I’m doing something different for my contribution to
Symposium this month: releasing a short film I’ve been nurturing on the film festival circuit for a year and a half.Ominous Horizon was shot two years ago this week. My wife’s friends from El Salvador were visiting and one of them is an actor. Her idea was that if we had a performer staying with us, we should use the opportunity to shoot something. She gave me the story she wanted, I banged out the script overnight, and when Teda and America arrived we went to the already surreal landscape of Corona Park and shot a guerrilla production (which I’ve always wanted to do once I saw those amazing flying saucer-like World Fair structures from the nearby highway).
It’s fitting to release Ominous Horizon on Substack because during the post-production, many members of the Soaring Twenties Social Club lent their voices, literally, to the production of the film, answering a call-to-action to record radio announcer snippets for the opening scene.
Post-production lasted far longer than production on this short, actually taking most of my free time and attention for several months. This timesuck was somewhat due to my experimentations with the video capabilities of Stable Diffusion that I accessed through Google DeForum. This was early ChatGPT-causing-the-world-to-freak-out era AI tools and I wanted at least to take a good look at the tools. The results were frequently underwhelming: inconsistent clip-by-clip, often blowing out of one style within a clip to become a different style within it, and even when I got around the area I was looking for (using a lot of blind guess-and-check work), oftentimes the framerate or timing of the output video was off.
In the end I found a satisfactory middle ground that one VFX person I showed Ominous Horizon put succinctly: “I actually don’t mind that the tracking doesn’t match.” A lot of how I personally view any sort of algorithmic video generator comes from my background as a glitch artist: it’s a processing tool that gives you assets, not final results. You have to take those assets and make them into something more under your own craft and vision. This falls far outside the expectations executives and tech people are envisioning “AI” capable of doing. Anyone who has heard of ‘processing art’ knows that there’s a relationship between the artists’ use of randomness (or often pseudorandomness or often very visibly not-even-randomness) and how the artist pulls it back into a form and function, and AI isn’t going to replace that relationship. Rather, it’ll give that relationship to its users in new media, rather than replace older media.
My main objection to AI models still remains that if they are trained off of artists’ copyrighted work, then they are, effectively, a form of stock library and should pay royalties. Despite the tech industry’s claims otherwise, I fully believe it is possible to do technically and also that there is a profitable way to do it. I also am not holding my breath that it will happen.
Anyway, the use of Stable Diffusion in this short film has possibly affected its acceptance to film festivals. I’m not sure, as there are a lot of conflating variables, and oftentimes festivals can feel like black boxes.
What I do know is that I’ve gotten shorts like Dry Beach into quite a few film festivals, and it has no actors or visual effects or composer for the score or nearly any of the amount of collaboration and work I put into Ominous Horizon, which feels counterintuitive. Storywise I rate them as roughly about the same. Yet Dry Beach showed at 9 different festivals and a handful of public showcases, a few outside New York, and Ominous Horizon is closing at only one festival and two public showcases, all in New York.
Andrea offered that I should cut the credit slate I give to Stable Diffusion to see if it gets its acceptance rate up, but I actually never got around to it — I’m already focused on my next short, Shroomery, and besides, I’m not all that put out by festivals deciding to reject anything that stinks of AI, whether out of personal editorial distaste for the tools or just fear the audience may complain. It’s also possible that Ominous Horizon was accepted less because it’s just longer and thus harder to program, or because post-pandemic ramp-up of indie filmmakers getting back in the game has resulted in a more competitive surge of short films than Dry Beach competed against. Who knows?
This is why I am sharing my film for “E_ect”: because storywise it’s about the Elect, and BTS-wise this short film is about the sElect. The fact that the theme coordinates with the two-year shootiversary of the short film also pleases my inner sense of symmetry.
Before I go, a last note about the showings of Ominous Horizon: the third and final public screening will be at FilmShop Presents: Last Day on Earth!
Date: November 9, 2024
Time: 5 PM (doors open at 4:30 PM)
Location: Chelsea Music Hall, 407 W 15th St, New York, NY 10011
If you’re in New York City, please swing by! I also edited another short film that is playing there entitled Fright Cardio!
This short film is presented for the Soaring Twenties Social Club (STSC) Symposium. The STSC is a small, exclusive online speakeasy where a dauntless band of raconteurs, writers, artists, philosophers, flaneurs, musicians, idlers, and bohemians share ideas and companionship. Each month STSC members create something around a set theme. This cycle, the theme was “E_ect.”
If you are a writer, you might consider joining us.
To experience my previous STSC Symposium submissions:
To watch my previous video posts:
Ominous Horizon