Tim’s AI Workflow for “The Bridge”

Tim Simmons of Theoretically Media made an CGAI (Computer Generated Artificial Intelligence) short film using Google’s new Veo 2 model:

He completes the package by taking us behind the scenes to reveal his workflow:

The software or services he used and their cost per month (or for this project)? See below:

  1. Midjourney – $30 (images)
  2. Gemini – free (prompts)
  3. ElevenLabs – $22 (voice)
  4. Hume – free (voice)
  5. Udio – $10 (music)
  6. Hedra – $10 (lip sync)
  7. Premiere – $60 (NLE)
  8. RunwayML – $30 (stylize)
  9. Magnific – $40 (creative upscale)
  10. Veo 2 – $1,500 (video at 50 cents/second)
  11. Topaz – $300 (upscale)
    TOTAL – $2,002 (plus 40 hours of Tim’s time)

In addition to the great AI news and advice, Tim is actually funny:

“At some point in the process Gemini and I definitely got into a bit of a groove and I just ended up ditching the reference images entirely. I have often said that working this way kind of feels a bit like being a writer/producer/director working remotely with a film crew in like let’s say Belgium and then your point of contact speaks English but none of the other department heads do. But like with all creative endeavours you know somehow it gets done.”

My take: Tim’s “shooting” ratio worked out to about 10:1 and there are many, many steps in this work flow. Basically, it’s a new form of animation — kinda takes me back to the early days of Machinima, that, in hindsight, was actually more linear than this process.

BONUS

Here is the Veo 2 Cheat Sheet by Henry Daubrez that Tim mentions.

1/ If you’re not using a LLM (Gemini, ChatGPT, whatever), you’re doing it wrong.

VEO 2 currently has a sweet spot when it comes to prompt length: too short is poor, too long drops information, action, description etc. I did a lot of back and forth to find my sweet spot, but once I got in a place I thought felt right, I used a LLM to help me keep my structure, length, and help me draft actions. I would then spent an extensive amount of time tweaking, iterating, removing words, changing order, adding others, but the draft would come from a LLM and a conversation I built and trained to understand what my structure looked like, what was a success, or a failure. I would also share the prompts working well for further reference, and sharing the failures also for further reference. This would ensure my LLM conversation became a true companion.

2/ Structure, structure, structure

Structure is important. Each recipe is different but same as any GenAI text-to something, it looks like the “higher on the prompt has more weight” rule applies. So, in my case I would start by describing the aesthetics I am looking for, time of day, colors, mood, then move to camera, subject, action, and all the rest. Once again, you might have a different experience but what is important is to stick to whatever structure you have as you move forward. Keeping it organized also makes it easier to edit later.

3/ Only describe what you see in the frame

If you have a character you want to keep consistent, but you want a close-up on the face for example, your reflex will be to describe the character from head to toe and then mention you want a close-up…It’s not that simple. If I tell VEO I want a face close-up but then proceed to describe the character’s feet, the close-up mention will be dropped by VEO… Once again, the LLM can help you in this by giving it the instruction to only describe what is in the frame.

4/ Patience

Well, it can get costly to be patient, but even if you repeat the same structure, sometimes changing one word can still throw the entire thing out and totally change the aesthetics of your scene. It is by nature extremely consistent if you conserve most words, but sometimes it happens. In those situations, trace your steps back and try to figure out which words are triggering a larger change.

5/ Documenting

When I started “Kitsune” (and did the same for all others), the first thing I did was start a Figjam file so I could save the successful prompts and come back to them for future reference. Why Figjam? So I could also upload 1 to 4 generations from this prompt, and browse through them in the future.

