June 2019 - ffmpeg-user - lists.ffmpeg.org

Re: [FFmpeg-user] JPGs->video1 ; MOV -> Mp4 video2 ; Concatenate - video2 audio OK but video is slow . .
by Philip Rhoades 18 Mar '23

18 Mar '23

Carl, There has been some improvement - I will top-post for clarity but the slow motion video has disappeared and the only remaining problem is that the audio and video are out of sync: I stuck with concat because it alllows me to specify image files more conveniently in the concat_demuxer?.txt file: ffmpeg -f concat -i concat_demuxer1.txt -vf scale=1024:576 -r 30000/1001 output1.mp4 ffmpeg -f concat -i concat_demuxer2.txt -vf scale=1024:576 -r 30000/1001 output2.mp4 ffmpeg -f concat -i concat_demuxer3.txt -vf scale=1024:576 -r 30000/1001 output3.mp4 - see attached output. for i in 1 2 3 do ffmpeg -ar 44100 -acodec pcm_s16le -f s16le -ac 2 -i /dev/zero \ -i output${i}.mp4 \ -vcodec copy \ -acodec aac \ -shortest output${i}_silent.mp4 done - see attached output. #!/bin/sh ffmpeg -f concat -i inputs.txt -vcodec copy -acodec copy 2019-03-22_#1_XR_ClimateEmergencyDeclarationDayNSW_LoRes.mp4 - see attached output. On 2019-03-27 10:48, Carl Eugen Hoyos wrote: > 2019-03-27 0:23 GMT+01:00, Philip Rhoades <phil(a)pricom.com.au>: > >> On 2019-03-27 07:22, Carl Eugen Hoyos wrote: >>> 2019-03-26 21:15 GMT+01:00, Philip Rhoades <phil(a)pricom.com.au>: >>> >>>> I converted 21 JPGs to a video with: >>> >>> Usually, converting jpgs with the image2 input works better than >>> concat. >> >> Right - I will try that. > > This is probably not your issue but the concat demuxer > has the disadvantage that the console output is not > very helpful. > >>>> ffmpeg -f concat -i concat_demuxer1.txt -vf scale=1024:576 -r >>>> 29.97 >>> >>> This framerate is never correct, a possible frame rate is 30000/1001 >> >> Hmm . . I was just using what was reported for the frame for the MOVs >> . >> . so using: >> >> "-r 30000/1001" >> >> is OK? > > It is better if that is the framerate of the mov (as I assume). > >>> You can add the audio already in this step. >> >> I did have a few attempts at that but couldn't get it working so I >> resorted to a two-step process - I will have another look. >> >> >>>> output1.mp4 >>>> >>>> I then added a silent audio track with: >>>> >>>> ffmpeg -ar 44100 -acodec pcm_s16le -f s16le -ac 2 -channel_layout >>>> 2.1 >>> >>> The channel number and layout do not match. >> >> Should it be just "stereo"? > > Or drop channel_layout, for -ac 2 it should not be needed. > >>>> -i /dev/zero \ >>>> -i output${i}.mp4 \ >>>> -vcodec copy \ >>>> -acodec aac \ >>>> -shortest output${i}_silent.mp4 >>> >>>> I had another MOV file (fr=29.97) that I converted with: >>>> >>>> newfn=`basename $fn .MOV` >>>> ffmpeg -i $fn -vf scale=1024:576 ./${newfn}.mp4 >>>> >>>> When I concatenate them with (and other converted JPGs and MOVS on >>>> the >>>> end) using: >>>> >>>> ffmpeg -f concat -i inputs.txt -vcodec copy -acodec copy \ >>>> 2019-03-22_#1_XR_ClimateEmergencyDeclarationDayNSW_LoRes.mp4 >>>> >>>> the concat succeeds OK but when I play the video, the audio is OK >>>> but >>>> then when it gets to the converted MOV section, people are moving in >>>> slow motion . . >>>> >>>> Another odd thing, the first 21 converted JPGs at 4 sec each >>>> correctly >>>> shows 84 sec when viewed with mpv but when the converted MOV starts >>>> it >>>> jumps to ~3m 50s . . >>>> >>>> So there is at least one problem but maybe two? Can anyone tell me >>>> what >>>> I am missing? >>> >>> You may have to re-encode, feel free to provide the complete, uncut >>> console output. > > ..feel free to provide the complete, uncut console output for the > command line you already tried and also ffmpeg -i output for > the input files (as concat doesn't show it). > > Carl Eugen -- Philip Rhoades PO Box 896 Cowra NSW 2794 Australia E-mail: phil(a)pricom.com.au

