January 2025 - ffmpeg-user - lists.ffmpeg.org

FFMPEG Integration with Intratrust DRM
by rameshk＠unisysinfo.in 25 Jan '25

25 Jan '25

Hi Team Please help us to integrate ffmpeg with Intertrust DRM. Please share documentation and flow for the same. Thanks & Regards Ramesh Kumar System Administrator Unisys Infosolutions C: +91 9416778187 Email:- <mailto:Rameshk@unisysinfo.in> Rameshk(a)unisysinfo.in

4 4

ffmpeg doesn't reconnect when capturing live connections
by arthurpeabody 25 Jan '25

25 Jan '25

I started identical captures on my home computer and at a remote server. The capture at home stopped capturing at 42 minutes; the capture at the remote computer finished. The process at home is still running; the process at the remote computer ended. No error was reported. https://pastebin.com/vuinwUJQ

2 1

Streaming music with just ffmpeg, how to let concat preserve ogg metadata?
by Mark Gaiser 24 Jan '25

24 Jan '25

Hi, I'm trying to make a music streaming server with just ffmpeg and mpv [1]. The architecture is a typical client (C) server (S) model. S has the music files, does pre-processing and library management. C plays these files. I've added some randomization so that S serves a random music file every time a new song starts playing. C plays music like this: mpv "tcp://0.0.0.0:8988?listen" And S streams music like this: ffmpeg -re -stream_loop -1 -f concat -safe 0 -i /tmp/playlist.txt -flush_packets 0 \ -c:a copy -vn -f ogg tcp://10.0.3.14:8988 The playlist "/tmp/playlist.txt" contains: ffconcat version 1.0 file /tmp/song1 file /tmp/song2 An external tool makes sure that song1 and song2 are pre-processed to actual songs. They are OGG files. This works! Mostly... Metadata from the OGG files isn't preserved. On the receiving end in mpv it's just one long massive stream. I've tried many [2] many [3] many [4, 5] options but all have their own issues (mostly with the stream stopping after an EOF of the music file itself. The concat option, however weird it is to use, is near perfect! The only downside to me is that it drops metadata. I'd like to preserve that! Any ideas on how to do that? As an aside as it's bound to be asked. No, I cannot easily mount the data from S on C and play music that way. On a technical level C is essentially an raspberry pi with a speaker attached. S is doing some audio pre-processing (normalizing) and monitoring for new files. It would get a bit too heavy for C to do that too. Best regards, Mark [1] https://mpv.io/ it uses ffmpeg under the hood [2] Streaming per ogg file does give metadata on the receiving end. However, after a song is done playing both ends stop (EOF). [3] A variation on [2], let ffmpeg output into a pipe (mkfifo) and cat that pipe over the network using cat and nc. That works again for just one (sometimes two) songs when using cat. [4] A variation on [3], using tail instead of cat (as tail doesn't send EOF). Only the first song plays, mpv simply hangs after the first one. Neither mpv nor ffmpeg close. [5] I also tried a variation where i'm streaming to shoutcast and play that using mpv. This suffers from the same issues but just in the shoutcast ingest point. I don't want to go the full shoutcast webradio server route.

1 0

File duration of MP3 generated by mp3_mf seems wrong
by Tomas Kalibera 22 Jan '25

22 Jan '25

Hello, when I try converting a wav file to mp3 using the mp3_mf codec, and then display the duration of the resulting file, it is surprisingly short. Am I doing or interpreting anything wrong, or is this a bug and should I submit a full bug report? The output file otherwise plays as expected for the expected duration. ffmpeg -i u.wav ## gives duration 5.22s ffmpeg -i u.wav -acodec mp3_mf u_mf.mp3 ffmpeg -i u_mf.mp3 ## gives duration 1.49s Now, when I use "-acodec libmp3lame", the resulting file duration is as expected. According "mp3info -x" on Ubuntu, "Length" is 5s in both cases, as it should be. But the surprisingly short duration is shown even on Ubuntu. Not sure if it helps, but when converting using mp3_mf, the output includes: Stream #0:0: Audio: mp3, 44100 Hz, stereo, s16, 128 kb/s size= 83KiB time=00:00:05.22 bitrate= 129.7kbits/s speed= 149x When converting using libmp3lame, the output includes: Stream #0:0: Audio: mp3, 44100 Hz, stereo, s16p size= 82KiB time=00:00:05.22 bitrate= 129.0kbits/s speed= 121x The part of "ffmpeg -i" on the result with mp3_mf is: Duration: 00:00:01.49, start: 0.011995, bitrate: 455 kb/s Stream #0:0: Audio: mp3, 44100 Hz, stereo, fltp, 128 kb/s And with libmp3lame it is: Duration: 00:00:05.25, start: 0.025057, bitrate: 128 kb/s Stream #0:0: Audio: mp3, 44100 Hz, stereo, fltp, 128 kb/s This is on Windows 10 (build 19045) with ffmpeg 7.1 from Msys2/ucrt64. Thanks, Tomas

