DEV Community

Cover image for How to build a video watermarking app using Python
Kushal Magar
Kushal Magar

Posted on • Originally published at shotstack.io

How to build a video watermarking app using Python

Watermarking is a common way to label digital content. It enables brands and creators add branding to their content. Especially in this hyper-connected era where videos spread in a blink of an eye, it's important that brands can easily add branding to their work.

Of course, you can use the available video editors or web apps. They sure are simple and easy to watermark a single or a few videos. But what if you need to watermark thousands of videos? Even better, what if you can build an app that enables users to watermark videos?

That's what this tutorial aims to teach - to build a program using Python to automatically watermark videos. After all, Python developers love to automate, don't they?

This tutorial has two parts:

  1. watermarking a single video
  2. watermarking multiple videos using a list

The Shotstack API and SDK

Shotstack provides a cloud based video editing API. Rendering videos is resource intensive and it can take hours to edit and generate videos at scale. Shotstack's rendering infrastructure makes it possible to build and scale media applications in days, not months.

We will also be using the Shotstack video editing Python SDK for this
tutorial. The SDK requires Python 3.

Install and configure the Shotstack SDK

If you want to skip ahead you can find the source code for this guide in our GitHub repository. Otherwise, follow the steps below to install dependencies and set up your API key.

First of all, install the Shotstack Python SDK from the command line:

pip install shotstack_sdk
Enter fullscreen mode Exit fullscreen mode

You may need to use pip3 depending on how your environment is configured.

Then, set your API key as an environment variable (Linux/Mac):

export SHOTSTACK_KEY=your_key_here
Enter fullscreen mode Exit fullscreen mode

or, if using Windows (make sure to add the SHOTSTACK_KEY to the path):

set SHOTSTACK_KEY=your_key_here
Enter fullscreen mode Exit fullscreen mode

Replace your_key_here with your provided sandbox API key which is free for testing and development.

Create a Python script to watermark a video

Create a file for the script in your favorite IDE or text editor. You can call it whatever you like, but for this tutorial, we created a file called watermark-video.py. Open the file and begin editing.

Import the required modules

Let's import the required modules for the project. We need to import modules from the Shotstack SDK to edit and render
our video plus a couple of built-in modules:

import shotstack_sdk as shotstack
import os
import sys

from shotstack_sdk.model.clip import Clip
from shotstack_sdk.api import edit_api
from shotstack_sdk.model.track import Track
from shotstack_sdk.model.timeline import Timeline
from shotstack_sdk.model.output import Output
from shotstack_sdk.model.edit import Edit
from shotstack_sdk.model.video_asset import VideoAsset
Enter fullscreen mode Exit fullscreen mode

Configuring the API client

Next, add the following, which sets up the API client with the API URL and key, this should use the API key added to
your environment variables. If you want, you can hard code the API key here but we recommend using environment
variables.

host = "https://api.shotstack.io/stage"
configuration = shotstack.Configuration(host = host)
configuration.api_key['DeveloperKey'] = os.getenv('SHOTSTACK_KEY')
with shotstack.ApiClient(configuration) as api_client:
    api_instance = edit_api.EditApi(api_client)
Enter fullscreen mode Exit fullscreen mode

Understanding the timeline architecture

The Shotstack API follows many of the principles of desktop editing software such as the use of a timeline, tracks, and
clips. A timeline is like a container for multiple clips that includes
different assets which plays over time.
Tracks on the timeline allow us to layer clips on top of each other.

Setting up the video clip

The video needs to be hosted online and accessible via a public or signed URL. We will use the following 10-second drone footage as our video asset. You can replace it with your own video URL from any online source.