2 2

Recording audio from mic and speakers from commandline
by Mohan G 28 Sep '21

28 Sep '21

Hi, I am using ffmpeg version 2.4.7 on Fedora-21 workstation. I want to record audio, simultaneously from mic and speakers as separate streams. I am using the following command for it. ffmpeg -f alsa -i pulse -f alsa -i default -acodec libmp3lame -map 0:0 -map 1:0 outfile.mkv But after executing this command in terminal, I have to goto "Volume Control" settings in GUI, then goto "Recording" tab, change the "capture from" of one channel to "Built-in Audio Analog Stereo" (i.e., Mic) and the "capture from" of another channel to "Monitor of Built-in Audio Analog Stereo" (i.e., Speakers). But how can this be done completely from command-line, for automating it to use in scripts. - Mohan

3 10

How to have a secure UDP BroadCast
by hassan shatnawi 19 Aug '20

19 Aug '20

Dears, How to have a secure connection with username and password while broadcasting "UPD" I've use this command ffmpeg -f dshow -video_size 1280x720 -rtbufsize 702000K -framerate 30 -i video="Video (00-0 Pro Capture Quad HDMI)" -r 30 -threads 4 -vcodec libx264 -crf 0 -preset ultrafast -f mpegts "udp://username:password@192.168.1.12:6666" but it didn't work can you advice me plz. Thanks in advance. *Hassan Shatnawi* Software Developer *Software Development Department* *Haupshy Establishment for Electronics* Al Madina Al Monawara St. bld 31 | P.O.Box 6875 Amman 11118, Jordan. *T: *+962 6 5820 927* | F: *+962 6 5820 926* | M: *+962 78 5792 096 *E*: hassan(a)haupshy.com *|* www.haupshy.com

7 7

Can FFmpeg convert from ArriRaw MXF sources?
by Harvey Pikelberger 19 Nov '19

19 Nov '19

Is FFmpeg capable of transcoding from ArriaRaw to h.264?

6 11

denoise after scaling or before?
by Jon bae 27 Oct '19

27 Oct '19

Hello, I just have an general question: As I know denoiser are specialized to specific noise pattern. Makes it a different when I denoise after scaling a video, or is the quality better when I denoise before scaling. In terms of speed it would be better for me, denoising after scaling, but when the quality is better, then I would go that way. Have a nice day! Jonathan

5 7

[avi @ 0x9873ba0] Application provided invalid, non monotonically increasing dts to muxer in stream 1: 1104 >= 1104 [avi @ 0x987d2a0] Application provided invalid, non monotonically increasing dts to muxer only when using tee muxer
by Michael Kohne 11 Sep '19

11 Sep '19

I'm using ffmpeg 4.1 to grab an rtsp stream (includes both H.264 video and G.711 audio) from a Sony EP580 IP camera. If I transcode to mpeg4v2 and aac to a single avi file, things seem to work fine. This command line will go for as long as I like (I regularly do 15 minutes at a go). ffmpeg -min_port 62000 -max_port 62004 -i rtsp://192.168.0.113/media/video1 -codec:v msmpeg4v2 -codec:a ac3 -ar 44100 -map 0 -f avi /data/PatientData/vidtmp/test.avi If I instead use the tee muxer to make two avi files: ffmpeg -min_port 62000 -max_port 62004 -i rtsp://192.168.0.113/media/video1 -codec:v msmpeg4v2 -codec:a ac3 -ar 44100 -map 0 -f tee "[f=avi]/data/PatientData/vidtmp/one.avi|[f=avi]/data/PatientData/vidtmp/two.avi" Then within 2 minutes of starting the stream, it fails with the errors: [avi @ 0x9873ba0] Application provided invalid, non monotonically increasing dts to muxer in stream 1: 1104 >= 1104 [avi @ 0x987d2a0] Application provided invalid, non monotonically increasing dts to muxer in stream 1: 1104 >= 1104 I've repeated these tests over and over for a week or so now, so I am fairly confident that it's not simply that the camera is janky some of the time. I've also tested using the rtsp stream 'rtsp:// 184.72.239.149/vod/mp4:BigBuckBunny_175k.mov'. That stream does not produce any errors with the tee muxer. Note that those errors both appear to come from the avi muxers, not the tee muxer, which is why I find it surprising: if the camera's output causes this, then why doesn't it happen with the single avi muxer? Test note: the tee invocation is the simplest command line that I could come up with that produces the problem. In the real world one avi muxer goes to a pipe for a live output display, and the other goes to the segment muxer to produce output files in manageable chunks. I did also try this with the latest ffmpeg snapshot (as of yesterday morning), and there was no difference in behavior. Any clues as to how to debug this would be greatly appreciated. ffmpeg version 4.1 Copyright (c) 2000-2018 the FFmpeg developers built with gcc 4.4.7 (GCC) 20120313 (Red Hat 4.4.7-11) configuration: --prefix=/data/SHR4634/ffmpeg/build --extra-cflags=-I/data/SHR4634/ffmpeg/build/include --extra-ldflags=-L/data/SHR4634/ffmpeg/build/lib --extra-libs='-lm -ldl -lpthread -lrt' --enable-gpl --enable-nonfree --disable-libfdk_aac --enable-libmp3lame --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libfreetype --enable-libspeex --enable-libtheora --cpu=i686 --enable-runtime-cpudetect libavutil 56. 22.100 / 56. 22.100 libavcodec 58. 35.100 / 58. 35.100 libavformat 58. 20.100 / 58. 20.100 libavdevice 58. 5.100 / 58. 5.100 libavfilter 7. 40.101 / 7. 40.101 libswscale 5. 3.100 / 5. 3.100 libswresample 3. 3.100 / 3. 3.100 libpostproc 55. 3.100 / 55. 3.100 OS is CentOS 6, 32 bit. Michael Kohne Senior Software Engineer Office: 215.283.0860 x208 mhkohne(a)moberg.com -- Celebrating 20 Years Transforming Neurocritical Care Moberg Research, Inc. 224 S Maple Street, Ambler, PA 19002 24/7 Customer Support: 888.662.7246 www.moberg.com <https://www.moberg.com/>

