What is HyperZoom™?
Imagine being able to seemingly fly from one location to the next, even if they are many kilometres apart, over and through trees, across lakes, over mountains, through buildings, into computer screens or posters in the street – all with no loss of visual continuity.
Effectively a continuous single take video that takes the viewer on an impossible, incredible journey.
This was the dream behind the development of HyperZoom™
In order to be able to achieve this, HyperZoom™ abandons conventional methods of capturing footage and also abandons conventional methods of post producing and editing that footage.
Paradoxically, even though extreme camera movement is a hallmark of the technique, HyperZoom™ does not use any drones, cranes, sliders, motion control rigs etc.
The camera doesn’t actually move at all during the capture.
All shots are produced with the camera on a static tripod or from a static hand held position.This lack of movement during the capture of the footage is essential to the three dimensional assembly of the final video and enables the virtual camera movement produced during post production to be completely independent of the action in the scene itself.
In other words we can change the flight path, speed and direction in post production without affecting the speed of motion in the footage itself (which can be timelapse, realtime or slow motion).
So how does it work?
HyperZoom™ takes advantage of the way we interpret the world we see in front of us, especially the way in which we use our past experience to interpret two dimensional images as three dimensional scenes. When we look at a photograph of a car for instance, we presume the car is the size cars usually are and this enables us to decide how far away the car is in the scene around it. It could of course be a much smaller car but much closer to us.
The only way we can tell the difference between the toy car and the real car, if viewing the scene from a single viewpoint, is to move the viewpoint itself and see what happens to the parallax shift between the car and its surroundings. This gives us depth information and enables us to correctly position and size the car in the scene.
This parallax shift depth perception is only really useful up to about 20metres away (unless we move the viewpoint a very large distance).
HyperZoom™ utilizes this understanding and deconstructs any given scene into depth layers – effectively foreground, middleground and background. The foreground needs to be more finely layered as the parallax shift here is greater and more critical than between the middleground and background.
If we shoot the scene as a series of layers related to their depth positions within the scene we can later, in postproduction, reassemble these layers in 3d space and create the ability to move a virtual camera to navigate the layers.
If the layers are positioned correctly, relative to the virtual camera movement, it is possible to create an effect that is very similar to actually flying through the scene. Furthermore by careful construction of the three dimensional layer puzzle we can create a completely new world where scenes that were originally many miles apart are brought together in a natural way with no cuts as the camera smoothly travels between one scene and the next.
German Television station Deutsche Welle talks to Geoff about the HyperZoom™ technique
Below is a short HyperZoom™ video produced for a construction company building a Hotel in Upper Austria