --no-mixed-refs option.
3809 8x8dct
3811 Enable adaptive spatial transform (high profile 8x8 transform) when
3812 set to 1. When set to 0, it has the same effect as x264's
3813 --no-8x8dct option.
3814 fast-pskip
3816 Enable early SKIP detection on P-frames when set to 1. When set to
3817 0, it has the same effect as x264's --no-fast-pskip option.
3818 aud (aud)
3820 Enable use of access unit delimiters when set to 1.
3821 mbtree
3823 Enable use macroblock tree ratecontrol when set to 1. When set to
3824 0, it has the same effect as x264's --no-mbtree option.
3825 deblock (deblock)
3827 Set loop filter parameters, in alpha:beta form.
3828 cplxblur (cplxblur)
3830 Set fluctuations reduction in QP (before curve compression).
3831 partitions (partitions)
3833 Set partitions to consider as a comma-separated list of. Possible
3834 values in the list:
3835 p8x8
3837 8x8 P-frame partition.
3838 p4x4
3840 4x4 P-frame partition.
3841 b8x8
3843 4x4 B-frame partition.
3844 i8x8
3846 8x8 I-frame partition.
3847 i4x4
3849 4x4 I-frame partition. (Enabling p4x4 requires p8x8 to be
3850 enabled. Enabling i8x8 requires adaptive spatial transform
3851 (8x8dct option) to be enabled.)
3852 none (none)
3854 Do not consider any partitions.
3855 all (all)
3857 Consider every partition.
3858 direct-pred (direct)
3860 Set direct MV prediction mode. Possible values:
3861 none (none)
3863 Disable MV prediction.
3864 spatial (spatial)
3866 Enable spatial predicting.
3867 temporal (temporal)
3869 Enable temporal predicting.
3870 auto (auto)
3872 Automatically decided.
3873 slice-max-size (slice-max-size)
3875 Set the limit of the size of each slice in bytes. If not specified
3876 but RTP payload size (ps) is specified, that is used.
3877 stats (stats)
3879 Set the file name for multi-pass stats.
3880 nal-hrd (nal-hrd)
3882 Set signal HRD information (requires vbv-bufsize to be set).
3883 Possible values:
3884 none (none)
3886 Disable HRD information signaling.
3887 vbr (vbr)
3889 Variable bit rate.
3890 cbr (cbr)
3892 Constant bit rate (not allowed in MP4 container).
3893 x264opts (N.A.)
3895 Set any x264 option, see x264 --fullhelp for a list.
3896 Argument is a list of key=value couples separated by ":". In filter
3898 and psy-rd options that use ":" as a separator themselves, use ","
3899 instead. They accept it as well since long ago but this is kept
3900 undocumented for some reason.
3901 For example to specify libx264 encoding options with ffmpeg:
3903 ffmpeg -i foo.mpg -c:v libx264 -x264opts keyint=123:min-keyint=20 -an out.mkv
3905 a53cc boolean
3907 Import closed captions (which must be ATSC compatible format) into
3908 output. Only the mpeg2 and h264 decoders provide these. Default is
3909 1 (on).
3910 udu_sei boolean
3912 Import user data unregistered SEI if available into output. Default
3913 is 0 (off).
3914 x264-params (N.A.)
3916 Override the x264 configuration using a :-separated list of
3917 key=value parameters.
3918 This option is functionally the same as the x264opts, but is
3920 duplicated for compatibility with the Libav fork.
3921 For example to specify libx264 encoding options with ffmpeg:
3923 ffmpeg -i INPUT -c:v libx264 -x264-params level=30:bframes=0:weightp=0:\
3925 cabac=0:ref=1:vbv-maxrate=768:vbv-bufsize=2000:analyse=all:me=umh:\
3926 no-fast-pskip=1:subq=6:8x8dct=0:trellis=0 OUTPUT
3927 Encoding ffpresets for common usages are provided so they can be used
3929 with the general presets system (e.g. passing the pre option).
3930 libx265
3932 x265 H.265/HEVC encoder wrapper.
3933 This encoder requires the presence of the libx265 headers and library
3935 during configuration. You need to explicitly configure the build with
3936 --enable-libx265.
3937 Options
3939 b Sets target video bitrate.
3941 bf
3943 g Set the GOP size.
3944 keyint_min
3946 Minimum GOP size.
3947 refs
3949 Number of reference frames each P-frame can use. The range is from
3950 1-16.