2 7

Is it possible to timeout a JPEG dump
by Martijn van Beurden 09 Jul '19

09 Jul '19

Hi all, Currently I have a script running that dumps snapshots from a security camera MJPEG stream with this command ffmpeg -i http://admin:admin@192.168.10.4/channel2 -c:v copy -f image2 -strftime 1 %Y-%m-%dT%H:%M:%S%z.jpg However, in case the network has a dropout for some reason, the connection is broken but FFMPEG hangs. The process isn't stopped (apparently the broken connection is not detected) and there is no way of detecting this. Is there any way to set a timeout between frames arriving or something similar? Console output is this: ffmpeg version N-93674-g1e01f66822 Copyright (c) 2000-2019 the FFmpeg developers built with gcc 7 (Ubuntu 7.3.0-27ubuntu1~18.04) configuration: --pkg-config-flags=--static --extra-cflags='-static -static-libstdc++ -static-libgcc -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -fPIC' --extra-ldflags='-static -static-libstdc++ -static-libgcc' --extra-libs='-lm -ldl -lpthread -lz -lrt' --enable-gpl --enable-nonfree --enable-libvorbis --enable-libtheora --enable-libvpx --enable-libopus libavutil 56. 26.100 / 56. 26.100 libavcodec 58. 52.100 / 58. 52.100 libavformat 58. 27.103 / 58. 27.103 libavdevice 58. 7.100 / 58. 7.100 libavfilter 7. 49.100 / 7. 49.100 libswscale 5. 4.100 / 5. 4.100 libswresample 3. 4.100 / 3. 4.100 libpostproc 55. 4.100 / 55. 4.100 Input #0, mpjpeg, from 'http://admin:admin@192.168.10.4/channel2': Duration: N/A, bitrate: N/A Stream #0:0: Video: mjpeg (Baseline), yuvj420p(pc, bt470bg/unknown/unknown), 1920x1080 [SAR 1:1 DAR 16:9], 25 tbr, 25 tbn, 25 tbc Output #0, image2, to '/srv/camera/snapshots/zuidwest/%Y-%m-%dT%H:%M:%S%z.jpg': Metadata: encoder : Lavf58.27.103 Stream #0:0: Video: mjpeg (Baseline), yuvj420p(pc, bt470bg/unknown/unknown), 1920x1080 [SAR 1:1 DAR 16:9], q=2-31, 25 tbr, 25 tbn, 25 tbc Stream mapping: Stream #0:0 -> #0:0 (copy) Press [q] to stop, [?] for help [2.0K blob data] [2.0K blob data] [2.0K blob data] [2.0K blob data] [2.0K blob data] [2.0K blob data] [2.0K blob data]

1 1

Equivalent of "-f rawvideo" but for audio?
by Peter B. 01 Jul '19

01 Jul '19

Hi everyone :) Is it possible somehow to transcode any source audio format type to its raw, uncompressed format that matches the source? I'm dealing with collections of mixed input format combinations, and am trying to find a way to bash-automate something. For video "-f rawvideo" seems to do the trick. But according to the "audio types" documentation [1] it seems I have to know define the (and therefore already know) the format. So, instead of specifying "-f s16le" or "-f s24le", is there something like "-f rawaudio"? I'm grateful for any hints or input! Thanks in advance, Peter B. ## Links: [1] https://trac.ffmpeg.org/wiki/audio%20types

7 11

global_sidx on IOs12
by James Northrup 28 Jun '19

28 Jun '19

i have dash encodings as follow working well in shaka player and the chunk files individually working well in android and desktop browsers. this file does not launch quickly only on IOs despite -global_sidx 1 https://iqbsug36tcyiy7ggk5jf-std-uploads.oss-ap-southeast-5.aliyuncs.com/in… also, master.hls does not play audio in ios or anything else i have tested (i don't test a lot of hls players) how do i get either of HLS to play or hinted fmpg for ios download conumption? running details below: ffmpeg version N-93817-ga6cd3ad Copyright (c) 2000-2019 the FFmpeg developers built with gcc 7 (Ubuntu 7.3.0-27ubuntu1~18.04) configuration: --prefix=/home/jim/ffmpeg_build --pkg-config-flags= --extra-cflags=-I/home/jim/ffmpeg_build/include --extra-ldfl ags=-L/home/jim/ffmpeg_build/lib --extra-libs='-lpthread -lm' --bindir=/home/jim/bin --disable-ffplay --enable-bzlib --enable-ffm peg --enable-gpl --enable-iconv --enable-libfdk-aac --enable-libmp3lame --enable-libopus --enable-libsnappy --enable-libssh --ena ble-nonfree --enable-openssl --enable-version3 --enable-zlib --enable-libwebp --enable-libfreetype --enable-libtwolame --enable-l ibvpx --enable-avresample --enable-librsvg --enable-libx264 --enable-libvorbis --enable-libtheora --disable-shared --enable-lto - -enable-static libavutil 56. 26.101 / 56. 26.101 libavcodec 58. 52.101 / 58. 52.101 libavformat 58. 27.103 / 58. 27.103 libavdevice 58. 7.100 / 58. 7.100 libavfilter 7. 50.100 / 7. 50.100 libavresample 4. 0. 0 / 4. 0. 0 libswscale 5. 4.100 / 5. 4.100 libswresample 3. 4.100 / 3. 4.100 libpostproc 55. 4.100 / 55. 4.100 ffmpeg -i ../1.mp3 -filter_complex asplit=4 -codec :a:0 libopus -ab:a:0 32k -vbr:a:0 off -frame_duration:a:0 60 -application:a:0 voip -codec:a:1 libopus -ab:a:1 128k -vbr:a:1 on -f rame_duration:a:1 20 -application:a:1 audio -codec:a:2 libfdk_aac -ab:a:2 32k -profile:a:2 aac_he_v2 -codec:a:3 aac -ab:a:3 128k -nostdin -strict -2 -f dash -seg_duration 3 -lhls 1 -hls_playlist 1 -streaming 0 -strict experimental -dash_segment_type auto -gl obal_sidx 1 -http_persistent 1 -index_correction 1 -init_seg_name init-eps-6-agenda-ham-untuk-pemerintah-dan-parlemen-terpilih$Re presentationID$-$Bandwidth$.$Ext$ -media_seg_name chunk-eps-6-agenda-ham-untuk-pemerintah-dan-parlemen-terpilih$RepresentationID$ -$Bandwidth$-$Number%08d$.$Ext$ -single_file_name stream-eps-6-agenda-ham-untuk-pemerintah-dan-parlemen-terpilih-$RepresentationI D$-$Bandwidth$.$Ext$ -adaptation_sets id=0,streams=0,1 id=1,streams=2,3 -y 1.mpd.xml manifests: https://iqbsug36tcyiy7ggk5jf-std-uploads.oss-ap-southeast-5.aliyuncs.com/in… https://iqbsug36tcyiy7ggk5jf-std-uploads.oss-ap-southeast-5.aliyuncs.com/in…

