Making a Grep Video Using FFmpeg

Making a Grep Video Using FFmpeg | Generated by AI

Home 2026.03

Sure! Here’s the full breakdown:

🦞 How I Made the Grep Tutorial Video

1. Install ffmpeg

sudo apt-get install -y ffmpeg

2. Generate slide PNGs using ffmpeg’s drawtext filter

Each “slide” is a PNG image created with ffmpeg’s lavfi (libavfilter) virtual input + drawtext overlays. No GUI needed — all pure CLI.

ffmpeg -y -f lavfi -i "color=c=0x1a1b26:s=1280x720:d=1" \
  -vf "drawtext=fontfile=FONT:text='Title':fontsize=42:fontcolor=0x7dcfff:x=(w-text_w)/2:y=30, \
       drawtext=fontfile=FONT:text='Some line':fontsize=24:fontcolor=0xc0caf5:x=70:y=105" \
  -frames:v 1 slide.png

Key tricks:

color=c=0x1a1b26 — Tokyo Night dark background
drawtext chained with commas — stacks multiple text lines on one image
(w-text_w)/2 — centers text horizontally
Color-coded: green=commands, blue=output, yellow=tips, grey=comments

3. Concatenate slides into video with the concat demuxer

Created a concat.txt listing each slide + how long to show it:

file '00_title.png'
duration 4
file '01_what.png'
duration 6
...

Then:

ffmpeg -y -f concat -safe 0 -i concat.txt \
  -vf "fade=t=in:st=0:d=0.5,fade=t=out:st=93:d=1.5" \
  -c:v libx264 -pix_fmt yuv420p -crf 22 -r 30 \
  grep_tutorial.mp4

4. The script itself

The whole thing is driven by tutorial/grep_tutorial.py (~280 lines of Python). It:

Defines slide content as structured data (auto-colors $, >, # prefixed lines)
Calls ffmpeg via subprocess for each slide
Writes the concat file
Runs final ffmpeg to stitch it all together

Want me to walk through any specific part, or tweak the video (add audio, different colors, more slides)?

Back Donate