1 0

header intact
by prabhavathis＠ridsys.in 22 Jan '25

22 Jan '25

header intact

1 0

ffmpeg compilation failed while compiling ffmpeg git at the point amf.
by Guddla Rupesh 21 Jan '25

21 Jan '25

Hi I am Rupesh from India and I brought a new system with AMD Ryzen zen3 5500GT and Asus prime B450 motherboard. I failed compiling the source code of ffmpeg which is downloaded from git at the point of AMF in Linux Operating System. I have installed the following Linux Operating Systems Arch Linux Debian Linux Trixie Fedora 41 Gentoo Linux I have downloaded and installed all the packages related to sound, video and images which are necessary to compile install and run ffmpeg properly. >From the past four months I am using ffmpeg along with the above Linux Operating Systems. At the starting point of compilation I have invoked the configure script of ffmpeg git source code through the following command ./configure --enable-amf The configuration succeeded so I have issued the following command make -j9 After sometime I got errors containing the words "amf" One month back I have downloaded ffmpeg source code from git and configured and compiled successfully using the same commands. As I received errors while compiling new ffmpeg source code obtained from git I have deleted the source code and copied old source code of ffmpeg which was obtained from git. I am able to compile old source code of ffmpeg but I can't compile the latest version of ffmpeg which is obtained from git. I am providing the commands which I entered previously and now below. [CODE] git clone https://github.com/FFmpeg/FFmpeg ffmpeg tar -czvf ffmpeg.tar.gz ffmpeg cd ffmpeg/ ./configure --arch=x86_64 --cpu=znver3 --prefix=/opt/ffmpeg --enable-gpl --disable-shared --enable-fontconfig --enable-frei0r --enable-gnutls --enable-gray --enable-libaom --enable-libfribidi --enable-libass --enable-libvmaf --enable-libfreetype --enable-libmp3lame --enable-librav1e --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libvorbis --enable-libopus --enable-libtheora --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libdav1d --enable-libxvid --enable-libzvbi --enable-libzimg --enable-libfdk-aac --enable-nonfree --enable-vaapi --enable-vdpau --enable-opencl --enable-opengl --enable-hwaccels --enable-runtime-cpudetect --disable-cuda-sdk --enable-avfilter --enable-bzlib --enable-libdrm --enable-libjxl --enable-libopenjpeg --enable-libplacebo --enable-librsvg --enable-librubberband --enable-libzmq --enable-pthreads --enable-vulkan --enable-amf --disable-libmfx > ../ffmpeg_output.txt make -j9 sudo make install [/CODE] I am providing the configuration command which I entered previously and now below. [CODE] ./configure --arch=x86_64 --cpu=znver3 --prefix=/opt/ffmpeg --enable-gpl --disable-shared --enable-fontconfig --enable-frei0r --enable-gnutls --enable-gray --enable-libaom --enable-libfribidi --enable-libass --enable-libvmaf --enable-libfreetype --enable-libmp3lame --enable-librav1e --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libvorbis --enable-libopus --enable-libtheora --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libdav1d --enable-libxvid --enable-libzvbi --enable-libzimg --enable-libfdk-aac --enable-nonfree --enable-vaapi --enable-vdpau --enable-opencl --enable-opengl --enable-hwaccels --enable-runtime-cpudetect --disable-cuda-sdk --enable-avfilter --enable-bzlib --enable-libdrm --enable-libjxl --enable-libopenjpeg --enable-libplacebo --enable-librsvg --enable-librubberband --enable-libzmq --enable-pthreads --enable-vulkan --enable-amf --disable-libmfx > ../ffmpeg_output.txt [/CODE] I am also providing the last few lines of the compilation output below. [CODE] CC libavcodec/acelp_filters.o CC libavcodec/acelp_pitch_delay.o CC libavcodec/acelp_vectors.o CC libavcodec/adpcm.o CC libavcodec/adpcm_data.o CC libavcodec/adpcmenc.o In file included from libavcodec/ac3dec_float.c:40: In function ‘spx_strategy’, inlined from ‘decode_audio_block’ at libavcodec/ac3dec.c:1144:24: libavcodec/ac3dec.c:856:37: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 856 | s->channel_uses_spx[ch] = get_bits1(bc); | ~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~ In file included from libavcodec/ac3dec_float.c:34: libavcodec/ac3dec.h: In function ‘decode_audio_block’: libavcodec/ac3dec.h:144:13: note: at offset 7 into destination object ‘channel_uses_spx’ of size 7 144 | uint8_t channel_uses_spx[AC3_MAX_CHANNELS]; ///< channel uses spectral extension (chinspx) | ^~~~~~~~~~~~~~~~ In file included from libavcodec/aacsbr.c:383: libavcodec/aacsbr_template.c: In function ‘read_sbr_grid’: libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In