So, Nikon's latest consumer camera, the Nikon D90, supports recording high-def video at 720p and 24fps, using the same excellent-quality camera and sensor used for general prosumer photography. This is pretty big news—nobody else has a product that can record video like this, using a wide line of available interchangeable lenses.

I was thinking while watching their high-quality demo video that the built-in microphone kind of sucks. I, however, have a great microphone at home. I realized that, if this camera doesn't support external microphones (though I assume it would), one could still record audio separately from the video, and synchronize it later. I'm not aware of the features of modern video software, but it actually seems like it would be reasonably straightforward to write some code that could take one segment of video synchronized with low-quality audio, and a separate high-quality audio track, and replace the low-quality audio with the high-quality audio automatically, without the user having to align audio tracks or specify any settings at all. All it would need to do would be to analyze the time positions of the relative peaks and valleys of the two sound waves, and match them up, finding the position that minimizes the mathematical difference between the two waves. Finding the best match would probably be processor-intensive, but not prohibitive.

Maybe some pro video editing software somewhere already does this, and I just haven't seen it. Feel free to go make some money with that "high-quality" idea I spent an exhaustive five minutes on...

Labels: photography, technology