+- Gimp-Forum.net (https://www.gimp-forum.net)
+-- Forum: GIMP (https://www.gimp-forum.net/Forum-GIMP)
+--- Forum: General questions (https://www.gimp-forum.net/Forum-General-questions)
+--- Thread: Help interpreting a histogram in Gimp (/Thread-Help-interpreting-a-histogram-in-Gimp)
Help interpreting a histogram in Gimp - linux_matt - 09-23-2017
There is a certain type of histogram which shows a very flat graph with few pixels for any tone (light, mid or dark). This type of histogram (shown below) relates to photos with pale/washed-out colours. I have often wondered why this histogram appears to indicate that the photo actually has less pixels than another photo with the same resolution. Shouldn't the histogram show the same number of pixels for any image? I think I am interpreting the information incorrectly and would be grateful for advice on this
![[Image: panorama_geometry1.jpeg]](http://www.temaroma.com/linux_stuff/panorama_geometry1.jpeg)
RE: Help interpreting a histogram in Gimp - mholder - 09-23-2017
It would help if I could see the actual image. Usually when I see a histogram like this, there either is a lot of pure white or pure black, and a very skinny line on the histogram at the edges, representing the white or black pixels. Is that a line at the far right, representing the white pixels? You did say the image was washed out, maybe with too much white?
If you don't already know, you can go [colors->info->colorcube analysis] and see how many colors are in the image, if that helps.
There are two images here. Before I adjust the contrast and after. After I raised the contrast the white pixels get blown out and the histogram changes like yours
RE: Help interpreting a histogram in Gimp - Ofnuts - 09-23-2017
(09-23-2017, 11:10 PM)mholder Wrote: It would help if I could see the actual image. Usually when I see a histogram like this, there either is a lot of pure white or pure black, and a very skinny line on the histogram at the edges, representing the white or black pixels. Is that a line at the far right, representing the white pixels? You did say the image was washed out, maybe with too much white?
Indeed. In addition, you can use the logarithmic histogram, often more readable (pair of icons at the top right).