Amplifying AI Technology - I created an AI Music Video in Under Six Hours

Working across multiple platforms unveils how amazing and consistent AI has become

CLICK ON THE VIDEO ABOVE TO WATCH THE FULL MUSIC VIDEO:

After checking out the “VIDEO SHORT” section of KLING AI’s community gallery, I thought I would give music video editing a try. After all, I’ve created an AI modeling agency and produced an AI movie trailer for a paying client of mine, so why not a music video?

First, I have been a song writer for over twenty years, with 19.5 of those years placing the opportunity on hold. Thanks to the AI capabilities of SUNO AI, I’ve been able to reignite this passion, taking my lyrics and creating fully produced songs out of decades old lyrics (in some cases).

In the last month, I have been furiously scribbling down new lyrics, and the song “GLASS HOUSE” was one of those new songs.

Using SUNO AI, I have developed numerous songs into full blown productions and have a SOUNDCLOUD page now created. (click here to check out numerous songs)

In SUNO AI, you are now able to generate “PERSONAS,” which allow the user to craft songs in similar voices and styles. What I’ve been able to do with my AI Models through Midjourney’s CREF function, I can now do with music artists using Personas in SUNO AI.

Speaking of CREF in Midjourney, the first step was to create my “music artist.” Prompting a young female singer with dark hair and dark makeup in Midjourney, produced the following image.

Not bad for a first generation. I could have brought the image into MAGNIFIC AI and improved the realism and facial quality, but I decided to work mostly with native image generations.

Midjourney’s CREF feature radically improved, when the platform moved from version 6.0 to 6.1. I found that with just one image prompted as a —CREF I was able to keep my character consistent throughout numerous generations.

The song is called GLASS HOUSE, so I needed to create the house that would serve as the main backdrop. Prompting a luxury glass house on an ocean bluff produced the following:

By adding an —SREF setting using the photo reference of my AI character, I was able to keep the overall tone of this photo consistent to the woman image generations.

Then, I just needed to create a few of my interior shots:

Again, note the consistency is tone and color. In fact, I never touched the color settings in ADOBE PREMIERE, which I used to put the whole video together.

Back to the process.

The next step was to animate the imagery, so I went into KLING AI and used the VIDEO GENERATION feature. First, you need to upload an image, and then prompt the image using video language. Here’s an example of one of the videos:

From there, it’s a quick add with the LIP SYNC button, to add voice over, or, in this case, a section of the song MP3 file. Within minutes, your character will regenerate and mouth the words in your MP3 file.

Not bad, huh?

Of course, it’s only set for five seconds of spoken (or sung) speech and it is not always perfect. But for a first try, and a limited time frame (Thanksgiving afternoon), it did the trick.

Below, I’ll share all of the artistic programs I used to create this video. If you are interested in having me create anything similar for you, or you want to learn more, just DM me and let’s start communicating!

Programs Used:

Midjourney - Image generation, mostly using CREF and SREF.

Kling AI - Video Generation and Lip Sync.

Runway GEN 3 - Runs faster than Kling for video generation, but KLING has superior lip sync capabilities, so it depends on the need.

Freepik AI - Now offers nearly all the abilities of the other three AI platforms, but does not have Lip Sync capabilities… yet.

SUNO AI - For the original music composition generation, and Persona Development

The bottom line harkens to a phrase I have used often:

“AI might not take your job, but someone who knows how to use AI just might.”

I have thousands of hours into these platforms, and have even created UDEMY COURSES on AI realism and fashion. Through all this time and study, AI opened me up to creative domains that just weren’t possible in the past.

Chances are a music video like this would cost millions of dollars. I completed it in six hours, for about 10 bucks in time on this platforms.

I wonder what that is worth?

Be sure to watch the full video at the top of this article if you haven’t already… I’m pretty proud of it!