Point, Shoot, Focus

October 24, 2011

Photography has always had a sort of class divide between the very large number of casual photographers that use simple, “point and shoot” cameras (starting with the original Kodak box camera in 1888, and exemplified today by simple cellphone cameras), and the smaller number of serious amateurs and professionals who use more complex cameras with adjustable focus, lens aperture, and shutter speed.   Photographers in this latter group could get images of much higher quality, but had to learn a good deal about how the whole system worked in order to do so, for image quality was dependent on getting those pre-exposure settings correct.

Now, according to an article at Technology Review, a new type of camera will allow some of that adjustment work to be carried out after the picture is taken.  A start-up company, Lytro, has developed a point-and-shoot camera with a new type of image sensor, which, because it captures more information than a conventional sensor (or film), will allow pictures to be focused after they are taken.

The camera has a novel design reminiscent of a telescope. It features only two buttons: one to turn the device on or off, and one to take a photo. Only after a photo is taken does the user need to worry about focusing the resulting image.

The “light field sensor”, as Lytro calls it, consists of a conventional image sensor with a  glass layer above it; the glass is etched to produce many small lenses.  (The idea is somewhat similar to that of a Fresnel lens.)

Lytro’s sensor is made by bonding a carefully etched sheet of glass on top of a conventional digital-camera sensor. The glass is patterned with tiny lenses, ensuring that specific pixels can receive light only from the specified angles.

This allows the Lytro camera to record not only the amount of light present bur also the angle from which the light comes; Lytro calls this the light field.  With this information, software can then compute what a conventionally focused camera would “see”, and regenerate the image.   The computations are similar to those used in generating synthetic three-dimensional images by ray tracing.  Since the light field incorporates depth information, production of 3-D images is also possible.  (The company’s Web site has a gallery of sample images.)  Ars Technica also has an article on the Lytro camera.

All of this comes at a price, of course.  Recording the extra information about the light field means that, for a given sensor, the light field image will have considerably lower resolution than a conventional image.  For some usage, such as photo sharing on the Web, this will not matter much;  current cameras have resolution far in excess of what is required.   Printed images (which of course have to be at a particular focus setting) may have less resolution than those from quality conventional cameras, but are probably good enough for snapshots.  And there are image sensors with considerably higher resolution under development.

The cameras are slated to be available sometime next year.  It’s a clever new approach to capturing images; it will be fascinating to see how it develops.

%d bloggers like this: