Audiovisual production
Magic Film Cloud
Images and video for production, with per-shot approval.
Let’s talkCreate, explore versions and approve each shot
Magic Film Cloud rethinks audiovisual creation from the outset with AI at its foundation, helping media, marketing and communications teams express their ideas through images and film. AI supports creation and versions; people describe each shot, direct, review and approve. Delivery includes a report showing how each take was made. Its initial version keeps evolving.
Who it is for
Media companies that produce content, such as broadcasters, studios and production companies. It is used to generate production material, try versions and approve it.
What it solves
Generated material is meant for pieces that are broadcast and distributed, so it matters to know what each take was made with. Delivery comes with a provenance report stating the model, the license and the cost per take.
Your description comes first
What you write stays at the heart of the shot. Your description is never silently replaced. That’s a guiding rule of the product.
How it works
From the project to delivery, step by step.
Projects and shots
Work is organized into projects and shots. Each shot has its own format, and that format is kept when generating and when viewing the result.
The shot description
The person writes the description of the shot and can edit it in place. It saves on its own, and generation always uses the latest text.
Images and video
Images and video are generated from that description. There is also editing by written instruction.
Approval and delivery
Approval happens per shot. On delivery, the material goes out together with its provenance report.
An initial version that we keep improving day by day.
Each shot has its own format, its versions and its approval.
Each company has its library and can use its own brand.
What it does, piece by piece
Explore the pieces of the platform. Open each one to learn more.
Each shot has a gallery of versions. You can view them, choose which one is the main version and remove the ones that do not work.
Each company has its own library of resources: images, videos and logos that can be reused in any shot.
A reference image can arrive in three ways: by pasting its address, by picking it from the company library, or by reusing another image on the platform, as long as it belongs to the same company.
The platform can be shown with each company's name, logo and color.
The interface rule is that wherever the address of an image or a video is pasted, a preview appears below it. A mistake shows when you look at the image, not when you read the text.
It is designed so that each generation job is recorded and can be reviewed later.
Notice before spending
If an option changes the price or the conditions, the platform must say so before anything is spent, not after.
Only what has been tried
The rule is that a new AI model is not offered until someone on the team has actually run it and seen the result.
Similar screens, same order
Screens that do the same job with different material, such as image and video, must follow the same order. What you learn on one should work on the other.