6/ VEO is the Midjourney of video

Currently, no text-to-video tool (Minimax being the closest behind) gave me a feeling I could provide strong art directions and actually get them. I have been a designer for nearly 20 years, and art direction to me has been one of the strongest foundations of most of my work. Dark, light, happy, sad, colorful or not, it doesn’t matter as long as you have a point of view and please…have a point of view. Recently watched a great video about the slow death of art direction in film (link in comments) and oh boy, did VEO 2 deliver on giving me the feeling I was listened. Try starting your prompts with different kinds of medium (watercolor for example), the mood you are trying to achieve, the kind of lighting you want, the dust in the rays of light, etc… which gets me to the next one

7/ You can direct your colors in VEO

It’s as simple as mentioning the hues you want to have in the final result, in which quantity, and where. When I direct shots, I am constantly describing colors for two reasons: 1. Well, having a point of view and 2. reaching better consistency through text-to-video. If I have a strong and consistent mood but my character is slightly different because of text-to-video, the impact won’t be dramatic because a strong art direction helps a lot with consistency.

8/ Describe your life away

Some people asked me how I achieved a good consistency between shots knowing it’s only text-to-video and the answer is simple: I describe my characters, their unique traits, their clothing, their haircut, etc..anything which could help someone visually impaired have a very precise mental representation of the subject.

9/ But don’t describe too much either…

It would be magical if you could stuff 3000 words in the window and have exactly what you asked for, right? Well, it turns out VEO is amazing with its prompt adherence, but there is always a moment where it starts dropping animations or visual elements when your prompt stretches for a tad too long. This actually happens way before the character limit allowed by VEO is reached, so don’t overdo it, it’s no use and will play against the results. For info, 200-250 words seems like a sweet spot!

10/ Natural movements but…

VEO is great with natural movements and this is also one of the reasons why I used it so extensively: people walking don’t walk in slow-motion. That being said, don’t try to be too ambitious on some of the expected movements: multiple camera movements won’t work, full 360 revolutions around a subject won’t work, anime-style crazy camera movements won’t work, etc… what it can do is already great, but there are still some limitations…

Advice for Indie Filmmakers from Stephen Follows

Stephen Follows offers Practical advice for indie directors who want to keep making films.

Here are the topics he discusses:

  1. Stay flexible and keep moving
  2. Start small but start now
  3. Think like a micro-budget filmmaker
  4. You need to create your own opportunities
  5. Learn by doing everything
  6. Understand where indie films truly live
  7. Consider if film school is for you
  8. You never know what an opportunity looks like
  9. Find your thing
  10. Learn what people want to watch
  11. Build the right team
  12. Stay ahead of the game
  13. Keep going, no matter what

I particularly like this:

“The filmmakers who get noticed are the ones who don’t wait for permission. They write, shoot, and edit their own work. Even a no-budget short is better than waiting for the perfect offer that may never come.”

My take: good advice! TL;DR? Just do it.

CAPTCHA frustration film wins 2025 Oscar best live action short

I’m Not a Robot, a 2023 Dutch-language short science fiction drama written and directed by Victoria Warmerdam, has just won the 2025 Academy Award for Best Live Action Short Film.

It stars Ellen Parren as a woman who has an identity crisis after failing a series of CAPTCHA tests.

Watch the full short film here.

See trailers for all fifteen Oscar-nominated short films.

Check out more great short films in The New Yorker Screening Room.

My take: Great concept! And great production! Watching it, I wondered how they’d execute the “jumping” scene. (A oner with a quick pan away to substitute the stunt performer.) I accomplished something similar with edits only in dust2dust. Can’t wait to see what additional themes Victoria tackles in the feature-length version!

The problem with independent film exhibition in Canada

…is Cineplex.

Writing in This MagazineJake Pitre lays bare the stark truth facing the movie-going public in Canada:

Cineplex’s role in Canada is, without a doubt, a monopoly. It runs 158 theatres with over 1,630 screens, and it controls approximately 75 percent of domestic box office. By contrast, no one company in the U.S., the UK, or Australia controls more than 30 percent.”

The Network of Independent Canadian Exhibitors (NICE) reports in its The State of Independent Film Exhibition in Canada that 60% of independent film exhibitors
operated at a loss at the end of their most recent fiscal year and highlights two policy suggestions that would improve the exhibition of films in Canada:

  1. Limit studio Clean Run demands: An overwhelming majority at 81%
    of independent film exhibitors are impacted by the “clean
    runs” required by major studios.
  2. Eliminate Zones: 53% of independent film exhibitors must
    wait for the Cineplex in their “zone” to finish playing a new
    release before they are allowed to show it.

