OUR VIEW ON AI'S ROLE IN VIDEOGRAPHY
Is Hybrid the Future?
Every filmmaker has a list of things they wanted in a shot and could not get. The location. The set. The costumes. The weather. The light. A second person to help. Most of the talk about AI in video is about one question: will it replace the shoot? We think that is the wrong question. The better question is what happens to that list.
A videographer can already make a good, watchable film without any AI. But there is always more in their head than what ends up on screen. With more time, more money or more energy, they would have gone further: a better location, better light, a bigger cast, a bolder camera move. Somewhere along the way, the vision was cut down to fit what was possible.
This is what we mean by the hybrid workflow. You keep doing the parts you know and love. The parts of your vision that never made it into the frame get completed afterwards with AI. You shoot real footage, and the final video is generated from that footage. We believe this is the best way to use AI in video, and every example on this page was made this way.
Look at the example above before reading on. One side is what we shot. The other is what we published. What we decided ourselves is on both sides. What we could not have is only on one. The rest of this post explains why it works.
THE MECHANISM
The model fills whatever you leave empty
Most of us met AI video the wrong way. It was shown to us as an all or nothing technology: you type one prompt, and the entire video comes out the other end. Honestly, for people like us, that was hard to stomach. We did not need something we already do well, and enjoy doing, to be taken over.
What we did not see at the time is that we had met one extreme end of a very wide range. At that end, you type a prompt and the whole video is generated, and for someone with no camera at all, that end is a real door. But AI in video is not a switch. It is a range, and you can use as little or as much of it as you want. That was not how it was presented, so most of us never looked for the middle.
The rule underneath is simple. The model fills whatever you leave empty. Give it only a sentence, and it fills everything: the camera, the actor, the timing, the room. Give it a finished film, and it has nothing left to fill. Everything between those two ends is hybrid, and where you stand is a decision you make shot by shot.
Everything you bring to the shoot is one less thing the model has to invent.
THE HYBRID WORKFLOW
Start from the video you already have
You shoot the scene for real. Real camera, real lens, real movement, real timing, and whatever is in front of the lens is really there. Then you give that footage to the AI and ask it to complete the shot: change the environment, the costumes, the props, the lighting, the time period, even the character, while keeping the motion and timing of what you filmed. One honest note on that last point: changing who appears on screen is something you agree with your performer first, not something you do to them.
If you have ever worked with film, the comparison is obvious. The shoot is your negative. The AI pass is your darkroom. The negative holds everything that was true about the moment. The darkroom decides how the world sees it. Nobody calls the darkroom the author of the photograph, because the photograph, the important part, was already made.
The examples on this page were shot with the means we had at the time: a studio, our own equipment, a small team. They did not look like the films we had in mind, and they were not supposed to. Our effort went into the camera movement and what happened in front of the lens. Look at how cleanly the finished version connects back to the footage.
Everything that carries over from one side to the other is a decision we made on set.
WHY IT WORKS
The most precise way to say what you want
Creative control comes down to one thing: how precisely you can tell the model what you want. And the most precise language you have is not words. It is footage.
The model reads your footage as structure: motion, timing, framing, the geometry of the scene. All of this is kept. What the model creates is appearance: surfaces, light, faces, environments, atmosphere. So when you give it footage, you are not describing the shot. You are showing it. There is nothing left to misunderstand about where the camera is, what happens in front of it, or when.
This matters even for people who came to AI video without a camera. Working from words alone looks like the easier path. But everything you want has to be written down, and the model has to guess from your sentences. You try, you look, you adjust, you try again. AI video is not free, so every one of those tries has a cost.
A camera, any camera, changes that. Filming a reference is not extra work. It is the most direct way to express what you mean. The more you can collect from real life, the less the model has to guess. For most films, that turns out to be both better and cheaper than building everything from text. It is worth thinking about, even if you never planned to pick up a camera.
The consistency between the footage and the result is not something we had to work for. The same frames went in.
THE HUMAN CORE
The gift: reality survives
There is one more thing this workflow gives you, and nobody puts it on a budget sheet.
Let us be honest about the discomfort many of us feel when watching video where nothing was recorded. It is not the visual errors; those are disappearing fast. It is that there is no real moment behind the film. Nobody stood anywhere, nobody looked at anyone, nothing actually moved. Audiences feel this even when they cannot name it. The worry is not that AI makes video. The worry is that video stops being connected to anything that really happened.
If you have no way to shoot, that connection is not available to you, and your work has to convince people in other ways. A lot of it does. But if you have a camera, you can keep that connection, and keeping it costs you almost nothing. In every example on this page, the movement you see is the movement that happened, and the timing is real timing.
How much reality you keep is a dial, not a switch. At one end, you can use this as an invisible compositing tool: keep the shot almost exactly as filmed and change only what a VFX studio would have changed, a background, a prop, a sky. At the other end, you can transform everything visible and keep only the human core: motion, performance, and time. Where you sit on that range for any shot is a decision that belongs to you.
THE CAMERA MOVE
Where the camera goes is part of the story
Where the camera is, and how it moves, is not decoration. It is one of the main ways a film speaks to its audience. A camera move works on people the way a performer does. It tells them where to look and in what order. It tells them what matters and what is only in the background. It sets the mood, and it decides whose eyes they are seeing through. It works whether the audience notices it or not.
This is also most of what videography is. Take away the location, the set, the costumes and the lights, and what remains is the craft of the frame: blocking, framing, and moving the camera. Deciding what each frame holds, when it moves, and where it goes next. That is what makes it film instead of theatre, and it is the part a videographer has spent years learning.
So when you hand parts of a production to AI, this is the part worth keeping in your own hands. A camera move can be generated, and it will often look convincing. But the same logic that applies to footage applies here. The most precise version of the move you want is the one you perform. Your timing, your pause, your speed into the turn. A move made with a real camera carries the weight of the equipment, the hand that started it, the room it travelled through, and none of that has to be described. It is simply there.
We should be open about where we stand. edelkrone was built on one belief: that the frame is part of the story, and that where the camera goes shapes what the audience feels. Every product we make comes from that. So we have an obvious interest in this argument. We are making it anyway, because it is the thing we believe most, and defending it is both our job and our pleasure.
PUTTING IT TO WORK
What this means if you make video
If you take one thing from this page, take this. The question is not whether to use AI. It is what to leave for it. Shoot everything you can and want to shoot: the performance, the frame, the camera move, and whatever else you have the means for. Then let the model complete what was out of reach. The more you bring to the shoot, the less it has to guess, and the more of the finished film is yours.
There is a bigger version of this, and it is the part we find most exciting. For as long as films have existed, what reached the screen was a reduced version of what someone had in their head, and most people accepted that as the price of making anything at all. That limit is coming off, and it is coming off first for the people who never had access. We are about to see into a lot of very different minds, and that is a bigger change than any tool on this page.
The camera still matters. The performance still matters. The director still matters. That turns out to be the best possible news for the people who love making things with a camera.
Keep what you love. Complete what you could not.
