histogram question

dubsider · 11-28-2014, 11:03 AM

I pretty much understand (and rely upon) the camera's histogram, but I have a question..
If the Y Axis is to represent the number of pixels in that tonal range, what does it mean with the Y Axis is very flat but spread out across the X Axis and neither end is clipped at 0 or 255 ?

Adam · 11-28-2014, 11:43 AM

That would suggest that the number of lighter, darker, and neutrally-lit pixels is more or less equal.

dubsider · 11-28-2014, 03:33 PM

Adam - contrast that with a graph where the Y Axis rises very quickly from 0 through the highlight section filling the Y Axis to 255.. still no clipping but the graph seems to be filled (# of pixels)

**GUB** · 11-28-2014, 06:13 PM

Originally posted by dubsider

If the Y Axis is to represent the number of pixels in that tonal range,

I use the histograms a lot and sort of understand them.
But if the y axis represents the number of pixels then wouldn't the area under the line be always equal?
Glancing at my camera that would not seem to be the case although it is hard to tell. And certainly doing the curves in gimp the area changes.
May be the scale changes on the y axis to fit things in.?

emalvick · 12-02-2014, 02:49 PM

Originally posted by GUB

I use the histograms a lot and sort of understand them.
But if the y axis represents the number of pixels then wouldn't the area under the line be always equal?
Glancing at my camera that would not seem to be the case although it is hard to tell. And certainly doing the curves in gimp the area changes.
May be the scale changes on the y axis to fit things in.?

I think this is the key. Some software, and even cameras, automatically scale the y-axis to fit things in to some extent.

As an example, when you are clipping an image, the end points of the histogram will be very high leaving the remaining portion of the histogram looking flat. As a result it's all relative.

In reality, I think you really only have to know it's relative and understand what the shape of the histogram means roughly and make sense of it.

For instance, the earlier post about a histogram that was essentially flat. Did this make sense for what the image looked like? When you end up with spikes at the ends when things are clipped, does that make sense? When you see a more typical curved histogram, where is the peak?

To me, the y-axis numbers for a histogram and the general shape of the histogram mostly serves as a qualitative guide. On the camera it helps you to readjust and reshoot if the exposure is too high or too low. In PP it helps for guiding adjustment of the brightness, contrast, curves, etc. I almost never look at specific numbers or worry about them.

BrianR · 12-03-2014, 07:21 AM

They definitely scale the y-axis. In the Gimp or Photoshop, you can see the pixel count for each of the brightness levels. Take an image with a big white area that makes a spike to the right. Check the pixel count at a couple of mid-tone levels then crop out the white area. The pixel count of the midtones won't have changed, but the bars representing these values will be taller if you've cut out the area that caused the spike.

A flat histogram just means all brightness levels are equally represented (or nearly so). You can 'cheat' one of these into existence by taking a picture of a smooth gradient that ranges from white to black, or artificially make one in photoshop.

11-28-2014, 11:03 AM	#1
dubsider Junior Member Join Date: Mar 2009 Location: Georgia Posts: 27	histogram question I pretty much understand (and rely upon) the camera's histogram, but I have a question.. If the Y Axis is to represent the number of pixels in that tonal range, what does it mean with the Y Axis is very flat but spread out across the X Axis and neither end is clipped at 0 or 255 ?

11-28-2014, 11:43 AM	#2
Adam Administrator Site Webmaster Join Date: Sep 2006 Location: Arizona Photos: Gallery \| Albums Posts: 51,595	That would suggest that the number of lighter, darker, and neutrally-lit pixels is more or less equal. Adam PentaxForums.com Webmaster (Site Usage Guide \| Site Help \| My Photography) PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Histogram question	SPM	Pentax K-5 & K-5 II	2	03-09-2014 01:26 PM
RAW+ and histogram question	snofox	Pentax DSLR Discussion	3	11-13-2009 08:09 PM
K200D histogram question	RichardS	Pentax DSLR Discussion	1	09-08-2009 12:38 PM
K20D In-Camera Histogram Question	Lucky Sky	Pentax DSLR Discussion	4	11-03-2008 09:39 PM

11-28-2014, 03:33 PM	#3
dubsider Junior Member Join Date: Mar 2009 Location: Georgia Posts: 27 Original Poster	Adam - contrast that with a graph where the Y Axis rises very quickly from 0 through the highlight section filling the Y Axis to 255.. still no clipping but the graph seems to be filled (# of pixels)

11-28-2014, 06:13 PM	#4
GUB Loyal Site Supporter Join Date: Aug 2012 Location: Wanganui Photos: Gallery \| Albums Posts: 5,760	Originally posted by dubsider If the Y Axis is to represent the number of pixels in that tonal range, I use the histograms a lot and sort of understand them. But if the y axis represents the number of pixels then wouldn't the area under the line be always equal? Glancing at my camera that would not seem to be the case although it is hard to tell. And certainly doing the curves in gimp the area changes. May be the scale changes on the y axis to fit things in.?

12-02-2014, 02:49 PM	#5
emalvick Veteran Member Join Date: Feb 2008 Location: Davis, CA Photos: Gallery Posts: 1,642	Originally posted by GUB I use the histograms a lot and sort of understand them. But if the y axis represents the number of pixels then wouldn't the area under the line be always equal? Glancing at my camera that would not seem to be the case although it is hard to tell. And certainly doing the curves in gimp the area changes. May be the scale changes on the y axis to fit things in.? I think this is the key. Some software, and even cameras, automatically scale the y-axis to fit things in to some extent. As an example, when you are clipping an image, the end points of the histogram will be very high leaving the remaining portion of the histogram looking flat. As a result it's all relative. In reality, I think you really only have to know it's relative and understand what the shape of the histogram means roughly and make sense of it. For instance, the earlier post about a histogram that was essentially flat. Did this make sense for what the image looked like? When you end up with spikes at the ends when things are clipped, does that make sense? When you see a more typical curved histogram, where is the peak? To me, the y-axis numbers for a histogram and the general shape of the histogram mostly serves as a qualitative guide. On the camera it helps you to readjust and reshoot if the exposure is too high or too low. In PP it helps for guiding adjustment of the brightness, contrast, curves, etc. I almost never look at specific numbers or worry about them.

12-03-2014, 07:21 AM	#6
BrianR Veteran Member Join Date: Dec 2010 Location: Ontario Photos: Gallery Posts: 3,332	They definitely scale the y-axis. In the Gimp or Photoshop, you can see the pixel count for each of the brightness levels. Take an image with a big white area that makes a spike to the right. Check the pixel count at a couple of mid-tone levels then crop out the white area. The pixel count of the midtones won't have changed, but the bars representing these values will be taller if you've cut out the area that caused the spike. A flat histogram just means all brightness levels are equally represented (or nearly so). You can 'cheat' one of these into existence by taking a picture of a smooth gradient that ranges from white to black, or artificially make one in photoshop.