1 1

pipe from /dev/video to /dev/video adding a delay - syncing issues
by Javier Infante Porro 27 Jun '19

27 Jun '19

Hello, I am trying to create a video device using another video device, adding a 20 seconds extra delay. Also, i will create on the fly a custom countdown video before my delayed stream starts. I create my virtual video device using v4l2loopback: modprobe v4l2loopback exclusive_caps=1 video_nr=20 card_label="LIVE" The command writting to file works pretty good (an initial 20 seconds countdown with bg.jpg background): ffmpeg -loglevel 43 -y -loop 1 -r 28 -t 20 -i bg.jpg -f v4l2 -thread_queue_size 512 -i /dev/video0 -filter_complex "[0:v:0]setpts=N/(28*TB),drawtext=fontfile='/usr/share/fonts/truetype/msttcorefonts/Georgia.ttf':fontcolor=yellow:fontsize=40:x=(w-text_w)/2:y=(h-text_h)/2:text='%{eif\:20-t\:d}'[count]; [count] [1:v:0] concat=n=2:v=1:a=0:unsafe=1 [v];[v]scale=852:480[out]" -map "[out]" -an -pix_fmt yuv420p -c:v libx264 -preset ultrafast s1.mp4 But whenever i try to push the stream back to the the /dev/video20 device, it works, but the countdown goes real fast (like 10x faster), and then video tries to catch up and keeps around a 2 second delay: ffmpeg -loglevel 43 -y -loop 1 -r 28 -t 20 -i bg.jpg -f v4l2 -thread_queue_size 512 -i /dev/video0 -filter_complex "[0:v:0]setpts=N/(28*TB),drawtext=fontfile='/usr/share/fonts/truetype/msttcorefonts/Georgia.ttf':fontcolor=yellow:fontsize=40:x=(w-text_w)/2:y=(h-text_h)/2:text='%{eif\:20-t\:d}'[count]; [count] [1:v:0] concat=n=2:v=1:a=0:unsafe=1 [v];[v]scale=852:480[out]" -map "[out]" -an -pix_fmt yuv420p -f v4l2 -r 28 /dev/video20 Any ideas on how to output as v4l2 correctly or achiving what i am trying using another approach? Thanks in advance! -- Jabi

3 6