PDF --> MP4 Settings

[jrdeee]

ConvertX appears great - I am converting PDF to MP4 and appears to work, other than each page only lasting 5ms (edit: sorry, it's 4ms). A 28 page PDF runs through very quick, as you can imagine.

I know there are no settings per conversion (none that I can see at least) but is there a way to modify conversion settings, perhaps in the compose file for example? In my use case, I would like to use a 3 seconds pause per page.

[jrdeee]

AI helped with a workaround, thought I'd share in case it helps others. There may be a much better / streamlined way of doing this, but I'm no expert hence relying on what I can find. I tested with one video, seems to work - time will tell with others.

AI explains the situation as and script function as:
ConvertX treats PDF pages as a rapid image sequence and defaults to a blindingly fast frame rate (like 25 frames per second), flashing each page on screen for a uselessly brief 4ms to 40ms

The script scans the output directories for these .mp4 video files, rips the video apart into its original raw image frames (the individual PDF pages) and temporarily saves them. It then stitches those images back together into a standard, smooth 30fps video, forcing each slide to freeze on screen for exactly 3 seconds. Once the new _fixed.mp4 file is safely saved right next to the original, the script automatically deletes the fast version and cleans up its temporary files to save storage space.

# The auto-fixer service that stretches your video
  ffmpeg-fixer:
    image: jrottenberg/ffmpeg:4.4-ubuntu
    container_name: ffmpeg-fixer
    user: root     # Bypasses permission locks on dynamic folders created by ConvertX
    restart: unless-stopped
    volumes:
      # Shares the same folder so it can access the fast video
      - ./data:/media   # same data folder as ConvertX
    entrypoint: ["/bin/bash", "-c"]
    command: >
      -c "
      echo 'Watching recursively for fast videos...';
      while true; do
        find /media -type f -name '*.mp4' ! -name '*_fixed.mp4' | while read -r file; do
          echo \"Found file: $$file\";
          
          # 1. Clear out any previous temp frames to stay clean
          rm -f /tmp/frame_*.png;
          
          # 2. Extract every single distinct frame out of the video as a raw image file
          ffmpeg -y -i \"$$file\" -vf \"select='eq(mod(n,1),0)'\" -vsync vfr /tmp/frame_%04d.png;
          
          # 3. Stitch the raw images back together, forcing each image to hold for exactly 3 seconds
          ffmpeg -y -framerate 1/3 -i /tmp/frame_%04d.png -c:v libx264 -r 30 -pix_fmt yuv420p \"$${file%.mp4}_fixed.mp4\" && rm \"$$file\";
          
          # 4. Clean up temp workspace files
          rm -f /tmp/frame_*.png;
          
          echo \"Done fixing! Saved right beside original as $${file%.mp4}_fixed.mp4\";
        done;
        sleep 5;
      done
      "

[Trithereon]

This is not possible in version 0.18.0 of ConvertX.

The way ConvertX handles converting pdf to mp4 in version 0.18.0 is to use ImageMagick by calling the command magick input.pdf output.mp4. What you need is to add the delay argument, with a value that matches your preferred delay before the image switches to the next image. The value is in "centiseconds" where 100 centiseconds = 1 second. So if you want a delay of 2 seconds per image, you would set the value to 200, for a full command of magick -delay 200 input.pdf output.mp4. If you want, you can install imagemagick on your machine and use that exact command yourself. You will probably need ghostscript and ffmpeg installed too, since imagemagick calls them to read the pdf and to create video, respectively.