Original Link: https://www.anandtech.com/show/1648




Index

Welcome back to another guide to better photos here at AT. In our past guides, we have looked at techniques that can be applied to take better pictures. In this guide, we will be focusing on ways to improve the photos that you have with the use of Photoshop or Photoshop Elements. If you don’t already have one of these programs, and you are at all serious about photography, you should really consider checking them out. Photoshop Elements 3 is a reasonably-priced alternative to Photoshop CS that offers many of the same features.

The quality of a picture right out of the camera will vary depending on the camera that you are using and its internal processing settings. For example, consumer digicams tend to produce images with high contrast and sharpness while higher-end models tend to be more conservative. It is a common misconception that every image needs to be post-processed. Depending on your camera’s settings and partly on your photographic ability, you may find that some pictures will look great right out of the camera. In such a case, there is no need to feel bad for not post-processing your images. However, it is far more likely that when you look through your photos you will see some that need straightening, lightening, higher saturation, red-eye reduction, sharpening, etc. This guide is designed to help you post-process your images with simple steps to produce impressive results. Feel free to work with our samples as we go through this guide. It will help you get a better idea of how each process is applied.




Getting Ready

Before we jump right into editing, there are a couple of things that are very important to do. First, we highly recommend making a backup of your original image. Whether it’s to a CD, DVD, external hard drive, etc., just make sure that you have the original in a safe place and only work on a copy. By doing this, you are giving yourself the option to start over with the original. The second thing to do is to set the RGB color space in Photoshop. By default, Adobe Photoshop will use the sRGB color space. If your original images are in the sRGB color space, you can simply leave it as is (most consumer cameras shoot in sRGB). The sRGB color space is best suited for viewing images on a monitor (such as on a web page). Higher-end cameras often have the option to shoot in either sRGB or Adobe RGB. If your images are in the Adobe RGB color space, you should configure your Photoshop working space as Adobe RGB (1998). The reason for this is that the Adobe RGB color space has a much wider gamut than sRGB – meaning that you can work with a larger range of colors.


Photoshop Elements 3.0
Click to enlarge.

In Photoshop Elements, go to the Edit menu and select Color Settings. The above dialog will be displayed. You can see in the description that “Limited Color Management” will use sRGB and “Full Color Management” will use Adobe RGB.


Photoshop CS
Click to enlarge.

In Photoshop CS, go to the Edit menu and select Color Settings. As you can see, the color settings dialog in Photoshop CS is a bit more involved than it is in Elements. Here, you will just select Adobe RGB (1998) or sRGB from the RGB pull-down menu in the “Working Spaces” section. If you use Adobe RGB to edit your images and you want to display them on the web, you’ll have to remember to convert the image back to sRGB after editing it (which we’ll discuss later). We should point out that if your images were taken in the sRGB space, you won’t gain anything by selecting Adobe RGB.




Cropping

After backing up your image and setting the appropriate color space, one of the first things that you will want to do is crop the image. At this point, if your image is a bit crooked, we can kill two birds with one stone. Also, since the Crop tool works the same in both Photoshop and Elements, we’ll discuss cropping as it pertains to both versions. In our example below, we will straighten our image with the Crop tool.


Crooked image
Click to enlarge.

First, select the Crop tool from the Tools palette. Then, draw a crop box within the image (the size and placement isn’t important right now). Now, look for a line in your image that should be perfectly vertical or horizontal. Move your mouse to the outside of the crop box until the cursor turns into a double-sided curved arrow. Rotate the crop box until it is parallel with the edge that should be vertical or horizontal as demonstrated below.


Rotating the crop box
Click to enlarge.

In our example above, we used the left edge of the building as a guide. Once you have the crop box rotated the way that you want it, simply resize the box until the image is framed the way to your liking. If at any time you want to cancel the crop, just press Escape. Otherwise, press Enter or double-click inside the box to finalize the crop.


Straightened and cropped image
Click to enlarge.