3951 preset
3953 Set the x265 preset.
3954 tune
3956 Set the x265 tune parameter.
3957 profile
3959 Set profile restrictions.
3960 crf Set the quality for constant quality mode.
3962 qp Set constant quantization rate control method parameter.
3964 qmin
3966 Minimum quantizer scale.
3967 qmax
3969 Maximum quantizer scale.
3970 qdiff
3972 Maximum difference between quantizer scales.
3973 qblur
3975 Quantizer curve blur
3976 qcomp
3978 Quantizer curve compression factor
3979 i_qfactor
3981 b_qfactor
3982 forced-idr
3983 Normally, when forcing a I-frame type, the encoder can select any
3984 type of I-frame. This option forces it to choose an IDR-frame.
3985 udu_sei boolean
3987 Import user data unregistered SEI if available into output. Default
3988 is 0 (off).
3989 x265-params
3991 Set x265 options using a list of key=value couples separated by
3992 ":". See x265 --help for a list of options.
3993 For example to specify libx265 encoding options with -x265-params:
3995 ffmpeg -i input -c:v libx265 -x265-params crf=26:psy-rd=1 output.mp4
3997 libxavs2
3999 xavs2 AVS2-P2/IEEE1857.4 encoder wrapper.
4000 This encoder requires the presence of the libxavs2 headers and library
4002 during configuration. You need to explicitly configure the build with
4003 --enable-libxavs2.
4004 The following standard libavcodec options are used:
4006 • b / bit_rate
4008 • g / gop_size
4010 • bf / max_b_frames
4012 The encoder also has its own specific options:
4014 Options
4016 lcu_row_threads
4018 Set the number of parallel threads for rows from 1 to 8 (default
4019 5).
4020 initial_qp
4022 Set the xavs2 quantization parameter from 1 to 63 (default 34).
4023 This is used to set the initial qp for the first frame.
4024 qp Set the xavs2 quantization parameter from 1 to 63 (default 34).
4026 This is used to set the qp value under constant-QP mode.
4027 max_qp
4029 Set the max qp for rate control from 1 to 63 (default 55).
4030 min_qp
4032 Set the min qp for rate control from 1 to 63 (default 20).
4033 speed_level
4035 Set the Speed level from 0 to 9 (default 0). Higher is better but
4036 slower.
4037 log_level
4039 Set the log level from -1 to 3 (default 0). -1: none, 0: error, 1:
4040 warning, 2: info, 3: debug.
4041 xavs2-params
4043 Set xavs2 options using a list of key=value couples separated by
4044 ":".
4045 For example to specify libxavs2 encoding options with
4047 -xavs2-params:
4048 ffmpeg -i input -c:v libxavs2 -xavs2-params RdoqLevel=0 output.avs2
4050 libxvid
4052 Xvid MPEG-4 Part 2 encoder wrapper.
4053 This encoder requires the presence of the libxvidcore headers and
4055 library during configuration. You need to explicitly configure the
4056 build with "--enable-libxvid --enable-gpl".
4057 The native "mpeg4" encoder supports the MPEG-4 Part 2 format, so users
4059 can encode to this format without this library.
4060 Options
4062 The following options are supported by the libxvid wrapper. Some of the
4064 following options are listed but are not documented, and correspond to
4065 shared codec options. See the Codec Options chapter for their
4066 documentation. The other shared options which are not listed have no
4067 effect for the libxvid encoder.
4068 b
4070 g
4071 qmin
4072 qmax
4073 mpeg_quant
4074 threads
4075 bf
4076 b_qfactor
4077 b_qoffset
4078 flags
4079 Set specific encoding flags. Possible values:
4080 mv4 Use four motion vector by macroblock.
4082 aic Enable high quality AC prediction.
4084 gray
4086 Only encode grayscale.
4087 gmc Enable the use of global motion compensation (GMC).
4089 qpel
4091 Enable quarter-pixel motion compensation.
4092 cgop
4094 Enable closed GOP.
4095 global_header
4097 Place global headers in extradata instead of every keyframe.
4098 trellis
4100 me_method
4101 Set motion estimation method. Possible values in decreasing order
4102 of speed and increasing order of quality:
4103 zero
4105 Use no motion estimation (default).
4106 phods
4108 x1
4109 log Enable advanced diamond zonal search for 16x16 blocks and half-
4110 pixel refinement for 16x16 blocks. x1 and log are aliases for
4111 phods.
