Wednesday, May 7, 2008

Future Thinking: Recording Data for Playback Media Yet to Be Invented

Recently researchers discovered what was billed as the earliest sound recording ever – a brief clip of a French folk song recorded by Édouard-Léon Scott de Martinville in Paris on April 9, 1860. Actually de Martinville never intended to record sound to be played back. He was simply interested in capturing on paper a visual representation of sound for analysis, but with modern technology this paper can now be scanned and converted into sound by means that he could never have anticipated. (See the New York Times for details, http://www.nytimes.com/2008/03/27/arts/27soun.html?_r=2&oref=slogin&oref=slogin)

This reminded me that very often there is a means to capture inputs suitable for some future playback mechanism long before this mechanism exists. For example, decades before color photography became a mainstream product some photographers were capturing color separated black and white negatives for display using colored light projectors. Later these negatives could be used to make prints using color printing paper. In the same way pioneers of moving film could easily have made sets of negatives that would have allowed people to make color movie prints later when the technology was available, though I know of no case where this was actually done.

Sometimes, like in the de Martinville case, the inputs are recorded accidentally, as in the case of early non-electrical sound recordings where multiple recorders were sometimes used to record many original masters. As a result it is occasionally possible to find two recordings from two separate microphones, giving a true stereo effect. There are some authentic stereo recordings of early bluesmen assembled in this way.

Animated cartoons, too, could have been created long before movie technology was invented. Leonardo da Vinci could conceivably have made hundred of drawings that when assembled in a flip book would give animated drawings; it was also not beyond the technology of the day to make a “What the Butler Saw” machine. With technology that developed 500 years later these drawings could have been photographed and projected as cartoons.

More recently photographers have started experimenting with High dynamic range photography ( http://en.wikipedia.org/wiki/HDRI ), where digital photographic techniques are used to capture visual information that cannot be viewed directly using today’s prints and displays but can be stored in anticipation of the development of this technology. At its simplest this consists of taking the same picture three times with three different exposure values – a technique used by traditional photographers to ensure one perfect exposure. One of the images will be on average underexposed but will show more detail in the highlights; another will be overexposed in the highlights but will show more details in the shadows. The data from all three images can be combined in a file which holds more information than we can use today, though there are techniques for mapping these images to a practical format which has a strange, shadow-less look.

Similarly, anyone can capture inputs for future 3d video projection technology: all you need to do is record the video using two cameras side by side and then just wait; eventually 3d projection video will be in every home and your baby videos will not look so dated. Suitable display and projection technologies may not be in stores next year, but most of you will still be around when it comes.

So what are the practical lessons from all of this? For important events and actualities maybe you need to be forward thinking and considering capturing some extra data that might be appreciated by future generations. In my own small way I have already started by recording all video in 16:9 format – old style 5:4 TV format looks sooo dated -- and by capturing some photos for future high dynamic range technologies. But if you are in a position to be capturing images that are likely to have value in the future I would seriously consider thinking ahead to future media and anticipating what kind of data will be needed. For example, at the inauguration of a head of state or a public building expected to last a long time I’d give some thought to recording some video in both stereo and high dynamic range. At home you might do the same for baby pictures. One day someone will thank you for this.

No comments: