World's first full HDR video system sees like the human eye
By Ben Coxworth
January 18, 2011
Anyone who regularly uses a video camera will know that the devices do not see the world the way we do. The human visual system can perceive a scene that contains both bright highlights and dark shadows, yet is able to process that information in such a way that it can simultaneously expose for both lighting extremes – up to a point, at least. Video cameras, however, have just one f-stop to work with at any one time, and so must make compromises. Now, however, researchers from the UK’s University of Warwick claim to have the solution to such problems, in the form of the world’s first full High Dynamic Range (HDR) video system.
HDR has been in development for some time – Sunnybrook Technologies unveiled a High Dynamic Range display system back in 2004, and just last year BenQ joined a list of several manufacturers to have released HDR still cameras. Even HDR video has been shot before, albeit on a limited, experimental basis. What the researchers at Warwick claim to have developed is the world’s first full-motion HDR video system, that covers everything from image capture through to display.
“HDR imagery offers a more representative description of real world lighting by storing data with a higher bit-depth per pixel than more conventional images,” explained Prof. Alan Chalmers, of Warwick’s WMG Digital Laboratory. “Although HDR imagery for static images has been around for 15 years, it has not been possible to capture HDR video until now. However, such HDR images are typically painstakingly created in computer graphics or generated from a number of static images, often merging only 4 exposures at different stops to build an HDR image.”
The new system, by contrast, captures 20 f-stops per frame of 1080p high-def video, at the NTSC standard 30 frames-per-second. In post-production, the optimum exposures can then be selected and/or combined for each shot, via a “tone-mapping” procedure. A process called Image-Based Lighting can also be utilized, in which computer-created objects can be added to real-world footage, where they will appear to actually be lit by the light given off by that footage – in one example, the light of real-world explosions is reflected on the sides of a computer-generated car.
Of course, all of that extra data takes up some space – each frame is 24MB in size, which works out to 42GB per minute. To address that rather large quandary, the researchers are collaborating with HDR tech firm goHDR to develop software that will compress the HDR footage by at least 100-fold. This should allow for existing editing systems to be able to handle the video files.
The final step in the process is the HDR monitor. It consists of an LED panel which projects through an LCD panel placed in front of it. The combination of the two screens is necessary to provide all of the lighting information.
The Warwick team believe that the technology would be useful for applications such as televised sports coverage (in which a football moves in and out of sunlight and shadows, for instance), conducting or recording surgery, or for security systems. It could also find use in feature film-making, as the researchers state that it could be used to create 3D images that don’t require viewers to wear special glasses.