4112 epzs
4114 Enable all of the things described above, plus advanced diamond
4115 zonal search for 8x8 blocks, half-pixel refinement for 8x8
4116 blocks, and motion estimation on chroma planes.
4117 full
4119 Enable all of the things described above, plus extended 16x16
4120 and 8x8 blocks search.
4121 mbd Set macroblock decision algorithm. Possible values in the
4123 increasing order of quality:
4124 simple
4126 Use macroblock comparing function algorithm (default).
4127 bits
4129 Enable rate distortion-based half pixel and quarter pixel
4130 refinement for 16x16 blocks.
4131 rd Enable all of the things described above, plus rate distortion-
4133 based half pixel and quarter pixel refinement for 8x8 blocks,
4134 and rate distortion-based search using square pattern.
4135 lumi_aq
4137 Enable lumi masking adaptive quantization when set to 1. Default is
4138 0 (disabled).
4139 variance_aq
4141 Enable variance adaptive quantization when set to 1. Default is 0
4142 (disabled).
4143 When combined with lumi_aq, the resulting quality will not be
4145 better than any of the two specified individually. In other words,
4146 the resulting quality will be the worse one of the two effects.
4147 ssim
4149 Set structural similarity (SSIM) displaying method. Possible
4150 values:
4151 off Disable displaying of SSIM information.
4153 avg Output average SSIM at the end of encoding to stdout. The
4155 format of showing the average SSIM is:
4156 Average SSIM: %f
4158 For users who are not familiar with C, %f means a float number,
4160 or a decimal (e.g. 0.939232).
4161 frame
4163 Output both per-frame SSIM data during encoding and average
4164 SSIM at the end of encoding to stdout. The format of per-frame
4165 information is:
4166 SSIM: avg: %1.3f min: %1.3f max: %1.3f
4168 For users who are not familiar with C, %1.3f means a float
4170 number rounded to 3 digits after the dot (e.g. 0.932).
4171 ssim_acc
4173 Set SSIM accuracy. Valid options are integers within the range of
4174 0-4, while 0 gives the most accurate result and 4 computes the
4175 fastest.
4176 MediaFoundation
4178 This provides wrappers to encoders (both audio and video) in the
4179 MediaFoundation framework. It can access both SW and HW encoders.
4180 Video encoders can take input in either of nv12 or yuv420p form (some
4181 encoders support both, some support only either - in practice, nv12 is
4182 the safer choice, especially among HW encoders).
4183 mpeg2
4185 MPEG-2 video encoder.
4186 Options
4188 profile
4190 Select the mpeg2 profile to encode:
4191 422
4193 high
4194 ss Spatially Scalable
4195 snr SNR Scalable
4197 main
4199 simple
4200 level
4201 Select the mpeg2 level to encode:
4202 high
4204 high1440
4205 main
4206 low
4207 seq_disp_ext integer
4208 Specifies if the encoder should write a sequence_display_extension
4209 to the output.
4210 -1
4212 auto
4213 Decide automatically to write it or not (this is the default)
4214 by checking if the data to be written is different from the
4215 default or unspecified values.
4216 0
4218 never
4219 Never write it.
4220 1
4222 always
4223 Always write it.
4224 video_format integer
4226 Specifies the video_format written into the sequence display
4227 extension indicating the source of the video pictures. The default
4228 is unspecified, can be component, pal, ntsc, secam or mac. For
4229 maximum compatibility, use component.
4230 a53cc boolean
4232 Import closed captions (which must be ATSC compatible format) into
4233 output. Default is 1 (on).
4234 png
4236 PNG image encoder.
4237 Private options
4239 dpi integer
4241 Set physical density of pixels, in dots per inch, unset by default
4242 dpm integer
4244 Set physical density of pixels, in dots per meter, unset by default
4245 ProRes
4247 Apple ProRes encoder.
4248 FFmpeg contains 2 ProRes encoders, the prores-aw and prores-ks encoder.
4250 The used encoder can be chosen with the "-vcodec" option.
4251 Private Options for prores-ks
4253 profile integer
4255 Select the ProRes profile to encode
4256 proxy
4258 lt
4259 standard
4260 hq
4261 4444
4262 4444xq
4263 quant_mat integer
4264 Select quantization matrix.
4265 auto
4267 default
4268 proxy
4269 lt
4270 standard
4271 hq
4272 If set to auto, the matrix matching the profile will be picked. If
4274 not set, the matrix providing the highest quality, default, will be
4275 picked.
