For a POV to be effective, it must match the perspective of the character in the scene. In other words, it should match their eye line height, position on the set, and the angle of direction they were looking in the previous shot.

In this gradient, the values of pixels move evenly from dark to light, but the human eye sees most of the change occurring in the dark regions. That’s because the human eye is most sensitive to dark and mid-range contrast and much less sensitive to contrast in bright regions of an image.

In addition, sense of contrast is contextual. These two squares actually have the same pixel value, but because of the intensity of the surrounding pixels they are perceived as having very different depths.

It’s important to keep this relative sense of contrast in mind when working with film and video content.

Computer engineers have exploited this understanding of human perception by encoding image files logarithmically, throwing away data that the human observer won’t be able to detect, thus reducing file size.