Of backgrounds, technology and convenience...

Last time, I mentioned in passing about optical front projection as a means to extending creative options in my studio. It's a system I've been using for over twenty years, so I'm pretty familiar with it, both the advantages and the disadvantages. And the technical limitations and hassles associated with it, too.

On the face of it, it's a huge plus to be able to pull up photo-realistic scenery into any image at the flick of a switch. In practice, we have to manage a huge collection of film transparencies, storing them in good condition as well as in an easily searchable manner. It can be a problem if it takes several minutes to locate a specific background while a client is waiting... We also have to maintain the screen itself. Frequent washing with warm, soapy water helps remove the inevitable fingerprints, but small creases are harder to deal with. Over time, a screen becomes damaged to the extent it becomes unusable, and these things are not cheap to replace.

The other most significant drawback is the need to employ unusually controlled lighting techniques. For most types of studio work, diffuse light is most commonly needed. But the direction of light has to be carefully controlled when using optical front projection. The screen typically has a cutoff point around 25 degrees or so from normal, where 'normal' is the angle of view of the camera. This means that any light hitting the screen at an angle of less than about 25 degrees to the camera's own view tends to wash out the background image. To provide this combination of diffuse yet directional light, I developed my own light boxes many years ago. These are about two feet deep and reflect the light from a flat panel in the back of the box, passing it through a grid at the front in order to cut out any wayward bits. This works well, but the boxes are heavy and quite cumbersome. Sadly, I've found no practical alternatives which work as well.

The biggest stumbling block for people using photo-realistic background systems of any sort always seems to be making it convincing. It really shouldn't be hard for anyone calling themselves a photographer to appreciate that matching the lighting direction and quality, together with the perspective is absolutely necessary. It's really not that hard to do, but I have seen some appalling attempts.

In recent years, I've obviously had one eye on the possibilities of colourkey as an alternative to my optical projection system. It's cheaper to use, in terms of the screen. It's less demanding in terms of lighting control. But it's also far less subtle and obviously relies heavily on software. It does though fit in with one development I have considered for a while, but have yet to find a practical way to implement.

The idea is to replace the part of the optical system which uses a flash head, a modelling lamp and a film transparency with a digital projector. The problems are primarily: (1) getting enough light from the projector to give an acceptable exposure time, especially when the output is filtered to match the colour temperature of my main lights. (2) getting enough resolution to provide a convincing background image. (3) keeping the whole system light enough, small enough and running cool enough to be practical. The big plusses would be the ability to file backgrounds on the nearest PC and to adjust them more easily for scale and perspective. The hybrid of digital projection with an optical background system might actually be very useful, having the most important advantages of both, but I'm beginning to wonder whether it will ever be economically achievable.