4276 bits_per_mb integer
4278 How many bits to allot for coding one macroblock. Different
4279 profiles use between 200 and 2400 bits per macroblock, the maximum
4280 is 8000.
4281 mbs_per_slice integer
4283 Number of macroblocks in each slice (1-8); the default value (8)
4284 should be good in almost all situations.
4285 vendor string
4287 Override the 4-byte vendor ID. A custom vendor ID like apl0 would
4288 claim the stream was produced by the Apple encoder.
4289 alpha_bits integer
4291 Specify number of bits for alpha component. Possible values are 0,
4292 8 and 16. Use 0 to disable alpha plane coding.
4293 Speed considerations
4295 In the default mode of operation the encoder has to honor frame
4297 constraints (i.e. not produce frames with size bigger than requested)
4298 while still making output picture as good as possible. A frame
4299 containing a lot of small details is harder to compress and the encoder
4300 would spend more time searching for appropriate quantizers for each
4301 slice.
4302 Setting a higher bits_per_mb limit will improve the speed.
4304 For the fastest encoding speed set the qscale parameter (4 is the
4306 recommended value) and do not set a size constraint.
4307 QSV encoders
4309 The family of Intel QuickSync Video encoders (MPEG-2, H.264, HEVC,
4310 JPEG/MJPEG and VP9)
4311 The ratecontrol method is selected as follows:
4313 • When global_quality is specified, a quality-based mode is used.
4315 Specifically this means either
4316 - CQP - constant quantizer scale, when the qscale codec flag is
4318 also set (the -qscale ffmpeg option).
4319 - LA_ICQ - intelligent constant quality with lookahead, when the
4321 look_ahead option is also set.
4322 - ICQ -- intelligent constant quality otherwise. For the ICQ
4324 modes, global quality range is 1 to 51, with 1 being the best
4325 quality.
4326 • Otherwise, a bitrate-based mode is used. For all of those, you
4328 should specify at least the desired average bitrate with the b
4329 option.
4330 - LA - VBR with lookahead, when the look_ahead option is
4332 specified.
4333 - VCM - video conferencing mode, when the vcm option is set.
4335 - CBR - constant bitrate, when maxrate is specified and equal to
4337 the average bitrate.
4338 - VBR - variable bitrate, when maxrate is specified, but is
4340 higher than the average bitrate.
4341 - AVBR - average VBR mode, when maxrate is not specified. This
4343 mode is further configured by the avbr_accuracy and
4344 avbr_convergence options.
4345 Note that depending on your system, a different mode than the one you
4347 specified may be selected by the encoder. Set the verbosity level to
4348 verbose or higher to see the actual settings used by the QSV runtime.
4349 Additional libavcodec global options are mapped to MSDK options as
4351 follows:
4352 • g/gop_size -> GopPicSize
4354 • bf/max_b_frames+1 -> GopRefDist
4356 • rc_init_occupancy/rc_initial_buffer_occupancy -> InitialDelayInKB
4358 • slices -> NumSlice
4360 • refs -> NumRefFrame
4362 • b_strategy/b_frame_strategy -> BRefType
4364 • cgop/CLOSED_GOP codec flag -> GopOptFlag
4366 • For the CQP mode, the i_qfactor/i_qoffset and b_qfactor/b_qoffset
4368 set the difference between QPP and QPI, and QPP and QPB
4369 respectively.
4370 • Setting the coder option to the value vlc will make the H.264
4372 encoder use CAVLC instead of CABAC.
4373 snow
4375 Options
4376 iterative_dia_size
4378 dia size for the iterative motion estimation
4379 VAAPI encoders
4381 Wrappers for hardware encoders accessible via VAAPI.
4382 These encoders only accept input in VAAPI hardware surfaces. If you
4384 have input in software frames, use the hwupload filter to upload them
4385 to the GPU.
4386 The following standard libavcodec options are used:
4388 • g / gop_size
4390 • bf / max_b_frames
4392 • profile
4394 If not set, this will be determined automatically from the format
4396 of the input frames and the profiles supported by the driver.