Curiously, a new book on corporate monopolies in Canada, The Big Fix, digs into the history of film exhibition in Canada and reveals that Cineplex was the upstart challenger in the 1980s!

My take: this situation can only be addressed with government intervention, as Cineplex has a vested interest in maintaining its monopoly. It’s a pity because competition is actually good.

How to create a TV Pitch Deck

Australia’s Special Broadcasting Service (SBS) has a great Scripted Pitch Guide.

SBS Scripted is looking for the same things all broadcasters are:

“We want to hear pitches for TV series that thrill us…. Great stories, brilliantly told. Entertaining, culture defining, fresh and unique. Bring us a show we have not seen before. Bring us a show that is brilliantly crafted; a show that has characters that audiences will fall in love with; a show with a propulsive narrative that starts with a bang and keeps audiences deeply engaged.”

That’s why I believe their pitch deck format will work for almost all broadcasters.

My take: download their PDF, start reading from Page 13 and then go back and study everything! Great for all your media projects too.

Funding + Screening, together now

Brian Welk reports on IndieWire that From Film Funding to Distribution, This Partnership Gives You Control: ‘Anyone Can Do This’.

The partnership in question is between Seed&Spark and Kinema.

Unlike other crowdfunding platforms Seed&Spark is purpose-built for film & TV projects. They have a phenomenal 82% success rate.

Kinema is a tech platform that enables non-theatrical exhibition — what you may call grassroots screening tours — of films in person and online. We make it easy and rewarding for anyone, anywhere to organize moving showings and share in the proceeds.”

Brian states:

“Those who fund programs on Seed&Spark get a dedicated Kinema account manager and custom distribution consultations. Fees are waived for filmmakers with over 500 followers or over 1,000 campaign backers.”

But hey, if you want to DIY, at least check out their free online resource The Distribution Playbook.

My take: well worth further investigation!

Riffusion generates full songs effortlessly

Riffusion has just opened a public beta and it rocks!

Riffusion is the brainchild of Hayk Martiros and Seth Forsgren.

“Our goal is to make everyone into a musician and bring a future where music is interactive and personalized.”

TechCrunch reported their $4M seed funding in October 2023.

My take: damn! Not only will this create full songs, it will also create stems you can download for further modification in your DAW of choice.

Indie TV: self-financing your own series

Elaine Low writes on The Ankler that Indie TV is a viable model to self-finance shows.

She writes:

“In Park City for the (Sundance) festival is Cooper Raiff, the writer and director of Cha Cha Real Smooth, which just three years ago was a festival competition title that got snapped up by Apple TV+ for $15 million. This year, he’s shopping Hal & Harper — not an indie film, which would be a tough enough sell in 2025, but rather an independently financed TV series, whose path is even less clear.”

Zack Sharf quotes Raiff on making Hal & Harper in Variety as saying:

“(Selling your second movie for $15 million to Apple) makes you too confident. Because of that deal, I thought, ‘Let’s just do it with television.’ That’s what I told everyone. We’re really braving the storm with the series. ‘Cha Cha’ was sold on the backs of so many indie movies. With this, we’re trying to sell this show… so it has to be this undeniable thing to these streamers and these networks. But it made me too confident.”

See the Collider review.

Wikipedia info.

My take: this seems extremely risky, unless your budget is practically zero, you have all the locations, and everyone is volunteering their time. Or, you’ve made so much, you need write-offs? I mean it’s a great way to keep total creative control — until Season Two, that is.

AI feature filmmaking

AiMation Studios has released the first feature fully made with AI on YouTube, Where the Robots Grow.

It took the crew of nine ninety days to create it. They used Wonder Dynamics to motion capture actors, replace them with assets, and then skin the assets, backgrounds and skies with AI generated replacements.

Haydn Rushworth interviewed Tom Paton, the director, in Coventry, England. Here he explains how he finances all his self-described “shitty” movies:

Watch Part One, Part Two, Part Three and a bonus studio tour.

My take: the feature definitely has some rough spots. But the democratization of production using AI to assist in all areas is something only those with vested interests will resist.