How I Used AI to Make a Music Video By Myself
In the music industry, music videos are a crucial element of promoting an artist's work. Music videos are often made with the help of videographers and actors, but what if there was a way to make a music video using just AI and basic editing software? In this blog post, we'll dive into how one artist created an animated music video using AI-generated images and simple editing techniques.
The artist in question is a small, independent artist who goes by the name "Assuming I Survive". They recently released a music video for their song "Bring it Back", which was entirely animated and made with the help of AI-generated images. While AI-generated art is becoming more common, creating complex animated sequences with AI is still in its infancy. However, Assuming I Survive was able to successfully use AI-generated images to create a unique and visually stunning music video.
The artist began by using an AI tool called "Mid Journey", which is prompt-based and uses programming language to create images. They used key prompts like "AR 16x9" to ensure the correct framing for the shot, and often used a reference image prompt or an artist's name to maintain consistency in the art style. The artist also had to consider color, and continuously told the AI where and when color could be used. While there were instances where this process didn't work as planned, the artist was able to generate over 1400 environment prompts to create a cohesive and consistent background for the video.
Once the backgrounds and environments were created, the artist used AI-generated images to create characters and people to go in them. They sometimes used the same environment prompts and then used the image prompt of the environment to generate people that would fit in it. For example, a woman playing a violin was created using various prompts, and the artist was able to build an animated sequence of movement that looked like it was always the same figure.
To add movement and feel to the individuals in the scenes, the artist used masking, a bit of Motion VFX M Puppet animation, and Gaussian blur to simulate focus shifting. They also used composite elements like dust, film grain, smoke, light leaks, and lasers to add a feeling of movement to the frames. These elements were obtained through a subscription to Envato Elements, which provides access to a massive library of resources for a project.
While the AI-generated images provided the base for the animated sequences, the artist used simple editing techniques to manipulate them and make them unique. For example, they added additional parts, manipulated areas of the images, and created unique animations with the help of Final Cut software.
Overall, Assuming I Survive was able to create a visually stunning music video using AI-generated images and basic editing software. While there were limitations to what the AI could do, the artist was able to use their creativity and ingenuity to make the most of what was available. With the continued development of AI-generated art, we may see more and more artists using this technique to create unique and visually striking music videos.
Its been a year since my journey, And things have come a long way
AI technology is rapidly evolving, and we can expect it to continue to make significant strides in the field of video creation. There are a few ways in which AI could evolve to allow for seamless video creation:
Increased automation: As AI algorithms become more advanced, they will be able to automate more and more of the video creation process. This could include everything from selecting the best shots and editing the footage to adding special effects and color grading.
Natural Language Processing (NLP): NLP is an area of AI that focuses on enabling machines to understand and interpret human language. As NLP technology improves, we may see AI systems that can take a script or a set of instructions in natural language and automatically create a video that matches those specifications.
Improved image and video recognition: One of the key challenges in video creation is identifying the best footage to use from a large collection of raw footage. As AI algorithms become better at recognizing images and videos, they will be able to more accurately select the best shots to use in a video.
Better integration with existing video creation tools: Many video editors already use AI-powered tools to assist with tasks like color grading and image stabilization. As AI technology continues to evolve, we can expect to see even tighter integration between AI and existing video editing software.
Generative AI: Generative AI is a type of machine learning that enables machines to create new content based on existing examples. As this technology continues to improve, we may see AI systems that can automatically generate entire video sequences based on a small set of inputs.
In conclusion, AI technology is already making a significant impact on video creation, and we can expect to see even more advances in the coming years. With improved automation, natural language processing, image and video recognition, and generative AI, video creation will become increasingly seamless and efficient.