Add the following code to create a VideoAsset using the video URL:

video_asset = VideoAsset(
    src = "https://d1uej6xx5jo4cd.cloudfront.net/sydney.mp4"
)
Enter fullscreen mode Exit fullscreen mode

Next, create a Clip. A clip is container for different types of
assets, including the VideoAsset. We can configure different properties
like length (duration to play for) and start time (when on the timeline the clip will play from). For our clip we use
the video_asset we just created, a start of 0 seconds and a length of 10 seconds:

video_clip = Clip(
    asset = video_asset,
    start = 0.0,
    length = 10.0
)
Enter fullscreen mode Exit fullscreen mode

Setting up the image clip

Next, we need to add an image which will be the watermark on the video. We will be using the attached logo. You can replace it with your own image url.

A simple watermark

Similar to setting up the VideoAsset, let's add the ImageAsset by adding the following code.

image_asset = ImageAsset(
    src = "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/logos/real-estate-black.png"
)
Enter fullscreen mode Exit fullscreen mode

Let's configure the clip properties including the ImageAsset. We will configure length, start time, position (position
of the ImageAsset in the viewport), scale (size of the asset relative to viewport size), and opacity (transparency of
the asset). Visit the clip documentation to learn more about
clip properties.

image_clip = Clip(
    asset = image_asset,
    start = 0.0,
    length = 10.0,
    scale = 0.25,
    position = 'bottomRight',
    opacity = 0.3
)
Enter fullscreen mode Exit fullscreen mode

Adding the video clip to the timeline

Next, let's create a timeline to add our clips. Before that, we need to create two seperate tracks. We shouldn't add two clips on the same track in the same timeframe as it will not know which asset to show on top. This causes asset to flicker.

Let's add the image_clip on track_1 and the video_clip on track_2 by adding the following script.

track_1 = Track(clips=[image_clip])
track_2 = Track(clips=[video_clip])
Enter fullscreen mode Exit fullscreen mode

Next, let's add both tracks to our timeline. Make sure they are ordered sequentially based on how they are layered. If the track consisting of the image_clip is second on the list, then it won't show on the video as it will stay behind the video_clip.

timeline = Timeline(
    background = "#000000",
    tracks = [track_1, track_2]
    )
Enter fullscreen mode Exit fullscreen mode

Configuring the final edit and output

Next, we need to configure our output. We set the output format to mp4. Let's set the video resolution to hd which generates a video of 1280px x 720px @ 25fps. You can also configure other properties like repeat for auto-repeat used in gifs, thumbnail to generate a thumbnail from a specific point on the timeline, destinations to set export destinations, and more.

output = Output(
    format = "mp4",
    resolution = "hd"
)

edit = Edit(
    timeline = timeline,
    output   = output
)
Enter fullscreen mode Exit fullscreen mode

Sending the edit for rendering via the API

Finally, we send the edit for processing and rendering using the API. The Shotstack SDK takes care of converting our objects to JSON, adding our key to the request header, and POSTing everything to the API.

try:
    api_response = api_instance.post_render(edit)
    message = api_response['response']['message']
    id = api_response['response']['id']
    print(f"{message}\n")
    print(f">> render id: {id}")
except Exception as e:
    print(f"Unable to resolve API call: {e}")
Enter fullscreen mode Exit fullscreen mode

Final script

The final script is below with a few additional checks and balances:

import shotstack_sdk as shotstack
import os
import sys

from shotstack_sdk.api import edit_api
from shotstack_sdk.model.clip import Clip
from shotstack_sdk.model.track import Track
from shotstack_sdk.model.timeline import Timeline
from shotstack_sdk.model.output import Output
from shotstack_sdk.model.edit import Edit
from shotstack_sdk.model.video_asset import VideoAsset
from shotstack_sdk.model.image_asset import ImageAsset

if __name__ == "__main__":
    host = "https://api.shotstack.io/stage"

    configuration = shotstack.Configuration(host = host)

    configuration.api_key['DeveloperKey'] = os.getenv("SHOTSTACK_KEY")

    with shotstack.ApiClient(configuration) as api_client:
        api_instance = edit_api.EditApi(api_client)

        video_asset = VideoAsset(
        src = "https://d1uej6xx5jo4cd.cloudfront.net/sydney.mp4"
        )

        video_clip = Clip(
            asset = video_asset,
            start = 0.0,
            length= 10.0
        )

