Recherche avancée

Sur d’autres sites (8013)

• Quickly split videos into parts with ffmpeg

  20 avril 2021, par RewossKy

  I'm doing splitting on FFMPEG. I divide 1.5 hour videos into 30 minutes. I have two ways of partitioning, the first is the method of rendering by recodecing on the top, I do not have any problem in this, but the time is quite long despite the GTX 1660s graphics card and the ryzen5 3500x processor.

  


  At the bottom, the time is short, but the first 3-4 seconds of the second part videos are frozen. Sometimes this causes problems like not reading in video editing programs.

  


  With recodec (slow, no problem)
ffmpeg -i 1.mp4 -vcodec libx264 -crf 30 -ss 00:00:00 -t 00:29:30 s1.mp4
ffmpeg -i 1.mp4 -vcodec libx264 -crf 30 -ss 00:29:31 -t 00:29:30 s2.mp4
ffmpeg -i 1.mp4 -vcodec libx264 -crf 30 -ss 00:59:01 -t 00:29:30 s3.mp4
ffmpeg -i 1.mp4 -vcodec libx264 -crf 30 -ss 01:28:30 -t 00:29:30 s4.mp4

  


  Without recodec (fast but have problem sometimes)
ffmpeg -i 1.mp4 -vcodec copy -ss 00:00:00 -t 00:29:30 s1.mp4
ffmpeg -i 1.mp4 -vcodec copy -ss 00:29:31 -t 00:29:30 s2.mp4
ffmpeg -i 1.mp4 -vcodec copy -ss 00:59:01 -t 00:29:30 s3.mp4
ffmpeg -i 1.mp4 -vcodec copy -ss 01:28:30 -t 00:29:30 s4.mp4

  


• Not able to run the final step of running supervisord for plumi

  1er avril 2021, par manic_startup

  on the github page https://github.com/plumi/plumi.app
I am on the final step of Step 6. see below section

  


  =========================

  


  Step 6 : start supervisord
With plumi buildout having finished and ffmpeg installed on our system, it's time to run supervisord :

  


  root@user :/home/plumi.app# ./bin/supervisord

  


  ========================

  


  When I type in the above on CentOS 7 it says "bash : ./bin/supervisord : No such file or directory

  


  I had to change some things in step 5 "Step 5 Install ffmpeg" for the centos section there are two urls in that section pointing to some repos. They did not work so I had to find some other ones on the Internet. The ones I used where.
https://dl.fedoraproject.org/pub/epel/7/x86_64/Packages/e/epel-release-7-13.noarch.rpm

  


  http://ftp.riken.jp/Linux/atrpms/el7-x86_64/atrpms/stable/atrpms-repo-7-7.el7.x86_64.rpm

  


  The first one looks the same but it is different also the only release available was 7-13 so i used that.
The second one is a different url but the file is the same.

  


  also in that section is asks to edit /etc/yum.repos.d/atrpms.repo. It asks to change the base URL to a URL similar to the what is asks when it asked to download from the second repo above. Obviously that url is different now so I changed it to this instead.
http://ftp.riken.jp/Linux/atrpms/el$releasever-$basearch/atrpms/stable/

  


  After this I asks to install ffmpeg this did not work so I used the below guide
https://linuxize.com/post/how-to-install-ffmpeg-on-centos-7/

  


  Then came the section 6 to install supervisord as explained at the start this did not work. Now im stuck !

  


  I have also realised that when I get to the following command in step 4.

  


  ./bin/python bootstrap.py && ./bin/buildout -v

  


  It fails and says "Error : couldn't find distribution for 'hachoir-parser==1.3.4' and bombs out. I have tried to find this file but cant install it and dont know if i even have the correct file. not sure what to do.

  


• m3u8 hls stream to wav in one step with ffmpeg

  17 février 2021, par Marinos K

  I'm using to ffmpeg to extract 10" snippets in WAV audio from a m3u8 hls stream in two steps like this :

  


  ffmpeg -i XXXX.m3u8 -t 10 -c copy -bsf:a aac_adtstoasc -vn output.acc
ffmpeg -i output.acc output.wav


  


  How can I do it in just one step to avoid the intermediate file and maybe speed the process ?

  


  I tried directly encoding the output with something like -c:a pcm_s24le, but this fails with a 'codec not supported by bitstream filter' error.

  


  I also tried piping the output of the first command to the next like this :

  


  ffmpeg -i XXX.m3u8 -t 10 -c copy -bsf:a aac_adtstoasc -vn pipe: | ffmpeg  -i pipe: output.wav


  


  and I get an 'pipe: : Invalid data found when processing input' error.

  


  any indications of what am I doing wrong, or any other ideas ?