file included from libavcodec/aacsbr.c:31: libavcodec/sbr.h:106:24: note: at offset [2147483649, 4294967295] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [2147483649, 4294967294] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [2147483649, 4294967293] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [2147483649, 4294967292] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ In file included from libavcodec/ac3dec_fixed.c:159: In function ‘spx_strategy’, inlined from ‘decode_audio_block’ at libavcodec/ac3dec.c:1144:24: libavcodec/ac3dec.c:856:37: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 856 | s->channel_uses_spx[ch] = get_bits1(bc); | ~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~ In file included from libavcodec/ac3dec_fixed.c:52: libavcodec/ac3dec.h: In function ‘decode_audio_block’: libavcodec/ac3dec.h:144:13: note: at offset 7 into destination object ‘channel_uses_spx’ of size 7 144 | uint8_t channel_uses_spx[AC3_MAX_CHANNELS]; ///< channel uses spectral extension (chinspx) | ^~~~~~~~~~~~~~~~ CC libavcodec/adts_header.o CC libavcodec/adts_parser.o CC libavcodec/adx.o CC libavcodec/adx_parser.o In file included from libavcodec/aacsbr_fixed.c:610: libavcodec/aacsbr_template.c: In function ‘read_sbr_grid’: libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In file included from libavcodec/aacsbr_fixed.c:60: libavcodec/sbr.h:106:24: note: at offset [-2147483647, -1] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [-2147483647, -2] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [-2147483647, -3] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ libavcodec/aacsbr_template.c:729:57: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] 729 | ch_data->t_env[ch_data->bs_num_env - 1 - i] = | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^ 730 | ch_data->t_env[ch_data->bs_num_env - i] - 2 * get_bits(gb, 2) - 2; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/sbr.h:106:24: note: at offset [-2147483647, -4] into destination object ‘t_env’ of size 9 106 | uint8_t t_env[9]; | ^~~~~ CC libavcodec/adxdec.o CC libavcodec/adxenc.o CC libavcodec/agm.o CC libavcodec/aic.o CC libavcodec/alac.o CC libavcodec/alac_data.o CC libavcodec/alacdsp.o CC libavcodec/alacenc.o CC libavcodec/aliaspixdec.o CC libavcodec/aliaspixenc.o CC libavcodec/allcodecs.o CC libavcodec/alsdec.o CC libavcodec/amfenc.o CC libavcodec/amfenc_av1.o CC libavcodec/amfenc_h264.o In file included from /usr/include/AMF/core/Data.h:37, from /usr/include/AMF/core/Buffer.h:37, from /usr/include/AMF/core/Context.h:37, from /usr/include/AMF/core/Factory.h:40, from libavcodec/amfenc.h:22, from libavcodec/amfenc_av1.c:23: libavcodec/amfenc_av1.c: In function ‘amf_encode_init_av1’: libavcodec/amfenc_av1.c:467:57: error: ‘AMF_VIDEO_ENCODER_AV1_ADAPTIVE_MINIGOP’ undeclared (first use in this function); did you mean ‘AMF_VIDEO_ENCODER_ADAPTIVE_MINIGOP’? 467 | AMF_ASSIGN_PROPERTY_BOOL(res, ctx->encoder, AMF_VIDEO_ENCODER_AV1_ADAPTIVE_MINIGOP, ((ctx->pa_adaptive_mini_gop == 0) ? false : true)); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/amfenc_av1.c:467:57: note: each undeclared identifier is reported only once for each function it appears in libavcodec/amfenc_av1.c:481:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement] 481 | AMFVariantStruct is_adaptive_b_frames = { 0 }; | ^~~~~~~~~~~~~~~~ libavcodec/amfenc_av1.c:490:66: error: ‘AMF_VIDEO_ENCODER_AV1_CAP_BFRAMES’ undeclared (first use in this function); did you mean ‘AMF_VIDEO_ENCODER_CAP_BFRAMES’? 490 | res = encoder_caps->pVtbl->GetProperty(encoder_caps, AMF_VIDEO_ENCODER_AV1_CAP_BFRAMES, &var); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | AMF_VIDEO_ENCODER_CAP_BFRAMES libavcodec/amfenc_av1.c:499:70: error: ‘AMF_VIDEO_ENCODER_AV1_MAX_CONSECUTIVE_BPICTURES’ undeclared (first use in this function); did you mean ‘AMF_VIDEO_ENCODER_MAX_CONSECUTIVE_BPICTURES’? 499 | AMF_ASSIGN_PROPERTY_INT64(res, ctx->encoder, AMF_VIDEO_ENCODER_AV1_MAX_CONSECUTIVE_BPICTURES, 3); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ libavcodec/amfenc_av1.c:510:74: error: ‘AMF_VIDEO_ENCODER_AV1_B_PIC_PATTERN’ undeclared (first use in this function); did you mean ‘AMF_VIDEO_ENCODER_B_PIC_PATTERN’? 510 | AMF_ASSIGN_PROPERTY_INT64(res, ctx->encoder, AMF_VIDEO_ENCODER_AV1_B_PIC_PATTERN, ctx->max_b_frames); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ make: *** [ffbuild/common.mak:81: libavcodec/amfenc_av1.o] Error 1 make: *** Waiting for unfinished jobs.... build@Rupesh:~/soft/temp/ffmpeg$ [/CODE] Kindly try to suggest what's wrong and how to compile the source code of the latest version of ffmpeg which is obtained from git. Regards, Rupesh.

