Photoshop - Processing Basics
Photoshop is the most widely used imaging processing
software as far as I know. The name alone is synonymous with any type of
processing and we have all heard the term "photoshopped" when referring to
images that have been obviously altered in some way. Your camera probably came
with some kind of photo processing software and it may even be some version of
Photoshop. There is no way for me to comment on any other piece of software.
Photoshop CS3 is what I'm currently using and the essays that follow will mostly
be using that as a reference. However the techniques described should be able to
be easily understood for most versions of Photoshop and hopefully to users that
are using other brands as well. The menus will be different of course but
hopefully your software has some of the features described and you can still use
the ideas here to process images.
Photoshop in general is a very powerful program that
has more features than most people can digest. For this reason it can be very
intimidating and difficult for the casual user to wrap their head around. It was
for me when I first started using it, but after playing with it for a while and
researching tutorials like this one, I was able to take charge and control it
rather than it controlling me. It is the same when using a camera. The better
you control it, the less it controls you.
What exactly is meant by 'processing' photos? In my
opinion, almost every photo needs to be processed. The image that was captured
by your camera is not the image I would want to show anyone. Virtually every
photo requires some degree of enhancement to bring out the best of what the
camera actually captured. The image has more information hidden in it than is
being displayed in the raw photo out of the camera. Processing cannot make a bad
photo good. A bad photo is a bad photo whether it is out of focus, blurry from
shake, too light, too dark or whatever and most of them are unsalvageable or not
worth the effort. I say most of the time because there can be rare occasions
where something interesting can be pulled out of an otherwise bad image.
What processing can do is make an ok image good or good
image great or great image spectacular. Most of the time, just a few simple
adjustments is all that is needed to bring out the best of the image.
The first thing I do is check the Levels. The Levels
adjustment can be opened by going to the top of the screen and opening the Image
menu. Under the Image menu you will see Adjustments which has a sub menu and the
first thing in the sub menu is Levels (hot key shortcut Ctrl+L). Opening the
Levels adjustment shows a 'histogram' which is a graphic representation of all
the pixel colors that are present in the image. The extreme left is total black,
the right is white and the middle is neutral gray.
 |
Looking at the the top we see the channel and
beside it says "RGB" which means red, green, blue which, along with
their opposites cyan, magenta and yellow, represent all the colors there
are in an image. We can change levels for the entire RGB spectrum by
moving the the little arrows under the graph. We can delve deeper by
clicking the down arrow to the right of RGB and a drop down menu will
allow us to play with adjustments for red, green or blue separately.
Notice how the graph comes up short on the right
side. This is the light side and it is telling us that there are no
white pixels anywhere in the image. In the actual image below, you can
see it is rather dark and lifeless. Even though the cat is white, in the
image, it is actually light gray. We can fix this by simply moving the
white arrow on the right so it is under the graph at the spot where we
first see black. We do not want to move it any farther to the left than
that because then we would be pushing the image farther than it is
capable of producing.
Remember this is all subjective and if you want
to push it farther and you think it looks good, you are the artist and
you decide. I certainly have done so on more than one occasion, but
often the results can add unwanted pixilation to the image. You can see
that there is a thin line at the bottom of the middle slope and
sometimes you can push the arrow passed that line to where it begins the
slope. It isn't recommended though.
The photographer however will take this
technique one step farther and do this process individually for red,
green and blue. The reason being would be to balance the colors more
precisely but for most people this isn't always necessary.
|
 |
 |
For the purposes of this essay I will keep it
simple and process it in RGB rather than individually. So I will grab
the white arrow and slide it to the left until it is directly under the
the first bit of black. Comparing the images side by side it is easy to
see we have improved the image however slightly and it is looking much
more realistic. If the image was too bright then everything above would
be done the same just opposite from left to right.
If we see that the graph extends all the way to the
edges in both directions then we do not need to make any changes to the
levels. However it is a good idea to check the red, green, blue levels
individually because it is quite possible that one or more of them don't
extend all the way to both sides and they will need adjustment to
balance the colors. Even minute adjustments here can balance the colors
and greatly improve the image.
If the all the colors do indeed extend all the
way, the image still might not look bright enough or dark enough and we
can slide the midpoint (gray) arrow either way to change the midpoint to
darken or lighten the image. Take care here and try not to do too much
and also consider it may look that way because of your monitor settings.
And again it is all subjective and you, the artist, is the only one that
it has to please.
