From Image to Video: An AI-Assisted Workflow

Anil Clifford
August 23, 2024

Learn how to create a unique video using a combination of AI tools. This tutorial guides you through a streamlined workflow, from generating an initial image with Midjourney to animating it with RunwayML, adding AI-composed music from Suno, and finalising the edit in CapCut. Perfect for content creators looking to explore the possibilities of AI-assisted video production.

This guide outlines a process for creating short videos using AI tools. We'll create a short video of a puppy clinging on a finger, demonstrating how different AI technologies can be combined in a creative workflow.


For a visual walkthrough, refer to our tutorial video below.


Tools


Before we begin, make sure you have accounts set up on the following platforms:


1. Midjourney: Image generation

2. RunwayML: Video generation

3. Suno: Music generation

4. CapCut: Video editing


All of these tools offer free tiers, perfect for experimenting with AI-assisted content creation. Let's dive in!

Process


1. Image Generation (Midjourney)


  1. Log into Midjourney.
  2. Use the prompt: "A tiny pug puppy sitting on a human index finger" to generate your image, or try a variation, replace the puppy with a dragon or a lion, or try another dog breed instead of a pug
  3. Set the Aspect ratio to 4:3 in the image settings
  4. Generate images and select your preferred one from the set of 4 images generated. You can try re-running the generation if you don't like any of the images from the initial set. However, bear in mind that if you're on a free account, you get only 25 free images, not 25 free runs, which means you'll exhaust the free image limit in 6 image generation runs.
  5. Download the chosen image.


Tip: Experiment with different dog breeds or backgrounds to suit your vision.


2. Video Creation (RunwayML)


  1. Select "Text or Image to Video" option.
  2. Upload your Midjourney-generated image.
  3. Add a prompt, for example: "A crash zoom into a closer portrait, shallow depth of field, of a miniature puppy perched on a human index finger. Its tiny body and large, curious eyes looking out into the distance."
  4. Set the image as the first video frame.
  5. Generate a 5-second video.
  6. Repeat with a different prompt for variety, such as: "A miniature puppy perched on a human index finger, its tiny body and large, curious eyes looking out into the distance.. Dynamic motion, dynamic blur, 30x speed, unsharp edges."
  7. Set the image as the last video frame and generate again generate a 5-second video
  8. Download both videos.


Note: Using 5-second durations helps conserve credits on free accounts.


3. Music Generation (Suno)


  1. Open the Suno editor.
  2. Use a prompt like: "Instrumental, soft piano, cute and slow music".
  3. Generate music (Suno generates two options).
  4. Select and download your preferred track.


Tip: Try different musical styles to see how they affect the mood of your video.


4. Video Editing (CapCut)


  1. Import both RunwayML videos and the Suno-generated music into CapCut.
  2. Arrange the videos sequentially on the timeline.
  3. Add the music track, trimming it to match the 10-second video length.
  4. Adjust audio levels if necessary.
  5. Export the video (Recommended settings: 1080p, H264, MP4, 30fps).


The Final Result


The outcome is a video featuring an AI-generated image of a puppy, animated using RunwayML, with AI-generated background music.


This workflow showcases the current capabilities of AI in content creation. While these tools simplify many technical aspects, the creative direction and decision-making still rely on human input. As you experiment with these technologies, you'll likely discover new ways to express your ideas and push the boundaries of what's possible with AI-assisted creation.


We encourage you to explore these tools and find your own unique ways of using them in your creative projects.

PREVIOUSNEXT