4397 • level
4399 • b / bit_rate
4401 • maxrate / rc_max_rate
4403 • bufsize / rc_buffer_size
4405 • rc_init_occupancy / rc_initial_buffer_occupancy
4407 • compression_level
4409 Speed / quality tradeoff: higher values are faster / worse quality.
4411 • q / global_quality
4413 Size / quality tradeoff: higher values are smaller / worse quality.
4415 • qmin
4417 • qmax
4419 • i_qfactor / i_quant_factor
4421 • i_qoffset / i_quant_offset
4423 • b_qfactor / b_quant_factor
4425 • b_qoffset / b_quant_offset
4427 • slices
4429 All encoders support the following options:
4431 low_power
4433 Some drivers/platforms offer a second encoder for some codecs
4434 intended to use less power than the default encoder; setting this
4435 option will attempt to use that encoder. Note that it may support
4436 a reduced feature set, so some other options may not be available
4437 in this mode.
4438 idr_interval
4440 Set the number of normal intra frames between full-refresh (IDR)
4441 frames in open-GOP mode. The intra frames are still IRAPs, but
4442 will not include global headers and may have non-decodable leading
4443 pictures.
4444 b_depth
4446 Set the B-frame reference depth. When set to one (the default),
4447 all B-frames will refer only to P- or I-frames. When set to
4448 greater values multiple layers of B-frames will be present, frames
4449 in each layer only referring to frames in higher layers.
4450 rc_mode
4452 Set the rate control mode to use. A given driver may only support
4453 a subset of modes.
4454 Possible modes:
4456 auto
4458 Choose the mode automatically based on driver support and the
4459 other options. This is the default.
4460 CQP Constant-quality.
4462 CBR Constant-bitrate.
4464 VBR Variable-bitrate.
4466 ICQ Intelligent constant-quality.
4468 QVBR
4470 Quality-defined variable-bitrate.
4471 AVBR
4473 Average variable bitrate.
4474 Each encoder also has its own specific options:
4476 h264_vaapi
4478 profile sets the value of profile_idc and the
4479 constraint_set*_flags. level sets the value of level_idc.
4480 coder
4482 Set entropy encoder (default is cabac). Possible values:
4483 ac
4485 cabac
4486 Use CABAC.
4487 vlc
4489 cavlc
4490 Use CAVLC.
4491 aud Include access unit delimiters in the stream (not included by
4493 default).
4494 sei Set SEI message types to include. Some combination of the
4496 following values:
4497 identifier
4499 Include a user_data_unregistered message containing
4500 information about the encoder.
4501 timing
4503 Include picture timing parameters (buffering_period and
4504 pic_timing messages).
4505 recovery_point
4507 Include recovery points where appropriate (recovery_point
4508 messages).
4509 hevc_vaapi
4511 profile and level set the values of general_profile_idc and
4512 general_level_idc respectively.
4513 aud Include access unit delimiters in the stream (not included by
4515 default).
4516 tier
4518 Set general_tier_flag. This may affect the level chosen for
4519 the stream if it is not explicitly specified.
4520 sei Set SEI message types to include. Some combination of the
4522 following values:
4523 hdr Include HDR metadata if the input frames have it
4525 (mastering_display_colour_volume and content_light_level
4526 messages).
4527 tiles
4529 Set the number of tiles to encode the input video with, as
4530 columns x rows. Larger numbers allow greater parallelism in
4531 both encoding and decoding, but may decrease coding efficiency.
4532 mjpeg_vaapi
4534 Only baseline DCT encoding is supported. The encoder always uses
4535 the standard quantisation and huffman tables - global_quality
4536 scales the standard quantisation table (range 1-100).
4537 For YUV, 4:2:0, 4:2:2 and 4:4:4 subsampling modes are supported.
4539 RGB is also supported, and will create an RGB JPEG.
4540 jfif
4542 Include JFIF header in each frame (not included by default).
4543 huffman
4545 Include standard huffman tables (on by default). Turning this
4546 off will save a few hundred bytes in each output frame, but may
4547 lose compatibility with some JPEG decoders which don't fully
4548 handle MJPEG.
4549 mpeg2_vaapi
4551 profile and level set the value of profile_and_level_indication.
4552 vp8_vaapi
4554 B-frames are not supported.
4555 global_quality sets the q_idx used for non-key frames (range
4557 0-127).
