Microcosmologist

Photoshop tutorial number three dishes the dirt about my personal favorite adjustment in Photoshop: selective color! This is a completely radical function that often gets overlooked.

When you have a photo that just seems kinda blah, the first instinct is, “beef up the color!” And yes, it’s not a bad idea. But there are a hundred different ways of going about it, some of them certainly better than others. Let’s start with the worst: saturation. Cranking up the saturation indiscriminately makes every pixel in your image more of a pure color. A boring red becomes bright red. What separates secret sauce from lame sauce here though, is that saturation will leave you with blown-out colors. There’s no subtlety and you start losing detail.  As an example, I’ve got a lousy photo here, with some dude doing a dirtbike jump.  At least this lousy photo does feature motocross, I guess that’s cool.  Have a look at the oversaturated image: the tree at left is RED, the grass still really isn’t green, and the sky looks a little fake.

A bit more refined than the saturation is the vibrance control in Photoshop. I think they basically put this in because they couldn’t stand people cranking on the saturation any longer. Adobe said, okay, if you really can’t be bothered to figure out how to properly adjust your colors, and you want to juice things up using only one slider–use this one. It’s basically a saturation control that only affects dull colors. So now, instead of blowing out your already blue sky into something totally unrealistic, you boost up your boring colors into something more interesting, and leave the ones that were already cool alone. Ahh, that’s better already.

But vibrance is still a blunt instrument. We need something with a delicate touch to really make our colors look great but not fake. Enter selective color.

Selective color, as the name implies, lets you individually edit colors. Is the grass in your photos just not bright enough? Edit the green! Is your sky sort of pale? Edit just the blues! Not only can you tweak the colors of your photo one at a time, you can change their hues as well. Selective color is like a hue/saturation/brightness control for each individual color. And it’s the secret sauce when it comes to beefing up ALL your colors in a controllable way that won’t leave you with some colors blown out and others still muted.

I originally started using selective color as a means to correct skin colors. To my eye, I don’t think I’ve ever seen a digital camera that correctly renders human skin tones. Which is kind of surprising, considering the popularity of people as a subject for photographs! It seems to me that skin is *always* too reddish, often pale looking with too much magenta in it. Not flattering. So what I like to do is select “Red” for my color, and then adjust the three sliders for Cyan, Magenta, and Yellow. You’ll get decent results if you drop down the cyan just a skoch, reduce the magenta about a quarter, and boost the yellow very slightly. Just about anyone’s skin color will now look better after that one step. As I did this to photo after photo, I started to realize how powerful and subtle of a tool selective color can be!  Allow me to demonstrate on John Medeski.  In the middle of a melodica solo:

What you’re doing is adjusting the hue of the reds, making them more yellow.  As you probably noticed, changing the “Red” color composition affects anything red anywhere in the image, not just a person’s skin. (Like the Nord Electro keyboard in the image above. If there are certain areas that need a lot of correction and others that should be left alone, use feathered selections around your critical areas.) You may recognize the fact that Cyan, Magenta, and Yellow (Black is just all three put together), are the three primary colors for printing or painting. In other words if you have these three colors of ink, you can make any color imaginable. For example, yellow and cyan make green. Cyan and magenta make blue. Etcetera.

See where this is going? You can use selective color for way more than just correcting facial tones. And you can use it to make each color more of a “pure” color, with very fine control. To make a nice deep blue with ink, we’d combine cyan and magenta. No yellow. So to intensify your sky, turn the yellow slider way down, and start boosting the cyan, and to a lesser extent, the magenta. Viola! Deep blue skies, with every other color in your image unaffected! It’s science.

The same idea applies to every other color. To make green with inks, we’d combine about equal parts cyan and yellow, with no magenta. So select “Green” in selective color, and drop the Magenta slider way down. BAM. Your greens just got a lot greener, without oversaturating them or messing up the delicate balance you just achieved with the blues.  To illustrate what I mean, I went overboard in the screenshot below to show how much you can change a single color.  The green grass is ridiculously green, while the sky–perfectly boring blue.  (full disclosure: the “yellows” have similar settings here to the greens, otherwise everything is neutral.  Yellow and green tend to be inter-related)

Maybe at this point I will interject the nuance that you can either be additive with this process or subtractive. I prefer to be mostly subtractive, because I think it avoids oversaturating the picture. That is to say, most of my sliders are below 0, and only a few go above it. If you go crazy on the positive end of the sliders, you can make just as obnoxious of a result with selective color as you would get by cranking the saturation like a rookie would. This is what we’re trying to avoid.

And I’ll add the usual caveats: don’t overdo it. Make sure your image still looks “real”. Check and uncheck the preview box now and then to keep yourself under control. It’s easy to get carried away. When I first discovered how to use “compression” doing audio mixing, I went nuts and compressed the daylights out of everything. At the time I thought it sounded awesome and “thick”. Now, looking back at some of the mastering I did shortly after discovering this powerful new tool, it seems way overdone and amateurish.

Subtlety. Is key.

Curves!  Probably the most useful and basic tool in Photoshop for making adjustments.  You can find it under image > adjustments > curves.  For the uninitiated, a common use of curves is to increase contrast in a photo.  If you make the good ol’ “S” curve using 3 points, you can add contrast to any image.  Just like you see below!  I’ve added a small change to my S curve: the middle point is raised above center, instead of being dead center.  This has the effect of brightening the overall image as well as adding contrast.  How does that work, you ask?  Read on!

