Automation Script to Edit Videos for TikTok using Python

autoclip.py

#    _____     _       _____ _ _     
#   |  _  |_ _| |_ ___|     | |_|___ 
#   |     | | |  _| . |   --| | | . |
#   |__|__|___|_| |___|_____|_|_|  _|
#                            |_| 
#
# By Abhishta (github.com/abhishtagatya)

# pip install gTTs
# pip install moviepy
# pip install rich
# pip install pyfiglet

import os
import random
from typing import List, Tuple

from gtts import gTTS
from moviepy.editor import (
    CompositeVideoClip, CompositeAudioClip,
    VideoFileClip, AudioFileClip, ImageClip, TextClip,
    concatenate_videoclips, concatenate_audioclips
)
import moviepy.video.fx.all as vfx
from rich.console import Console
from rich.progress import track
import pyfiglet


def generate_speech(
        text: str, 
        lang: str = 'en', 
        filename: str = 'audio.mp3'):
    """
    Generate Speech Audio from gTTS

    text: str - Text to be synthesized
    lang: str - Language of text
    filename: str - Filename of output
    """
    myobj = gTTS(text=text, lang=lang, slow=False, tld='ca') # Change per settings https://gtts.readthedocs.io/en/latest/module.html
    myobj.save(filename)
    return


def clip(
        content: str, 
        video_file: str, 
        outfile: str, 
        image_file: str = '', 
        offset: int = 0, 
        duration: int = 0):
    """
    Generate the Complete Clip

    content: str - Full content text
    video_file: str - Background video
    outfile: str - Filename of output
    image_file: str - Banner to display
    offset: int - Offset starting point of background video (default: 0)
    duration: int - Limit the video (default: audio length)
    """
    audio_comp, text_comp = generate_audio_text(split_text(content))

    audio_comp_list = []
    for audio_file in track(audio_comp, description='Stitching Audio...'):
        audio_comp_list.append(AudioFileClip(audio_file))
    audio_comp_stitch = concatenate_audioclips(audio_comp_list)
    audio_comp_stitch.write_audiofile('temp_audio.mp3', fps=44100)

    audio_duration = audio_comp_stitch.duration
    if duration == 0:
        duration = audio_duration

    audio_comp_stitch.close()

    vid_clip = VideoFileClip(video_file).subclip(offset, offset + duration)
    vid_clip = vid_clip.resize((1980, 1280))
    vid_clip = vid_clip.crop(x_center=1980 / 2, y_center=1280 / 2, width=720, height=1280)

    if image_file != '':
        image_clip = ImageClip(image_file).set_duration(duration).set_position(("center", 'center')).resize(0.8) # Adjust if the Banner is too small
        vid_clip = CompositeVideoClip([vid_clip, image_clip])

    vid_clip = CompositeVideoClip([vid_clip, concatenate_videoclips(text_comp).set_position(('center', 860))])

    vid_clip = vid_clip.set_audio(AudioFileClip('temp_audio.mp3').subclip(0, duration))
    vid_clip.write_videofile(outfile, audio_codec='aac')
    vid_clip.close()


def split_text(text: str, delimiter: str = '\n'):
    """
    Split the Text

    text: str - Text to split
    delimiter: str - Delimiter of split (default: \n)
    """
    return text.split(delimiter)


def generate_audio_text(fulltext: List[str]):
    """
    Generate Audio and Text from Full Text

    fulltext: List[str] - List of splitted Text
    """
    audio_comp = []
    text_comp = []

    for idx, text in track(enumerate(fulltext), description='Synthesizing Audio...'):
        if text == "":
            continue
            
        audio_file = f"temp_assets/audio_{idx}.mp3"
        generate_speech(text.strip(), filename=audio_file)

        audio_duration = AudioFileClip(audio_file).duration

        text_clip = TextClip(
            text,
            font='Helvetica', # Change Font if not found
            fontsize=32,
            color="white",
            align='center',
            method='caption',
            size=(660, None)
        )
        text_clip = text_clip.set_duration(audio_duration)

        audio_comp.append(audio_file)
        text_comp.append(text_clip)

    return audio_comp, text_comp


if __name__ == '__main__':

    console = Console()
    banner = pyfiglet.figlet_format(text='AutoClip', font='rectangles')
    console.print()
    console.print("[bold][red1]" + banner)
    console.print("[dark_red] By Abhishta (github.com/abhishtagatya)")

    if not os.path.exists("temp_assets"):
        os.mkdir("temp_assets")

    video_background_file = "<SET YOUR VIDEO FILE>.mp4" # Your video background file
    video_background_offset = random.randint(0, 5000) # Starting Position of Video : 0 for Beginning
    image_banner_file = "<SET YOUR IMAGE BANNER>.png" # Your image banner file
    output_file = "AutoClip_Out.mp4" # The output filename

    content = """<YOUR STORY>
    <RECOMMENDED TO SPLIT PER LINES>
    <IF YOU WANT TO DISPLAY SUBTITLES IN THIS ORDER>
    <THANK YOU>
    """

    console.print("\n\n[light_green] Task Starting\n\n")
    clip(content=content, 
         video_file=video_background_file, 
         image_file=image_banner_file,
         outfile=output_file, 
         offset=video_background_offset)

    console.print("\n\n[light_green] Completed!")

requirement.txt

gTTS==2.2.4
moviepy==1.0.3
pyfiglet==0.8.post1
rich==13.4.2

Startup Guide

1. Download the script (autoclip.py)
2. Download all the required packages (the list of package requirements are inside the requirements.txt)

pip install [package-name]

3. Fill in the variables. They are marked inside the main section. You will need a background video, an image, and a content (a story to tell)
4. Run the script

python autoclip.py

Ta-da! Everything should be generating and rendering the video.

If somethings looks off with the video, please tweak the variable inside the clip() function to adjust.

nice script! are you posting these videos? do you think this approach is suitable for yt shorts also?

@aproxis hey there, thank you. i don't see why not, go ahead :)