1 0

how to put DTS to mp4
by KoSza 17 Jan '25

17 Jan '25

Hello, I am going forward with my imagined workflow but now I faced a new problem: how to put DTS to mp4. I tried some possibilities wich I found but nothing happened. What is the exact way for it? Thank you, kosza

1 0

Live Transcoding: What Causes Large Time Gaps from Input to Output
by Shane Warren 17 Jan '25

17 Jan '25

I am transcoding live multicast using hardware (decoding using nvidia and encoding using nvidia (h264_nvenc) and going out multicast at fixed framerate and constant bitrate. Normally the output is about 3-5 seconds behind the input. I have some streams that cause the output to eventually get minutes behind the input. I've narrowed down the input buffering being the culprit of holding a very large buffer of input frames in these situations. Here is my command: ffmpeg -y -threads 2 -nostats -nostdin -loglevel verbose -progress pipe:1 -probesize 5M -filter_threads 4 -threads 2 -re -fflags +genpts -fflags discardcorrupt -hwaccel_device 0 -extra_hw_frames 2 -hwaccel_output_format cuda -hwaccel cuda -heavy_compr 1 -thread_queue_size 512 -i "udp://@225.105.0.27:10102?fifo_size=1073440&buffer_size=2129920&timeout=800000" -autoscale 0 -threads 2 -filter_complex "[0:v:0]fps=30000/1001,yadif_cuda,scale_npp=1920:1080:interp_algo=super:format=yuv420p [vout]" -map "[vout]" -map "0:a:0" -map 0:d? -filter:a:0 "aresample=async=10000,volume=1.00" -c:a:0 aac -threads 2 -ac:a:0 2 -ar:a:0 48000 -b:a:0 192k -flush_packets 0 -c:v h264_nvenc -b:v 6000k -minrate:v 6000k -maxrate:v 6000k -bufsize:v 12000k -a53cc 1 -tune ll -zerolatency 1 -cbr 1 -forced-idr 1 -strict_gop 1 -threads 2 -profile:v high -level:v 4.2 -bf:v 0 -g:v 15 -f fifo -drop_pkts_on_overflow 1 -attempt_recovery 0 -fifo_format mpegts -format_opts muxrate=6800440:pes_payload_size=1528 "udp://@225.105.0.65:10102?pkt_size=1316&bitrate=6800440&burst_bits=10528&ttl=64" I suspect the input stream does something odd and causes ffmpeg to buffer frames and it gets slowly behind. I would like to make ffmpeg drop frames if possible, instead of forever buffering frames and getting further and further out of sync. Is there anything anyone can recommend to force ffmpeg to stop buffering input frames?

3 2

ffmpeg doesn't reconnect when capturing live connections
by arthurpeabody 15 Jan '25

15 Jan '25

Whoops! Ignore that example. The stream went down. I'll have an example in which it didn't eventually.

1 0

ffmpeg doesn't reconnect when capturing live connections
by arthurpeabody 14 Jan '25

14 Jan '25

ffmpeg quit 14 minutes into this capture, didn't reconnect: https://pastebin.com/uGHPqiBy

1 0