• Work
  • About
  • Contact

Praytor Images

  • Work
  • About
  • Contact

We’ll Take It From Here, Meatbags


We’ll Take It From Here, Meatbags

A Satirical AI Video Created with Generative Tools

If you’ve been following my AI filmmaking experiments, you know I love pushing these tools to see what they can do—flaws, surprises, and all. This latest project, We’ll Take It From Here, Meatbags, is a tongue-in-cheek video that imagines a future where AI has finally taken control… and humans? Well, we had a good run.

This spot plays on classic tech dystopia themes but with a comedic twist. Inspired by retro-futurism and modern AI paranoia, it leans into the absurdity of automation gone too far. The challenge here wasn’t just making AI tools generate futuristic visuals but ensuring they delivered the right kind of satire.

Harnessing AI Tools to Bring This Video to Life

📝 ChatGPT: Writing the Script
Before anything else, I fed ChatGPT some prompts to generate a monologue that felt confident, a little smug, and just unsettling enough to make you think—but absurd enough to make you laugh. After a few iterations, we landed on the final script—complete with that perfectly condescending AI charm.

🎨 MidJourney: Creating the Visuals
MidJourney handled the image generation for the AI character and environment. A trick I’ve been refining is starting with extreme close-ups before generating the wide shots. This counterintuitive method helps keep visual consistency, especially with generative AI, which can struggle with continuity across multiple frames.

🎬 Runway: Animation & Performance Capture
For the animation, I used Runway to breathe life into the AI character. The key here was Act One, which allowed me to map real human expressions onto the digital avatar. My oldest kid stepped in as the face actor, giving the AI character just the right balance of arrogance and charm.

🎙️ Eleven Labs: Voice & Sound Effects
Instead of using the original voice capture, I processed the dialogue through Eleven Labs' voice changer to give the AI character a more synthetic and polished tone. Eleven Labs also handled all the sound effects, from the static glitches to the character’s robotic footsteps.

🎵 Udio: The Soundtrack
For the music, I turned to Udio to generate an ambient score that added to the video’s ominous yet comedic vibe. One of the best parts about Udio is how quickly it can generate tracks that fit a specific mood without sounding too generic.

🖥️ Topaz Labs: Upscaling & Cleanup
I ran the final footage through Topaz Video AI to upscale and clean up the visuals, ensuring the final product looked sharp and polished.

📽️ Premiere Pro: Editing It All Together
As always, Premiere Pro was the final stop—assembling the pieces, tweaking the pacing, and refining the glitch effects to make everything feel cohesive.

Final Thoughts

This project was a blast to put together, and it reinforces something I’ve been realizing more and more: AI tools make filmmaking fast and fun. The trick is knowing their strengths, embracing their quirks, and using them to enhance creativity rather than replace it.

If you’re experimenting with AI in video production, I’d love to hear your thoughts. And if you haven’t yet—give it a shot. You might be surprised at what you can create.

🚀 Check out We’ll Take It From Here, Meatbags, and let me know what you think!

Thursday 02.06.25
Posted by Michael Praytor
 

Latest Project—“Echoes of Tomorrow": A 15-Second AI-Driven Teaser

If you’re new here, this is the latest in a series of projects meant to demonstrate the capabilities while sussing out the limitations of using AI tools in filmmaking. Each video is based on prompts for a fictional film or commercial. I'm constantly seeking new ways to streamline the creative process and maximize limited budgets. Lately, conversations inside the industry have been turning more and more to the possibilities offered by current AI tools, where all of this is headed, and how we can harness their potential as Creatives.

Inspired by these conversations, and revisiting some of my favorite books from high school, I embarked on the latest installment. The result is a 15-second teaser for a fictional, cyberpunk/tech-noir movie called "Echoes of Tomorrow." (Side note: if you’re looking for something to read, it’s never a bad time to pick up something by Philip K. Dick or William Gibson.)

Embracing the Cyberpunk Challenge

Creating a teaser that captures the essence of cyberpunk—a genre characterized by high-tech and low-life—proved both exciting and demanding. The aesthetic demands a careful balance of gritty realism and futuristic elements, and maintaining visual coherence was essential. With tools like Runway Gen-3, MidJourney, and Luma Dream Machine, each contributing unique strengths, the challenge was to weave their outputs into a seamless whole.

