- author: Matt Wolfe
Recreating the Runway ML CEO's Viral Video Tutorial Using Gen 1 Tool
Have you seen the viral video that Runway CEO shared on Twitter? If not, it's a must-watch! With over 1.7 million views, the video showcases the power of Runway's Gen 1 tool, which can create stunning deepfake videos in minutes.
As a tech enthusiast, I was curious to see if I could recreate the video myself. So, I decided to give it a try and share my experience with you. In this tutorial, I will take you through the steps involved in recreating the viral video.
Splitting the Video into Two Parts
The first step in recreating the Runway ML CEO's video is splitting the video into two parts. Since the Gen 1 tool can only process videos up to 15 seconds long, we need to split the video into two parts.
To do this, I used DaVinci Resolve, a free video editing software that works on both Mac and PC. I created a new project and pulled in the video I recorded of me walking around my house and snapping my fingers.
On my timeline, I counted how many scenes I needed to create in between each snap, which turned out to be five. I split the video at the 15-second mark and rendered out two videos: one with the first 15 seconds, and one with the last 12 seconds with a little overlap.
Generating Different Styles with Runway ML's Gen 1 Tool
With both parts of the video ready, it was time to use Runway ML's Gen 1 tool. I signed in to my Runway account and clicked on Gen 1 video to video. I uploaded the first 15-second video and waited for it to process.
One thing to note is that the Gen 1 tool will only process 15 seconds of the video you upload. So, if your video is longer than 15 seconds, you need to split it into parts like I did.
I generated multiple styles using different images of a samurai warrior, homelander, Lady Gaga, an astronaut, and Thanos. I uploaded these images to Runway and previewed the generated styles before selecting the one I liked the most.
Piecing the Clips Together in DaVinci Resolve
After generating the different styles, I downloaded each video clip and pulled them all into DaVinci Resolve. I lined up the clips and made sure the samurai clip started at the beginning and the astronaut clip ended at the last snap.
One thing I noticed was that Runway claims it can process up to 15 seconds, but it actually cuts it at around 14.2 seconds. So, if you want to match the timing perfectly, you need to keep this in mind.
Creating a Synthetic Deepfake Video with AI
If you've been on social media lately, you may have seen some impressive deepfake videos. Deepfakes are synthetic videos or images created with the help of artificial intelligence (AI), where one person's face or body is swapped with another person's, making it seem like they are saying or doing something they are not. In this article, we will walk you through the process of creating a deepfake video using AI.
To create a deepfake video, you will need to gather several clips that you will merge together. For this example, we will be using five clips: a personal clip, a samurai clip, a Homelander clip, a Lady Gaga clip, and an astronaut clip. It's essential to have these clips ready before you start working on the deepfake.
Once you have gathered your clips, you will need to align them. In this case, we will need to align the clips to a specific sound, known as a "snap." First, we will line up the personal clip to the snap, followed by the samurai clip, then the Homelander clip, the Lady Gaga clip, and the astronaut clip. Finally, we will end it with the Thanos clip and then back to the personal clip.
Now that you have aligned your clips, it's time to add some voiceovers. We use an AI voice generator from 11 Labs to create voiceovers. We create three different voice clips with three different accents, male older, American, and female young. We add the voice clips to specific clips depending on which accent we want, and then we download the clips.
To give your deepfake video a final touch, you need to add some music. We use a subscription to Motion Array, which provides different royalty-free music styles to choose from. We also change the music to match each clip to make it seem more natural.
How to Create a Video Like a Pro: The Costs Add Up
When it comes to creating professional-grade videos, the costs can add up quickly. In this tutorial, we learned how the creator used various services to produce a 30-second video.
Here are the services and costs involved:
- Motion Array: $30/month for background music.
- 11 Labs: Used for other stuff.
- Runway: Used from time to time. This is where most of the costs add up since it's a sunk cost in other services used, and creating a video can quickly eat at your credits.
It's important to note that while the costs can add up, creating a quality video can lead to millions of views and even viral success.
If you're interested in experimenting with AI tools to create your videos, check out Futuretools.io. This is a website where all the latest AI news is curated, so you can stay up-to-date and inspired.
Don't forget to tag the creator on Twitter if you use their strategy, and make sure to like and subscribe. Thanks for tuning in!Recreating the viral video using runway's gen 1 tool was a fun and exciting experience. while my deepfake videos may not be as polished as the original, they still turned out pretty impressive, and i learned a lot about the possibilities of deepfake technology.
if you're interested in trying out deepfake technology for yourself, i highly recommend giving runway ml a try. with their user-friendly interface and powerful tools, you'll be able to create stunning deepfake videos in no time.
Creating a deepfake video might seem complicated, but there are ai tools available that can make it much easier. the cost can stack up, but it can be a fun way to experiment with ai and hone your video editing skills. ultimately, the ethics behind deepfake videos are concerning, and it's essential to be responsible and use them with care.