Evaluating Images

You can form a quick judgment of the integrity of the data in your image by looking at the histogram. To help you form that judgment, look for the following hints when you read the histogram:

■ A histogram made up of smooth contours and data distributed evenly across the full 256 levels means that you have a good scan or capture from a digital camera (Figure 3-15).

■ A histogram that is weighted toward one side, to the left or right, almost certainly means that the full 256 levels are not being used (Figure 3-16).

Figure 3-15: A histogram showing a properly exposed image with good shadow and highlight detail and full use of the 256 levels. It also shows that at level 128 there are 82,447 pixels with the same intensity value.
Figure 3-16: A histogram showing an underexposed image; not all the levels are being utilized.

■ A histogram made up of peaks and troughs (Figure 3-17) is not necessarily a sign of a bad scan or capture; it could be just a reflection of the image content.

■ A histogram containing gaps (Figure 3-18) means that no pixels are present at those levels because either your scanner or your camera didn't capture them to begin with or they were eliminated in Photoshop when edits, transforms, or adjustments were applied.

Figure 3-17: A histogram showing a properly exposed image but unusual image detail.

You can also back that histogram up with statistical data; to see statistical data, hover the pointer over the histogram or click with the mouse anywhere in the histogram. Furthermore, you can view data from a range of values by clicking in the histogram and then dragging. Here's a quick breakdown of the statistical information:

■ Mean—The average brightness value of the total pixels in the image or in a selection.

■ Level—The numeric level underneath the pointer; levels are graphed horizontally from 0-255.

■ Std Dev (Standard deviation)—The variance between the highest and lowest intensity values.

■ Count—The total number of pixels corresponding to the numeric level underneath the pointer.

■ Median—The middle brightness value of all the pixels in the image, or in a selection.

■ Percentile—The total number of pixels at or below the numeric level underneath the pointer, that is, darker than the pixel underneath the pointer.

■ Pixels—The total number of pixels in the image, or selection, used to calculate the histogram.

■ Cache—The level of the cache (between 1 and 8) that the histogram is based upon. Level 1 represents the histogram based on actual pixels.

Figure 3-18: A histogram showing missing levels or clipped colors.
0 0

Post a comment