Figure 14.16. The Histogram dialog
The Histogram dialog shows you information about the statistical
distribution of color values in the active layer or selection. This
information is often useful when you are trying to
an image. However, the Histogram dialog is purely informational: nothing
you do with it will cause any change to the image. If you want to perform
a histogram-based color correction, use the
2.5.1. Activating the dialog
The “Histogram” dialog is a dockable dialog; see the section
Section 2.3, “Dialogs and Docking”
for help on manipulating it.
You can access it:
In the detached windows which exists
only if at least one dialog remains open. In this case, you can raise
the “Histogram” dialog from the image-menu:
menu, there is a list of
In GIMP, each layer of an image can be decomposed into one or more color
channels: for an RGB image, into R, G, and B channels; for a grayscale
image, into a single Value channel. Layers that support transparency
have an additional channel, the alpha channel. Each channel supports a
range of intensity levels from 0 to 255 (integer valued). Thus, a black
pixel is encoded by 0 on all color channels; a white pixel by 255 on all
color channels. A transparent pixel is encoded by 0 on the alpha
channel; an opaque pixel by 255.
For RGB images, it is convenient to define a Value "pseudochannel". This
is not a real color channel: it does not reflect any information stored
directly in the image. Instead, the Value at a pixel is given by the
V = max(R,G,B).
Essentially, the Value is what you would get at that pixel if you
converted the image to Grayscale mode.
For more information on channels, please consult the
Section 1, “
2.5.3. Using the Histogram dialog
The active layer name is shown at the top of the dialog.
Channel options for an RGB layer with alpha channel
This allows you to select which channel to use. The possibilities
depend on the layer type of the active layer. Here are the entries
you might see, and what they mean:
For RGB and Grayscale images, this shows the distribution of
brightness values across the layer. For a grayscale image,
these are read directly from the image data. For an RGB
image, they are taken from the Value pseudochannel.
For an indexed image, the “Value” channel
actually shows the distribution of frequencies for each
colormap index: thus, it is a “pseudocolor”
histogram rather than a true color histogram.
Red, Green, Blue
These only appear for layers from RGB images. They show the
distribution of intensity levels for the Red, Green, or Blue
This shows the distribution of opacity levels. If the layer
is completely opaque or completely transparent, the
histogram will consist of a single bar on the left or right
Combined histograms of R, G, and B channels.
This entry, only available for RGB layers, shows the R, G,
and B histograms superimposed, so that you can see all of
the color distribution information in a single view.
The histogram shown at the top, changed to logarithmic mode.
These buttons determine whether the histogram will be displayed
using a linear or logarithmic Y axis. For images taken from
photographs, the linear mode is most commonly useful. For images
that contain substantial areas of constant color, though, a linear
histogram will often be dominated by a single bar, and a
logarithmic histogram will often be more useful.
Dialog aspect after range fixing.
You can restrict the analysis, for the statistics shown at the
bottom of the dialog, to a limited range of values if you wish.
You can set the range in one of three ways:
Click and drag the pointer across the histogram display area,
from the lowest level to the highest level of the range you
Click and drag the black or white triangles on the slider
below the histogram.
Use the spinbutton entries below the slider (left entry:
bottom of range; right entry: top of range).
At the bottom of the dialog some basic statistics are shown
describing the distribution of channel values, restricted to the
Mean : the mean value of the interval
in the selected channel.
Std Dev : Standard deviation. Gives an
idea about how homogeneous the distribution of values in the
Median : For example, the value of the
fiftieth peak in a 100 peaks interval.
Pixels : The number of pixels in the
active layer or selection.
Count : The number of pixels in a peak
(when you click on the histogram) or in the interval.
Percentile : The ratio between the
number of pixels in the interval and the total number of
pixels in the active layer or selection.