After that little bit of work, we’ve straightened our image and cropped out some unnecessary space.


Options palette
Click to enlarge.

Another cool thing that you can do with the Crop tool is to specify the output dimensions. For example, let’s say that we plan on printing this image as an 8” x 10” print. Simply type the dimensions in the Options palette. You may have to type in the unit after the number (ex: “in” for inches or “cm” for centimeters). You can also type in the output resolution to the right of the “height” box (ex: 300 ppi). Then, rotate and resize your crop box like we did previously. This time, you will notice that the crop box will retain the appropriate ratio automatically. Now when you press “Enter”, your image will be in the right dimensions for an 8” x 10” print. If your digicam shoots images with an aspect ratio of 4:3 instead of 3:2, you will find this cropping method invaluable in getting your pictures ready for print. We use the Crop tool all the time to make thumbnails for our articles, which must be 60x75 pixels. Just specify “px” as the unit instead of “in”.




Basic Color Correction

In this section, we are going to discuss one simple way to remove color casts from your images. Although this method cannot always correct heavy color casts, it does a very good job with subtle ones. It can also add a good amount of contrast to your image. This method assumes that your image has a point that should be pure black and a point that should be pure white. If your image is comprised of all subtle midtones, you might want to skip this method.

When most people think of color correction, they might think of the “Color Balance” option under the Image/Adjustments menu in Photoshop. Unfortunately, using the “Color Balance” option can be very imprecise. Our method takes the guesswork out of the process so that you can simply find and specify the white and black points of your image. In this method, we will be using the Curves dialog box. Although Photoshop Elements does not offer Curves out of the box, we ran across this interesting website that shows you how to get some Curves functionality in Elements. However, since this Curves manipulation works a bit differently, we will be focusing on Curves in the standard version of Photoshop.

1) Open up the image that needs some color correction and set the view to “Fit to Screen”.


Original image
Click to enlarge.

In this image, our Auto WB setting produced a strong yellowish color cast. Using our method, we will try to get the colors to be more accurate.

2) Click the “New Adjustment Layer” icon (half white/half black circle on the layers palette) and select “Threshold”.


Create a new adjustment layer (Threshold)

3) In the Threshold dialog box that pops up, drag the slider on the bottom all the way to the left. Now, move the slider very slowly to the right and stop when you first see some black patches appearing on your image. In this step, you are actually revealing the darkest areas of your image. These will most likely be shadow areas.


Find the black point
Click to enlarge.

4) Click “OK” in the Threshold dialog. Then, change the magnification of the image to “Actual Pixels”.

5) Hold down the mouse button over the Eyedropper tool on the Tool palette. Then, select the Color Sampler Tool from the flyout menu.


Select the Color Sampler Tool

6) Click inside the dark patch with the Color Sampler Tool.


Sample the black point
Click to enlarge.

7) Set the view to “Fit to Screen” and double-click the Threshold Adjustment layer again. Now, drag the slider at the bottom all the way to the right. Slowly bring it back to the left until you see some white patches appearing. These are the brightest areas in your image. Repeat steps 4 – 6 to mark the white point with the Color Sample Tool. You do not want to set your white point on specular highlights. A specular highlight is an area that is particularly bright because of a reflection. These spots may be overexposed and it is usually desirable to leave them this way. Instead, select a white patch that corresponds to a white object in the image. You could repeat this process for a midtone as well, but these tend to be harder to spot. Don’t worry, we’ll adjust the midtones in a second.


Sample the white point
Click to enlarge.

8) Now that you have marked the points for the darkest and brightest parts of the image, we are finished with the Threshold layer. So, simply drag that layer into the Trash to delete it. Our next step will include inputting our black and white points into the Curves dialog.

9) We can either work directly on our image or we can use a Curves Adjustment layer. The benefit of using Curves on an Adjustment layer is that we can always go back and edit the Curves layer or delete it entirely. If you decide to work directly on the image, you will not be able to change the Curves settings later. So, click on the New Adjustment Layer icon and select Curves.