        image_asset = ImageAsset(
        src = "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/logos/real-estate-black.png"
        )

        image_clip = Clip(
            asset = image_asset,
            start = 0.0,
            length = 10.0,
            scale = 0.25,
            position = 'bottomRight',
            opacity = 0.3
        )

        track_1 = Track(clips=[image_clip])
        track_2 = Track(clips=[video_clip])

        timeline = Timeline(
            background = "#000000",
            tracks     = [track_1, track_2]
        )

        output = Output(
            format      = "mp4",
            resolution  = "hd"
        )

        edit = Edit(
            timeline = timeline,
            output   = output
        )

        try:
            api_response = api_instance.post_render(edit)

            message = api_response['response']['message']
            id = api_response['response']['id']

            print(f"{message}\n")
            print(f">> render id: {id}")
        except Exception as e:
            print(f"Unable to resolve API call: {e}")
Enter fullscreen mode Exit fullscreen mode

Running the script

Use the python command to run the script.

python watermark-video.py
Enter fullscreen mode Exit fullscreen mode

You may need to use python3 instead of python depending on your configuration.

If the render request is successful, the API will return the render id which we can use to retrieve the status of the
render.

Checking the render status and output URL

To check the status we need another script that will call the API to render status endpoint. Create a file called
status.py and paste the following code:

import sys
import os
import shotstack_sdk as shotstack

from shotstack_sdk.api import edit_api

if __name__ == "__main__":
    host = "https://api.shotstack.io/stage"
    configuration = shotstack.Configuration(host = host)
    configuration.api_key['DeveloperKey'] = os.getenv("SHOTSTACK_KEY")

    with shotstack.ApiClient(configuration) as api_client:
        api_instance = edit_api.EditApi(api_client)
        api_response = api_instance.get_render(sys.argv[1], data=False, merged=True)

        status = api_response['response']['status']

        print(f"Status: {status}")

        if status == "done":
            url = api_response['response']['url']
            print(f">> Asset URL: {url}")
Enter fullscreen mode Exit fullscreen mode

Then run the script using command line:

python status.py {renderId}
Enter fullscreen mode Exit fullscreen mode

Replace {renderId} with the ID returned from the watermark-video.py script. Re-run the status.py script every 4-5
seconds until the status is done and a URL is returned. If something goes wrong the status will show as failed.
If everything ran successfully you should now have the URL of the final video, just like the one at the start of the
tutorial.

Watermarked video example

We can see the watermarked video below:

Accessing your rendered videos using the dashboard

You can view your rendered videos inside the Shotstack dashboard under Renders. Videos are deleted after 24 hours and need to be transferred to your own storage provider. All files are however copied to Shotstack hosting and you can configure other destinations including S3 and Mux.

Shotstack Dashboard

Watermarking multiple videos

As you can see how easy it is to watermark a video using the Shotstack Python SDK. However, the big advantage of using the Shotstack API is how easily we can scale without having to build and manage the rendering infrastructure.

To demonstrate the scalability, we will watermark the following list of videos with the same logo we used above. You can also use other data resources like the CSV and other apps integrating with their APIs.

video_links = [
    'https://d1uej6xx5jo4cd.cloudfront.net/slideshow-with-audio.mp4',
    'https://cdn.shotstack.io/au/v1/msgtwx8iw6/d724e03c-1c4f-4ffa-805a-a47aab70a28f.mp4',
    'https://cdn.shotstack.io/au/v1/msgtwx8iw6/b03c7b50-07f3-4463-992b-f5241ea15c18.mp4',
    'https://cdn.shotstack.io/au/stage/c9npc4w5c4/d2552fc9-f05a-4e89-9749-a87d9a1ae9aa.mp4',
    'https://cdn.shotstack.io/au/v1/msgtwx8iw6/c900a02f-e008-4c37-969f-7c9578279100.mp4'
]
Enter fullscreen mode Exit fullscreen mode