The next step I would take is adding some
'saturation' to the image. Saturation brightens all the colors by
removing gray and bringing all the colors closer to their more perfect
state. This needs to be done in moderation, but as above, it is totally
up to you and what you like. |
 |
To bring up the adjustment screen, go to the
top of the screen and choose Image-Adjustments-Hue/Saturation (Ctrl+U).
On this screen the only thing I'm interested in is the slider in the
middle for Saturation. Moving the slider to the right will add color, to
the left will take away color. If you move it all the way to the left it
will become black and white. For most images, moving it to the right no
more than to a level of 10 or 15 should be enough to brighten up most
images. Any more than that and the image will stop looking real. Again
it's totally up to you and again colors can be affected by your monitor
settings so it is best to tread lightly here. For this image I'm going
to bump it up to 50 so you can see a major difference. There isn't a lot
of color in this image to begin with so a little more won't hurt.
Comparing it with our level adjusted image side by
side you can see we have added a lot of color. The green leaves are
bright, the window lettering is bright and the stuffed bears now have
some color as does the brown paper. Again this is probably too much but
on a small image like this it is still looking pretty good to illustrate
what saturation does.
The next and last step for basic processing is
increasing 'Contrast'. Contrast is the enhancement of dark and light
areas and getting this right will mean the difference between a good
photo and a great photo. |
|
|
 |
There are several ways to adjust the contrast.
The easy way is to use the Contrast adjustment
(Image-Adjustments-Brightness/Contrast). I don't use this because it is
simply terrible. I don't get the results I want from it. The hard way is
to use the Curves adjustment (Image-Adjustments-Curves). I used to use
this but no longer. It is complicated and again I don't get the results
I want because I don't have the patience to use it correctly.
The method I use is a little known secret that
is easy to use and gives me the results that I like (again, subjective).
I use the strangely named "Unsharp Mask". The Unsharp Mask is a very
useful tool that is designed to sharpen an image in a very specific way.
Because it is used for sharpening, I have no idea why it is called 'unsharp'.
Anyways it's indended purpose is to find pixels in an image where dark
ones meet light ones. Then it will darken the dark ones and lighten the
light ones according to the parameters we set and and it will create a
nicely sharpened image if we don't overdo it. If you overdo it it can
create halos around objects which generally looks very bad.
To open the Unsharp Mask go to Filter-Sharpen-Unsharp
Mask. As noted above this filter isn't designed to add contrast. It is
designed to subtly add sharpness. You can see in the image below we have
the Amount, Radius and Threshold. |
 |
The amount being the amount of the effect we
want to have and be set anywhere from 0 to 500%. For sharpening purposes
this Amount could be anywhere from 100 to 200%.
The Radius is how far the it will look for
light and dark pixels to affect. It can be set anywhere from 0.1 to 250
pixels. For sharpening purposes in most digital situations this would be
a number like 0.5 to 1 at the most (again being careful how we use this
to avoid halos. 0.8 is probably as high as you want to go).
The Threshold will limit the lightness/darkness
comparison so pixels that are closer together in lightness/darkness
contrast will not be affected. The higher the number the more subtle the
effect will be because the less pixels will be affected. Only the pixels
that have a high contrast will be affected. This can be set from 0 to
255. I don't think I have ever used a number higher than 5 with 0 being
the norm and 1, 2 or 3 being enough.
You can see though, the numbers that are there
now are 20 and 20. This is a number that I stumbled upon that works for
me in a general way for many images I process. I can and will change
these values depending on the situation, but often times this is my
starting point for an overall contrast adjustment. So I'm going to apply
the effect to 20% and spread the affect to 20 pixels wide. The effect
will find the light and dark areas that meet each other and increase
them 20 pixels in any direction 20%. 20 pixels is obviously way more
than the 0.5 recommended above, but the results are what matters and
this is way easier than trying to do this with the Curves function.
For the purposes of this discussion, I have
bumped up the Amount to 40% so the change is more noticeable in these
small examples. If these were enlarged or printed it would be obvious
that I have overdone it, but in these examples it better illustrates the
point.
|
|
|
 |
In the comparison to the left, we can clearly
see the contrast has increased evenly throughout the image and details
are much sharper. Looking at the cat, there is better contrast between
the gray patches and the white. Within the white we can see more detail.
Same with the bears. Studying the 'hair', the facial features etc. all
have much more detail. |
|
|
|
Let's make one more side by side comparison
from where we started to where we finished.
Not bad! This whole process would take me no
longer than 5 to 10 minutes if that.
|