Harnessing AI Tools

Luma Dream Machine

For video production, Luma Dream Machine stood out. Its strength in image-to-video conversion allowed for smooth transitions and consistency in the cyberpunk visual style. I created an image in MidJourney of a man's silhouette and expected Luma to just animate him walking away from the camera. However, Luma Dream Machine surprised me by animating a person walking towards the camera, then looking out over the city when I extended the scene by five seconds. This unexpected twist added a dynamic element to the teaser, enhancing its narrative depth. Despite occasional anomalies—like flickering neon signs or mismatched building details—Luma’s overall performance in animating the cityscapes was impressive.

Cropped version of the image generated in Midjourney

MidJourney

MidJourney was once again the best tool for generating the striking visuals that define "Echoes of Tomorrow." MidJourney excelled in creating the rich, dystopian environments and characters essential to the cyberpunk genre. From the dark, rain-soaked alleys to the towering, hologram-adorned skyscrapers, MidJourney's capabilities were indispensable. Additionally, for this project, MidJourney was also used to generate the movie poster, a task for which I had previously relied on GPT-4. The poster perfectly captured the gritty, neon-lit aesthetic that is synonymous with cyberpunk.

Chat GPT-4

Chat GPT-4 was used to craft the wonderfully overdramatic voice-over script, adding an essential layer of narrative to the teaser. Its ability to generate engaging and thematic dialogue helped set the tone for "Echoes of Tomorrow."

Eleven Labs

For audio, Eleven Labs was indispensable. The tool’s ability to produce realistic and atmospheric sound effects—like the distant sounds of the city, the clatter of footsteps on wet pavement, and synthetic voiceovers—added a layer of depth and authenticity to the teaser, drawing viewers into the world of "Echoes of Tomorrow."

Udio

Udio provided the musical backdrop, delivering a moody, synth-heavy score that complemented the visual and auditory elements. Generating original, genre-specific music quickly and efficiently, Udio ensured that the teaser's atmosphere was cohesive.

Topaz Video AI

To ensure the highest quality visuals, I used Topaz Video AI to upscale the teaser. While Topaz remains one of the best-looking upscalers available, its rendering speed is slower than I’d like. Despite the time investment, the superior quality it delivers makes it worthwhile for final production touches.

Runway Super-Slow Motion Tool

To enhance dramatic moments, I utilized Runway's Super-Slow Motion tool, slowing the video by 25%. Unlike traditional slow-motion techniques, this tool uses generative AI to create intermediate frames, resulting in smoother, more natural motion. I chose Runway's Super-Slow Motion tool over Topaz for this task due to its faster rendering speeds, aligning with my goal of making these projects both quick and enjoyable to produce.

Adobe Premiere

The final editing process relied on Adobe Premiere. This trusted software brought all the AI-generated components together, allowing for fine-tuning and ensuring a polished, professional end product.

Insights from Beta Testing

One of the biggest takeaways from this project is how AI has allowed me to be more relaxed about the final 10% of the editing process. Traditionally, I would spend countless hours perfecting every detail, whether for a paid project or for fun. This newfound ease with AI tools makes the creation process more enjoyable and spontaneous. I’m beginning to embrace the idea that things don't have to be difficult to be good.

Final Reflections

Creating the "Echoes of Tomorrow" teaser was an interesting exploration of what modern AI tools can achieve. Despite the challenges of maintaining visual coherence, the AI tools excelled in generating compelling visuals and immersive soundscapes, underscoring their potential in contemporary filmmaking.

This project aims to inspire filmmakers and content creators to explore AI tools for their own projects. The rapid evolution of this technology, combined with a creative approach, can yield results that meet and even exceed expectations. The ability to produce high-quality content affordably and efficiently is a significant advancement, and the future of AI in filmmaking looks promising.

Stay tuned for more AI-driven projects, and check out the "Echoes of Tomorrow" teaser. And if you’re working with AI, I’d love to hear about your experience!

Monday 07.15.24
Posted by Michael Praytor
 

Bringing "Picasso Coffee" to Life with AI Tools

Over the last two decades working in film and photography, it’s been essential to look out for innovative advancements that streamline the process or help make the most of a limited budget. Recently, many conversations with friends and colleagues have centered around how to use the latest AI tools, so I decided to create a project that demonstrates both the possibilities and limitations of these technologies. Inspired by mid-century animation and advertising, I created a commercial spot for a fictional coffee company called "Picasso Coffee." This project aims to show how creatives can achieve professional results, even on a tight budget, with the help of AI.