Basically, if you think about the labels on each axis here, you can understand what’s happening.  You have a point with a certain brightness, let’s call it “50” out of 100, smack in the very center.  That’s your “input”.  The way the photo is, already.  Your “output” is what you want to do to it, i.e. make it brighter or darker.  So for “50” which should be exactly in the center of the grid, we’ve raised it a bit, meaning it’s now brighter, maybe a 58 or so.  With the S curve shown, our darks are just a touch darker, our midtones are a touch brighter, and our brights got a noticeable boost even brighter.

The photoshop trick that’s really been growing on me lately is using curves to tweak white balance. Even if you had the right white balance selected when you took a given photo, there’s often a subtle color cast. Messing around with this color cast can sometimes give a real nice feel.  Or, if you’re like me and you sometimes forget to set your camera back to the right setting when you walk from indoors to out (and auto-WB never seems to get tungsten right), you can use this technique to correct photos that are wayyy off.

There are three eyedroppers in here that aren’t really labeled. I put a nice bright red box around them in the screenshot so you’ll notice them–I overlooked them myself for a long time. You can use them to pick the black, gray, and white points for a photo, assuming your picture has those three colors. (Most do.) It’s pretty simple. Picking the white point usually yields the most dramatic effect. Click the white eyedropper in the Curves window, then start clicking on places that are completely saturated, or the pixels which are immediately adjacent to a saturated area.  In the image below, I picked a bright white spot on one of the apples, reflected from the lights above.  Now this point is the “new white”.  Viola-instant white balance correction!  Note that you can lose detail here by selecting something darker than absolute white, because whatever lightness you click on now BECOMES absolute white with the resulting new curves. There are some really interesting changes in the color cast of the photo when you start clicking around.

Try a lot of points, and notice that your color cast is the opposite of whatever you click on. For example, if you click on an area that’s blue-ish, your resulting white balance will shift the colors toward red-ish. It’s trying to compensate against whatever your clicking on. If you click on a “cool” area, it will “warm” the photo. The inverse is also true.

You can do the same process with the black and the gray eyedropper, but a lot of times I find that the black point really doesn’t do much to affect the colors, and it’s hard to really get a good shade of gray in your average color photo. Maybe my technique is still maturing in this territory though.   If you do have a shot with a nice even gray in it, like concrete, definitely try using the gray eyedropper on that area.  It can produce some nice changes in the shades of your colors.

Finally, advanced students will note that if you wanted to, you could achieve the same thing by manually moving the individual color curves.  Sometimes, for examples like the picture below, where your camera was completely on the wrong white balance setting when you took the shot, you might want to apply the curves by hand to get it just so.  Go give it a try!

This is the first in a series of Photoshop tutorials which will detail how I like to spice up (“let’s enhance!”) my photos.

The very first thing I like to do to just about any picture I edit is to hit it with some Shadows/Highlights. This option can be found under the Image -> Adjustments menu. What it does is brighten up the darker areas of your photos (that’s the “shadows” part), and darkens down the bright areas (that’s the highlights part). You might say, well, I already do that with the ‘Curves’ or some other adjustment. As with many image adjustments, there’s a hundred ways to roughly accomplish the same thing. Personally, I feel that Shadow/Highlight gives the most pleasing appearance, and allows me to push the levels as hard as I want without making the image look obviously altered.

The more important of the two options is definitely shadows. There’s three sliders here: Amount, Tonal Width, and Radius. Amount is how hard you want to apply the overall effect, and Tonal Width is how wide of a range of the dark colors do you want to apply it over. A small Tonal Width would affect only the darkest of the blacks, versus a wide Tonal Width which would affect dark blacks and some medium blacks. There’s really no golden formula with these; I think every photo I do has slightly different settings. A good basic guideline is simply to get a noticeable boost of detail in the dark areas of your picture without making it look like it was edited. Notice in the tire above, you can start to see some detail there, but it’s not bright as daylight.  That’s what you want; a little extra detail. If you crank the sliders your picture will start to look like one of those overdone HDR pictures that everyone loves to hate on. Don’t crank the sliders. Don’t be that guy. Remember: just because you CAN get detail inside every shadow definitely does not mean that you SHOULD. Such a photo will look unrealistic, like a special effect in a movie that sticks out because it’s just too clean looking. Get a modest boost in the detail, and then start adjusting your Radius slider.  The image below shows TOO MUCH of a good thing.  Drowning your steak in steak sauce does not enhance the flavor.

The Radius controls how small regions adjacent to one another will be affected. Start moving it around and you’ll see what works best. What you want to avoid is what I’d call “bloom”. That is, gradients which are obviously caused by some kind of brightness editing, such as a noticeable brightness shift at the horizon in your sky. Again, the results should look natural.

The same concepts discussed above apply to the Highlights section. The main thing I’ll say about highlights is that I rarely use it! And if I do, it’s usually with a small Tonal Width and a small Amount. Only few and far between will come the photo that actually benefits from the Highlights enhancement. Even if you’ve got a strongly overexposed image, it often looks better to just make it look like you’re intentionally blowing things out, rather than trying to tame it into being a normal exposure.  The above image of the Metra train was made with excessive highlights applied, creating the halo effect that plagues so many poorly done HDR images.  Avoid that halo whenever possible.

Lastly, the controls at the bottom of the window midtone contrast and color correction are best left at neutral settings. Maybe a mild color boost through using the color correction slider, but please, go easy. You can get far, FAR superior color using selective color, which we’ll talk about soon! As for contrast, the best way to adjust that is in the curves dialog, and I think it makes the most sense to do that at the end, when you’re done tweaking other things like color and sharpness.