Bash Script for creating videos with FFMPEG, ImageMagick and gTTS

video_generator.sh

#!/bin/bash

# Check if the right number of arguments were passed
if [ $# -ne 6 ]; then
    echo "Usage: $0 \"primary_text\" \"secondary_text\" "KURAL_ENN" "Athigaram" input_image.png output.mp4"
    echo "Example: $0 "இருள்சேர் இருவினையும் சேரா இறைவன்\nபொருள்சேர் புகழ்புரிந்தார் மாட்டு" "கடவுளின் மெய்மைப் புகழையே விரும்புபவரிடம் அறியாமை இருளால் வரும் நல்வினை தீவினை என்னும் இரண்டும் சேருவதில்லை" "5" "கடவுள் வாழ்த்து"  bg.png kural5.mp4"
    exit 1
fi

# Assign arguments to variables
PRIMARY_TEXT="$1"
SECONDARY_TEXT="$2"
KURAL_ENN="குறள்: $3"
ATH_ENN="அதிகாரம்: $4"
INPUT_IMAGE="$5"
OUTPUT_VIDEO="$6"
FONT="Samyak-Tamil"


# Check if input image exists
if [ ! -f "$INPUT_IMAGE" ]; then
    echo "Error: Input image $INPUT_IMAGE does not exist."
    exit 1
fi

# Generate TTS audio using gTTS from SECONDARY_TEXT
AUDIO_FILE=$(mktemp --suffix=.mp3)
echo "Generating audio..."
gtts-cli "$SECONDARY_TEXT" --lang ta --output "$AUDIO_FILE"

# Get duration of audio in seconds
DURATION=$(ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 "$AUDIO_FILE")
DURATION=${DURATION%.*} # convert to integer
DURATION=$((DURATION + 1))
if [ -z "$DURATION" ]; then
    echo "Error: Could not determine audio duration."
    rm "$TEMP_IMAGE" "$AUDIO_FILE"
    exit 1
fi

# Get image dimensions
WIDTH=$(identify -format "%w" "$INPUT_IMAGE" 2>/dev/null)
HEIGHT=$(identify -format "%h" "$INPUT_IMAGE" 2>/dev/null)
if [ $? -ne 0 ]; then
    echo "Error: Could not get dimensions of $INPUT_IMAGE. Is ImageMagick installed?"
    echo "Install with: sudo apt-get install imagemagick"
    exit 1
fi

echo "Creating image with overlaid texts..."

# Create temporary image file
TEMP_IMAGE=$(mktemp --suffix=.png)

# Calculate a suitable text width based on image width (you might need to adjust this)
TEXT_WIDTH=$(( WIDTH * 80 / 100 ))

# Create base composite with primary text using Pango and wrapping within geometry
convert "$INPUT_IMAGE" \
    \( -background none -font $FONT -pointsize 32 \
       -fill white -stroke black -strokewidth 4 \
       -undercolor "rgba(1,0,0,0.1)" \
       -size "${TEXT_WIDTH}x" pango:"$PRIMARY_TEXT" \) \
    -gravity north -geometry +0+200 -composite "$TEMP_IMAGE"

# Add secondary text with Pango and wrapping within geometry
convert "$TEMP_IMAGE" \
    \( -background none -font $FONT -pointsize 32 \
       -fill white -stroke black -strokewidth 4 \
       -undercolor "rgba(1,0,0,0.1)" \
       -size "${TEXT_WIDTH}x" pango:"$SECONDARY_TEXT" \) \
    -gravity center -geometry +0-40 -composite "$TEMP_IMAGE"

convert "$TEMP_IMAGE" \
    \( -background none -font $FONT -pointsize 32 \
       -fill white -stroke black -strokewidth 4 \
       -undercolor "rgba(1,0,0,0.1)" \
       -size "${TEXT_WIDTH}x" pango:"பொருள்:" \) \
    -gravity center -geometry +0-190 -composite "$TEMP_IMAGE"

convert "$TEMP_IMAGE" \
    \( -background none -font $FONT -pointsize 32 \
       -fill white -stroke black -strokewidth 4 \
       -undercolor "rgba(1,0,0,0.1)" \
       -size "${TEXT_WIDTH}x" pango:"$KURAL_ENN" \) \
    -gravity north -geometry +0+20 -composite "$TEMP_IMAGE"

convert "$TEMP_IMAGE" \
    \( -background none -font $FONT -pointsize 32 \
       -fill white -stroke black -strokewidth 4 \
       -undercolor "rgba(1,0,0,0.1)" \
       -size "${TEXT_WIDTH}x" pango:"$ATH_ENN" \) \
    -gravity north -geometry +0+80 -composite "$TEMP_IMAGE"

echo "Creating video from the composite image with audio..."

ffmpeg -hide_banner -y -loop 1 -i "$TEMP_IMAGE" -i "$AUDIO_FILE" \
    -c:v libx264 -t "$DURATION" -pix_fmt yuv420p \
    -c:a aac -b:a 192k -shortest "$OUTPUT_VIDEO"

# Check for success
if [ $? -ne 0 ]; then
    echo "Error: Failed to create video with audio."
    rm "$TEMP_IMAGE" "$AUDIO_FILE"
    exit 1
fi

# Clean up
rm "$TEMP_IMAGE" "$AUDIO_FILE"

echo "Video successfully created: $OUTPUT_VIDEO"