When I first started doing video, I thought it would be an easy transition from still photography. They both capture light through a lens, iris, and shutter. One moves; the other doesn’t. That’s no big difference, right? Wrong. I’ve been surprised by how very different these two media are. I’ll be detailing these differences in a series of upcoming posts. For now, let’s start with some technical differences.
No cropping
Cropping is incredibly common in still photography. Most photographers crop their final images to a portion of the original image to strengthen the composition, or to eliminate unwanted objects such as a band of mad clowns leaping into an otherwise tranquil frame. Still photographers get away with this because they capture at such a high resolution that even if they print only a portion of the original, the final print will still be at a high enough resolution to look good. A 6 megapixel camera gives you 300 dots per inch in the final 8x10 print. 300dpi easily surpasses the eye’s ability to see individual dots.
Video cameras, on the other hand, capture images at the exact same resolution as the output medium. This resolution is pretty small: 720x480 for DV - that’s a whopping 0.3 megapixels. If you crop and try to blow the image up to fill the screen, it looks terrible. This means you don’t get to crop video. You’ve got to get it right in the camera the first time.
No cheating
Still photographers also can rescue images (aka “strengthen composition”) in photoshop - erasing blemishes, fixing blown highlights, moving trees around, etc. They get away with this because they only have to fix one image. Video moves. Video has 30 frames per second with the same object appearing in different positions in each frame. If you’ve got a 30 second scene, you’ve got nine hundred images to go clean up.
That’s not to say that photoshop-like fixes are impossible - they’re just much harder. You’ve got to animate your fix - moving it around the screen, changing its intensity as the object you’re fixing moves through the light. This is tedious and difficult. If you look closely at the last scene of the Definition Fitness video[url], you’ll see that when the man looks straight at the camera, the right side of his face is completely blown out. I “fixed” it by animating a skin-colored patch, moving and morphing it to fit the shape of his face as he talks. It took me about an hour to get right. It’d take me about 30 seconds if I had to do it for one image in photoshop. Fixing images in video is expensive. You’ve got to get it right in the camera the first time. (sense the trend here?)
