If the black (left most) triangle for the output is now draged to the right, the effect is that all lower input pixel valuess are mapped to a higher output pixel value with all values above that mapped proportionately. This make all darker regions brighter giving a small contrast range. The transform now looks like:
Similarly the white (right most) triangle for the output can be dragged to the left decreasing the maximum output pixel value. This tends to darken the image. The transform now looks like:
