Discussion in 'Beyond the Basics' started by flightless_beaker, Nov 12, 2009.
How exactly do you read a histogram and what benefits/improvements will I have if I can read it?
The first thing to realize is there are no 'bad' or 'wrong' histograms... What they will tell you is how much of your image is highlights, midtones and shadows and also how much (if any) is blown out (highs and/or mids), meaning 0 detail.
There are lots of good write-ups on this.
Digital ProTalk: histograms
From there, you'll find other useful links that will further guide you on learning about histograms, their use and implementation.
Depends on what part of the histogram you are looking at. Most cameras come with a luminosity histogram, as well as a histogram for the individual Red, Green and Blue channels. The luminosity (just think of it as the "light" channel) histogram will tell you what part of your image is overexposed and underexposed. The individual histograms for the color channels will tell you which of the different colors themselves are "overexposed": these will lead to color shifts in your image. So its possible you have a luminosity histogram that only has a few pixels to the far right edge of the scale (0 to 255 - these would read as all 255's - pure white), and will have a SH*T TON of "overexposed" Red pixels. Color shift.
Being able to read a histogram is quite important, as you'll be able to tell w/o benefit of calibrating your monitor, what the colors and contrast in an image is doing. You'll be able to tell whether you've got a good histogram (with each of the color channels matching up NEARLY perfectly - this is not needed, but the more they don't match up the more of a color shift you're going to have) or whether you have one that is going to need a little bit of work in Photoshop.
The luminous landscape has one of the better articles which is rife with examples:
Have a quick read over that. It tells you what the histogram does, why to use it, and why it ultimately is meaningless for the final image beyond detecting any abnormalities.
Wow! That Luminous Landscape site is great! Thanks for the tip Garbz.
After reading 'Understanding Histograms', read this; Expose Right
Separate names with a comma.