Sidebar Title goes Here
In the world of statistics, a histogram is a kind of bar graph
that shows the distribution of data. In the Levels dialog box,
it’s a bit simpler. The central histogram in the Levels dialog
box contains exactly 256 vertical bars. Each bar represents
one brightness value, from black (on the far left) to white
(on the far right). The height of each bar indicates how many
pixels in your image correspond to that particular brightness
value. The result is an alternative view of your image, one that
focuses exclusively on the distribution of luminosity values.
Consider the annotated histogram below. I’ve taken the liberty
of dividing it into four quadrants. If you think of the histo-
gram as a series of steep sand dunes, a scant 5 percent of that
sand spills over into the far left quadrant; thus, only 5 percent
of the pixels in this image are dark. Meanwhile, fully 25 per-
cent of the sand resides in the big peak in the right quadrant,
so 25 percent of the pixels are light. The image represented
by this histogram contains more highlights than shadows.
One glance at the image itself (opposite page, top) confirms
that the histogram is accurate. The photo so obviously con-
tains more highlights than shadows that the histogram may
seem downright redundant. But the truth is, it provides an-
other helpful glimpse into the image. Namely, we see where
the darkest colors start, where the lightest colors drop off,
and how the rest of the image is weighted.
With that in mind, here are a few ways to work with the his-
togram in the Levels dialog box:- Black and white points: Keeping in mind the sand dune
analogy, move the black slider triangle below the histo-
gram to the point at which the dunes begin on the left.
Then move the white triangle to the point at which the
dunes end on the right. (See the graph below.) This makes
the darkest colors in the image black and the lightest col-
ors white, which maximizes contrast without harming
fragile details inside the shadows and highlights. - Clipping: Take care not to make too many colors black
or white. This will result in clipping, in which Photoshop
renders whole regions of your image flat black or white.
That’s fine for graphic art but bad for photography, where
you need continuous color transitions to convey depth
and realism.
How to Read and Respond to a Histogram
Luminosity valuesGamma pointBlack point White point256192128640NoneLotsNumber of colors200 Lesson 6: Adjusting Color and Luminance