UNUSED

The current version is a much improved version upon this by Gaurav. Find the project here: https://github.com/igauravsehrawat/meetup-utilities

Chennaipy-video-helpers

This repository contains helper scripts and commands to create meetup videos for Chennaipy.

Steps

• Install ffmpeg, sox, imagemagick, ghostscript

  • On OSX, you can directly use brew for the above:

      $ brew install ffmpeg sox imagemagick ghostscript
    

  • On Ubuntu, installing ffmpeg is a bit tricky, but these commands should get you started:

      $ sudo add-apt-repository ppa:mc3man/trusty-media
      $ sudo apt-get update
      $ sudo apt-get dist-upgrade
      $ sudo apt-get install ffmpeg imagemagick ghostscript sox
    

    You can read more about the ffmpeg ppa here

  • On Windows, you can individually download and install prebuilt binaries for the packages:

    • ffmpeg, 64-bit static should do
    • sox
    • imagemagick
    • ghostscript
    • you might have to manually add them to PATH if not done automatically.

• Convert the audio to a wav (wav files are faster to process, and easier to manipulate using sox)

    $ ffmpeg -i bigdata_audio.m4a audio.wav
  

• Remove noise from audio

    $ sox audio.wav -n trim 0 1.5 noiseprof noise.profile
    $ sox audio.wav cleaned.wav noisered noise.profile 0
  

• Convert the pdf to individual images using

    $ convert -density 150 -resize 1920x1080 bigbigdata.pdf individual_images/%d.png
  

• Notedown timestamps from audio at which slide is to be changed

• Generate segments of video for those times

    $ python create_video_segments.py individual_images\ individual_video\ 00:02 00:40 1:25 2:10 3:35 3:50 5:30 7:22 7:42 8:02 10:03 12:10 13:10 15:12 16:00 17:25 18:30 20:00 21:18 21:23
  

• Trim the audio if required

    $ sox cleaned.wav final.wav trim 00:02 =21:20
  

• Generate a concatvideos file to be used by ffmpeg to join the videos to form a single one

    $ generate_concatvideos.py individual_video\
  

• Run the ffmpeg script to create the joined video

    $ ffmpeg -f concat -i concatvideos -c copy joined.mp4
  

• Join the generated video with the final audio

    $ ffmpeg -i joined.mp4 -i final.wav -shortest -preset ultrafast -q 0 final.mp4
  

• Adjust volume in the video

    $ ffmpeg -i <input_video_name> -af 'volume=2.5<or 1x>' -preset ultrafast <output_video_name>
  

• Trim the video if required

    $ ffmpeg -ss <start_time_in_hh:mm:sec> -i <input_video_name>  -t <time_duration_in_hh:mm:sec> <output_video_name>
  

• scale video to 16:9

    $ ffmpeg -i final.mp4 -vf pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2" -preset ultrafast -acodec copy final2.mp4
  

• Generate the title image

    $ sed -e 's/%title%/My Python Experience/g' -e 's/%speaker%/Shrikant Giridhar/g' -e 's/%month%/Feb/g'  -e 's/%year%/2017/g' title-template.svg > title.svg
  

• convert the svg to a png

    $ convert title.svg -density 150 -resize 1920x1080! title.png
  

• create title video with empty audio

    $ ffmpeg -loop 1 -i title.png -f lavfi -i aevalsrc=0 -acodec libvo_aacenc -ab 128k -map 0:0 -map 1:0 -t 7 -preset ultrafast -qp 0 title.mp4
  

• create the uploadable version

    $ ffmpeg -i title.mp4 -i final2.mp4 -filter_complex "[0:0] [0:1] [1:0] [1:1] concat=n=2:v=1:a=1 [v] [a]" -map "[v]" -map "[a]"  uploadable.mp4
  

  Note: Use -preset ultrafast if cpu is consistently high, but increases overall filesize. Using it is not recommended.