This AI smartphone can pull off Hitchcock's hardest camera trick in Vertigo without a film crew
The great film director Alfred Hitchcock spent Vertigo solving a problem most would rather avoid: how do you make a room feel like it's stretching away from someone without actually moving them?
The answer, dolly the camera one way while zooming the lens the other, has become known as the Hitchcock zoom. It's one of the hardest shots in film-making to pull off. It needs a rig, a crew and several takes to get the timing right.
So it's a sign of how far tech has come that the Honor 600 Pro smartphone offers it as a preset. Rather than filming anything, you just start from a single still image and let the model generate the motion around it. It's incredible, really, and reason enough to add it to our list of the best camera phones.
But it's not all. Other camera movement options inspired by professional filmmaking, including bullet-time shots and drone-style follow shots, are also now possible just by uploading a still.
Unlock your content
As Peng Jianteng, AI engineer at Honor and technical lead on its AI Image-to-Video work, explains, the appeal of image-to-video isn't novelty so much as unlocking content that's already sitting idle. "Many images, such as illustrations, old photos or designs, were originally created as still content," he points out. "AI Image-to-Video 2.0 can animate them with character movement, camera motion or environmental changes, giving the original image a new and totally unprecedented level of expression."
That doesn't just save time, it opens up new creative possibilities. "Some scenes are difficult for ordinary users to film in real life," says Peng, who joined Honor in 2023. "For example, bullet-time camera movement or playful, human-like actions from pets. Starting from a single image, users can easily explore different styles, motions and visual ideas."
Before launch, Honor handed the tool to students from art universities to see what they'd do with it. The results, Peng says, ranged well beyond what the marketing brief might have predicted. "These films included short music videos, with both the music and the music video created by AI, or expressive hyper-surreal movie-style concepts. Some of the most interesting uses of AI happen when creative talent is given the right tools and a platform to experiment."
Sign up to Creative Bloq's daily newsletter, which brings you the latest news and inspiration from the worlds of art, design and technology.
Meanwhile, everyday users have found their own, less experimental habits: turning generated clips into Live Photos for social media, or stitching several scene templates together so they appear to walk through a string of different environments as the same character.
Peng Jianteng is an AI engineer at Honor and technical lead on the company's AI video generation work. He joined Honor in 2023 after a career in computer vision, and now leads development of the AI Image-to-Video technology featured on the Honor 600 series.
Just getting started
This doesn't mean, though, that the technology has matured; quite the reverse. "Today, features on the HONOR 600 can already animate a single image and generate a short video," Peng says. "However, there are still limitations. Characters or objects may occasionally change during generation, motion can sometimes appear unstable, and the clips are still relatively short."
The next stage of AI Image‑to‑Video tech, he predicts, will focus on four main areas: controllability, consistency, longer video generation, and personalisation.
"The first direction is stronger control. Users will be able to specify camera movement, character actions and the timing of events more precisely, rather than relying mainly on the model’s automatic generation. The second is better consistency. Characters, clothing and environments should remain stable across longer clips or multiple shots.
"The third is longer video generation. Instead of short clips of a few seconds, the technology will gradually support longer videos with more continuous storytelling. Finally, personalisation will play a bigger role. By learning a user’s preferences, the system can generate videos that better match their creative style and aesthetic."

Tom May is an award-winning journalist specialising in art, design, photography and technology. He is the author of the books The 50 Greatest Designers (Arcturus) and Great TED Talks: Creativity (Pavilion). Tom was previously editor of Professional Photography magazine, associate editor at Creative Bloq, and deputy editor at net magazine.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.