Monday, 23 November 2009

Color levels adjustments to your images.

Using software to make basic adjustments to your photos.




I originally wrote this article for a friend.   I hope it is useful for others too.

The difference between the image your camera captures and the image you wanted can be affected by many things. One of the most important things to remember is that your camera is a mechanical device (in an electronic way of course) and your are an emotional creature (in biological way, naturally).
A camera image is a collection of information about the scene that when viewed is arranged in a way we recognise as an image. The information that is captured is a digital representation of the subject.

Your camera will probably have a display mode that can tell you about the data it has captured for an image.


One of the simplest and most useful is a colour distribution histogram.




An example of a basic colour
distribution histogram.



Check your camera's manual to learn how to display the histogram.
(this can usually be found by pressing the button for playback, then the display mode button one or more times).
The histogram can help you adjust the color level of you images at shoot time and afterwards on a computer with an advanced graphics editing program.

(In this brief example I use donation-ware software called the GIMP - GNU image manipulation program).

On the histogram the leftmost horizontal value represents a value of 0 and increments to the right most side with a maximum value of 255. Vertically, the values start at the bottom line with a value of 0 and extends to the top with a value of 255.



Here is an example of a picture along with it's colour distribution histogram.









 

In this example we use the sliders beneath the histgram to enhance this image.







NB. How one views a picture (or which one perfers) is a subjective matter. These are to illustrate how dramatic changes can be made just by adjusting the histogram distribution.

Basic colour distribution histogram.


To illustrate how the stacking of the colours shown in the histogram works, I have taken two pictures of my power ball (below) one on a dark background and one on a light background.

Notice that in the first Picture the histogram shows that largest concentration of pixels are to the left of the histogram and in the picture with the light back ground most of the pixel stack up more to the right representing more dark pixels.

Pixels to the left









Pixels to the right.





NB. Notice where the stack of pixels are so tall that they go vertically off the scale. This tells us that details in that portion of the picture is greater than 255. 255 is the maximum value of a colour that can be stored, the value represents pure white. The fact these values  go so abruptly of the scale, shows that those pixels are over exposed and the missing detail is assuned to be 255.


An example showing adjustment using the level command.




In this picture of this Catherdal in Poland, the white of the building has caused the sky to become over exposed making the sky a much paler blue than I remember it.

For this example I have use the donation-ware software, GIMP (GNU image manipulation program) to make slight changes to bring back the deep blue sky colour and introduce more contrast into the images

I have opened this picture in the GIMP and selected Colour / thresholds to bring up the simple histogram that shows the spread of colour values across the picture.




There are three tiny triangles beneath the graph.


The left most arrow is the extent of details captured in the dark range, the middle is the midpoint a kind of relative “white” from which the other colours in the picture are calculated. And the right side is the extent of detail captured in the “light” side of the graph.


By moving the tiny (just the left and right triangles for now – the mid point will automatically adjust.) triangles to a point on the histogram where the histogram values start proper, we effectively compress the spread of values through the picture which gives deeper blues.

Notice that in the “before” picture above the range of values across which the colours are spread are 0 on the left and 255 on the right.


After moving the triangles until the left value is 39 and the right is 242 we have effectively compressed the colours, this has restored the blue sky and allow the clouds to be more defined.


Below you can see the effect of the adjustments made represented by the values on the histogram below A and B.


Before





 

After
(Deeper blue sky, more contrast)








Another example
Carnival Liberty in the Carribean Sea
As Delivered by Nikon D50



After







 Conculsion
  • Use your cameras histogram display regularly to check the colour distribution across the graph.
  • Also check for blown out details, values that are shown going abruptly off the scale. These represent lost detail that currently cannot be easily recovered.
  • As long as the details exist in the image you can use software like the gimp or Photo Shop to enhance your images.
  • Experiment with levels in your digital dark room to see the dramatic changes this tool can make to your images.

(Next flash photography tips).


No comments:

Post a Comment