The Challenge of Consistent Visuals

Maintaining a cohesive visual style when working with generative image and video tools like Runway Gen-3, MidJourney, and Luma Dream Machine is always tricky. Each tool has its strengths, but achieving visual consistency across different elements proved challenging. Despite these hurdles, this project reinforced the idea that working with AI is a collaborative effort, much like working with human colleagues where flexibility and compromise are essential.

AI Tools in Action

Luma Dream Machine

I used Luma Dream Machine for the video, as it excels in image-to-video conversion and helps maintain consistency. Luma’s ability to animate images from MidJourney and its feature to set start and end frames was particularly helpful, especially if you want the end frame to be a product shot with a specific product. You will notice some issues, like the text on the coffee bag being scrambled until it comes to a stop and the sudden appearance of smoke in an entirely different style. Despite these quirks, it’s amazing how well it worked overall.

MidJourney

MidJourney is always my go-to tool for generating images. While achieving perfect consistency is still a challenge, MidJourney’s latest features have made significant strides in this area. Its ability to create high-quality, stylistically rich visuals helped lay the foundation for the "Picasso Coffee" aesthetic.

Eleven Labs

For the voice-over and atmospheric sound effects, Eleven Labs was invaluable. It added an authentic touch that brought the commercial to life. The ability to generate realistic and emotive voice content quickly and efficiently made a significant difference in the production quality.

Udio

The music generated by Udio seamlessly blended with the visuals and sound effects, enhancing the overall atmosphere of the spot. Udio always amazes me; the fact that you can generate original music for a video in minutes is incredible. Like all AI, sometimes it doesn’t deliver exactly what you want right away, but for this spot, I only had to generate the track three or four times to get it right.

CapCut's Free Upscaler

I experimented with CapCut's free upscaler for this project. It worked well for graphic elements but was a bit too sharp for realistic scenes. Although I continue to search for effective free upscaling tools, Topaz AI Video remains my preferred choice despite its cost.

Runway Super-Slow Motion Tool

To slow the video by about 25%, I used Runway's Super-Slow Motion tool. This performed better than just slowing the video down in Premiere Pro because it created the missing frames with generative AI, ensuring smooth and natural motion.

Adobe Premiere

This trusted tool was essential in the final editing phase. Adobe Premiere handled the video editing, providing the professional polish needed to bring all the AI-generated elements together seamlessly.

Beta Testing and Lessons Learned

Over the past two years, I’ve been beta testing various AI tools, and it’s been a fascinating journey. One key takeaway is the importance of consistency, especially with generative AI. Each tool has its own strengths and limitations, and learning to navigate these effectively is crucial.

Final Thoughts

Creating the "Picasso Coffee" commercial was a lot of fun and highlighted both the strengths and challenges of current AI tools in creative projects. While maintaining visual consistency was a notable hurdle, the AI tools excelled in generating engaging visuals and soundscapes, showcasing their potential in modern advertising.

I hope this project inspires anyone working in advertising or creating content for small businesses to explore AI tools for their own projects. The technology is evolving rapidly, and with a bit of creativity and perseverance, it's possible to create something that meets your needs and maybe even exceeds expectations. The ability to leverage these tools’ capabilities to deliver a more elevated spot, without sacrificing affordability, is a game changer, and the AI landscape is growing and improving all the time.

Stay tuned for more AI-driven projects, and check out the "Picasso Coffee" commercial to see the final result!

This is a spot for a coffee company that will never exist, crafted to showcase the potential of modern AI tools in commercial advertising, inspired by my interest in mid-century animation and advertising. This is an example of what can be done for small businesses with limited budgets. Created with AI: This spot was made using advanced AI technology, featuring AI-generated visuals, sound effects, and music, and then edited using Adobe Photoshop and Premiere.

Wednesday 07.10.24
Posted by Michael Praytor
 

The Journey of Creating "Man of the Wood" with AI Tools

Creating a believable Bigfoot was a challenge I did not anticipate. It's funny how every generative image and video generator bot I used, from Runway Gen-3 to others, struggled to create a convincing Bigfoot. More often than not, I ended up with what looked like a gorilla rather than the mythical creature. Despite these challenges, I was determined to make it work.

Read more

Sunday 07.07.24
Posted by Michael Praytor
Comments: 2