BetterPhoto Q&A
Category: New Answers

Photography Question 

NAVNITH KRISHNAN
 

The Histogram: What It Is, How It's Used


I am a bit confused by the histogram. Can somebody explain to me how to read histogram and how one can use it to improve photography? I have a Nikon D70s.


To love this question, log in above
May 26, 2006

 

doug Nelson
  A histogram divides the available tones of light and dark in an image into 256 shades. It shows the distribution of shades in the overall image. It is important to know that a histogram shows you nothing about where in the image the tones happen to be, just how many of each shade is there, relative to the other shades.

So, for a practical example, all of the histogram bunched up on the left tells you the image is a preponderance of dark tones. Bunched up on the right is a high-key image, with mostly bright tones. An even distribution tells you that there are about as many brights as darks.

A histogram will also show you where there is no tonal information, usually at the extreme ends of the scale. Imaging operators usually will move the slider on each end over to just the point at which tonal information is being shown. Most images, whether out of a digital camera, or from a scanner, look a lot better when you do this.

Editor's Note: For anyone interested, BetterPhoto offers this June online course:
What the Histogram Tells You About Exposure


To love this comment, log in above
May 26, 2006

 

doug Nelson
  One more tip, the histogram with the adjustable sliders will be found in Levels, under Image/Adjust, depending on which version of Photoshop or Elements you have. See www.cambridgeincolour.com for some nice tutorials on this subject.


To love this comment, log in above
May 26, 2006

 

Slim Brady
  A light meter reading tells you what exposure will render a standard 18% gray reference card as a mid tone. This reading may have been made because the camera read a variety of areas of the scene and averaged them out, or because you read the highlights, the shadows and some other areas and decided that a particular setting would yield the best compromise exposure for that scene.

This setting, like every other that you or your automated camera makes, is a compromise. In most real world situations there is no such thing as an ideal or “perfect” exposure. There is simply one that places the tonal values found in the scene most appropriately within the capability range of the camera’s imaging chip. And "most appropriately" means that the mid-tones found in the image fall roughly half way between the darkest and the brightest values.


To love this comment, log in above
May 26, 2006

 

Sabrina Dionne
 
 
 
A histogram is simply a graph of a scene's brightness values as recorded by your camera, including their distribution and relative amounts. Darkest tones on the left, brightest on the right; height of the graphs peaks has no numeric rating.

Peaks and valleys bunched against the left the image is probably underexposed. And bunched to the right, overexposed.

Then again:
Depends on the content: a predominantly dark picture with a bright highlight, vs predominantly bright picture with a few deep shadows, both "correctly" exposed, will have histograms pushed to opposite ends.

Most shots you take, the peaks and valleys should be in the middle with a little more to the right. keep in mind however there is no "perfect" histogram.(See picture)Here the highlights are OK. Losing a bit of the dynamic range in the highlight area but a good photo editing program can correct for this.

Good reading:http://www.sphoto.com/techinfo/histograms/histograms.htm , American Photo Magazine, March/April issue. Site to visit: http://www.luminous-landscape.com/tutorials/understanding-series/understanding-histograms.shtml or take an online course through betterphoto.com http://www.betterphoto.com/photocourses/ELL02.php .

Hope this helps,
Sabrina Dionne


To love this comment, log in above
May 30, 2006

 

Sabrina Dionne
 
 
  Ok Highlights
Ok Highlights
For the above comment by Sabrina Dionne.

Sabrina Dionne

 
 
Image:


To love this comment, log in above
May 30, 2006

 
This old forum is now archived. Use improved Forum here

Report this Thread