4558 loop_filter_level
4560 loop_filter_sharpness
4561 Manually set the loop filter parameters.
4562 vp9_vaapi
4564 global_quality sets the q_idx used for P-frames (range 0-255).
4565 loop_filter_level
4567 loop_filter_sharpness
4568 Manually set the loop filter parameters.
4569 B-frames are supported, but the output stream is always in encode
4571 order rather than display order. If B-frames are enabled, it may
4572 be necessary to use the vp9_raw_reorder bitstream filter to modify
4573 the output stream to display frames in the correct order.
4574 Only normal frames are produced - the vp9_superframe bitstream
4576 filter may be required to produce a stream usable with all
4577 decoders.
4578 vc2
4580 SMPTE VC-2 (previously BBC Dirac Pro). This codec was primarily aimed
4581 at professional broadcasting but since it supports yuv420, yuv422 and
4582 yuv444 at 8 (limited range or full range), 10 or 12 bits, this makes it
4583 suitable for other tasks which require low overhead and low compression
4584 (like screen recording).
4585 Options
4587 b Sets target video bitrate. Usually that's around 1:6 of the
4589 uncompressed video bitrate (e.g. for 1920x1080 50fps yuv422p10
4590 that's around 400Mbps). Higher values (close to the uncompressed
4591 bitrate) turn on lossless compression mode.
4592 field_order
4594 Enables field coding when set (e.g. to tt - top field first) for
4595 interlaced inputs. Should increase compression with interlaced
4596 content as it splits the fields and encodes each separately.
4597 wavelet_depth
4599 Sets the total amount of wavelet transforms to apply, between 1 and
4600 5 (default). Lower values reduce compression and quality. Less
4601 capable decoders may not be able to handle values of wavelet_depth
4602 over 3.
4603 wavelet_type
4605 Sets the transform type. Currently only 5_3 (LeGall) and 9_7
4606 (Deslauriers-Dubuc) are implemented, with 9_7 being the one with
4607 better compression and thus is the default.
4608 slice_width
4610 slice_height
4611 Sets the slice size for each slice. Larger values result in better
4612 compression. For compatibility with other more limited decoders
4613 use slice_width of 32 and slice_height of 8.
4614 tolerance
4616 Sets the undershoot tolerance of the rate control system in
4617 percent. This is to prevent an expensive search from being run.
4618 qm Sets the quantization matrix preset to use by default or when
4620 wavelet_depth is set to 5
4621 - default Uses the default quantization matrix from the
4623 specifications, extended with values for the fifth level. This
4624 provides a good balance between keeping detail and omitting
4625 artifacts.
4626 - flat Use a completely zeroed out quantization matrix. This
4628 increases PSNR but might reduce perception. Use in bogus
4629 benchmarks.
4630 - color Reduces detail but attempts to preserve color at
4632 extremely low bitrates.
4633
4635 dvdsub
4636 This codec encodes the bitmap subtitle format that is used in DVDs.
4637 Typically they are stored in VOBSUB file pairs (*.idx + *.sub), and
4638 they can also be used in Matroska files.
4639 Options
4641 palette
4643 Specify the global palette used by the bitmaps.
4644 The format for this option is a string containing 16 24-bits
4646 hexadecimal numbers (without 0x prefix) separated by commas, for
4647 example "0d00ee, ee450d, 101010, eaeaea, 0ce60b, ec14ed, ebff0b,
4648 0d617a, 7b7b7b, d1d1d1, 7b2a0e, 0d950c, 0f007b, cf0dec, cfa80c,
4649 7c127b".
4650 even_rows_fix
4652 When set to 1, enable a work-around that makes the number of pixel
4653 rows even in all subtitles. This fixes a problem with some players
4654 that cut off the bottom row if the number is odd. The work-around
4655 just adds a fully transparent row if needed. The overhead is low,
4656 typically one byte per subtitle on average.
4657 By default, this work-around is disabled.
4659
4661 ffmpeg(1), ffplay(1), ffprobe(1), libavcodec(3)
4662
4664 The FFmpeg developers.
4665 For details about the authorship, see the Git history of the project
4667 (git://source.ffmpeg.org/ffmpeg), e.g. by typing the command git log in
4668 the FFmpeg source directory, or browsing the online repository at
4669 <http://source.ffmpeg.org>.
4670 Maintainers for the specific components are listed in the file
4672 MAINTAINERS in the source code tree.
4673 FFMPEG-CODECS(1)