The following script watermarks the list of videos. If you want to test it, create a new file called
watermark-videos.py, paste the script, and save it.

import shotstack_sdk as shotstack
import os
import sys

from shotstack_sdk.api import edit_api
from shotstack_sdk.model.clip import Clip
from shotstack_sdk.model.track import Track
from shotstack_sdk.model.timeline import Timeline
from shotstack_sdk.model.output import Output
from shotstack_sdk.model.edit import Edit
from shotstack_sdk.model.video_asset import VideoAsset
from shotstack_sdk.model.image_asset import ImageAsset

if __name__ == "__main__":

    host = "https://api.shotstack.io/stage"

    configuration = shotstack.Configuration(host = host)

    configuration.api_key['DeveloperKey'] = os.getenv("SHOTSTACK_KEY")

    video_links = ['https://d1uej6xx5jo4cd.cloudfront.net/slideshow-with-audio.mp4',
        'https://cdn.shotstack.io/au/v1/msgtwx8iw6/d724e03c-1c4f-4ffa-805a-a47aab70a28f.mp4',
        'https://cdn.shotstack.io/au/v1/msgtwx8iw6/b03c7b50-07f3-4463-992b-f5241ea15c18.mp4',
        'https://cdn.shotstack.io/au/stage/c9npc4w5c4/d2552fc9-f05a-4e89-9749-a87d9a1ae9aa.mp4',
        'https://cdn.shotstack.io/au/v1/msgtwx8iw6/c900a02f-e008-4c37-969f-7c9578279100.mp4'
    ]

    with shotstack.ApiClient(configuration) as api_client:
        for link in video_links:
            api_instance = edit_api.EditApi(api_client)

            video_asset = VideoAsset(
            src = link
            )

            video_clip = Clip(
                asset = video_asset,
                start = 0.0,
                length= 10.0
            )

            image_asset = ImageAsset(
                src = "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/logos/real-estate-black.png"
            )

            image_clip = Clip(
                asset = image_asset,
                start = 0.0,
                length = 10.0,
                scale = 0.25,
                position = 'bottomRight',
                opacity = 0.3
            )

            track_1 = Track(clips=[image_clip])
            track_2 = Track(clips=[video_clip])

            timeline = Timeline(
                background = "#000000",
                tracks     = [track_1, track_2]
            )

            output = Output(
                format      = "mp4",
                resolution  = "hd"
            )

            edit = Edit(
                timeline = timeline,
                output   = output
            )

            try:
                api_response = api_instance.post_render(edit)

                message = api_response['response']['message']
                id = api_response['response']['id']

                print(f"{message}\n")
                print(f">> render id: {id}")
            except Exception as e:
                print(f"Unable to resolve API call: {e}")
Enter fullscreen mode Exit fullscreen mode

Then use the python command to run the script.

python watermark-videos.py
Enter fullscreen mode Exit fullscreen mode

To check the render status, run the status.py file we created in the first part and run it using the command line:

python status.py {renderId}
Enter fullscreen mode Exit fullscreen mode

Replace the renderId from the IDs returned from the watermark-videos.py.

Final thoughts

This tutorial should have given you a basic understanding of how to programmatically edit and generate videos using Python and the Shotstack video editing API. As a next step, you could learn how to add other assets like text and images to build a simple media application. This is just an introductory tutorial to programmatically working with media but we can do so much more. We can use this for many use cases like

You can check out our other Python tutorials and YouTube videos to learn programmatic media and building video apps faster.

Top comments (0)

Read next

stevenacoffman profile image

HTTP/2 streaming and you, HTTP/3 streaming and wheee!

Steve Coffman -

sharminshanta profile image

Product of Array Elements Except Self in PHP

Sharmin Shanta -

rdalbuquerque profile image

Building a Cost-Effective Valheim Server on Azure with Serverless Discord Bot Integration

Rodolfo Albuquerque -

sreeni5018 profile image

From Classroom Attention to Transformer: A Journey Through Language Learning and AI Evolution

Seenivasa Ramadurai -