10) The first thing that we want to do is to tell Photoshop how to define “Black” and “White”. Starting with black, double-click the black eye-dropper icon on the left (circled in red in the image below).


Curves dialog


Define RGB values

Here, we will be telling Photoshop what the RGB values are for the color black. For an image destined for the web or monitor, you can leave the RGB value at “0 ,0,0”. However, if you plan on printing your image, we have found that a value of “20 ,20,20” will produce better results. Otherwise, detail in the darkest part of the image will not be visible. Click OK and double-click the white eye-dropper icon. For the white point, specify an RGB value of “255,255,255” for the web or monitor, and “245,245,245” for print. When printing your images, the most effective values will vary depending on the output device. Therefore, we recommend making some test prints with differing values to see which is the best for you.

11) Now that you have defined the RGB values for black and white, we can tell Photoshop where these points are in our image. From the main Curves dialog, select the black eye-dropper tool. Now, move over to your image and click on the sample point for the darkest part of your image. Then, click the white eye-dropper tool and select the sample point on the brightest part of your image. You can also use the “ Set Gray Point” eye-dropper if you have a neutral midtone in your image. If everything went as planned, you should see a difference and an increase in contrast. If you make a mistake at any point and want to start over, simply hold down the Alt key (Mac: Option) and the Cancel button will turn into a Reset button.


Image after setting black and white points

12) Next, we want to boost the midtones. To do this, simply click on the middle of the diagonal line in the Curves dialog. Then, drag it up very slowly. You will notice the midtones increasing in brightness.


Brightening midtones

13) Finally, to add a bit more punch to the image, you can add a second point on the curve and create an “S” shape. You will see the image increase in contrast dramatically. You will probably need just a very subtle “S-curve” before your image has plenty of contrast.


Creating an “S-Curve”

 
Before curve adjustment
 
After curve adjustment
Click images to enlarge.


Hold mouse over image.





Levels

Almost every photo can benefit from a levels adjustment. We are going to show you some examples that demonstrate how a levels adjustment can bring life back into your photos. This is particularly effective for images that are too dark. The levels command in Photoshop and Elements allows you to remap the brightness levels in a photo resulting in a high contrast image.


Original image
Click to enlarge.

When we decided to photograph the image above, the leaf attracted our attention because of its bright red color contrasting against the hood of a light blue Chevrolet Caprice. Unfortunately, the image right out of the camera doesn’t look like how we remembered it. With two simple levels adjustments, we can fix this problem. Once you are familiar with this process, it should take you less than 30 seconds to perform.

1) Open the Levels dialog by pressing Ctrl-M (Mac: Command-M). Again, you could just as easily use a Levels Adjustment Layer if you think that you might want to alter it later. The first thing that you will see is a histogram. A histogram displays the brightness of an image starting with the darkest information on the left to the brightest information on the right.

By looking at the above histogram of our leaf, we can see that most of the light information is in the midtones. There is a considerable amount of shadow detail (left), but there is virtually no highlight detail (right). What we want to do is “stretch” the light information over the whole histogram. In other words, we want to remap the brightness values so that our histogram looks more balanced with shadow, midtones, and highlight detail. As you will see, this is actually very easy to do.

2) Click on the black slider on the left directly under the histogram (circled in red below). Slowly drag the slider to the right.

To see exactly what you are doing, make sure that you have the “Preview” box checked and hold down the “Alt” (Mac: Option) key. With the Alt key held down, you can see exactly when a color channel begins to clip. Drag the slider to the right until just before you see the clipping.


Dragging the black slider to the right while holding “Alt”
Click to enlarge.

In the screen capture above, you can see that we stopped just as we started to clip some of the color.

Now, use the same technique to drag the white slider to the left and stop just before clipping any color channels. You should already see a tremendous difference if your photo is anything like ours.


Dragging the white slider to the left while holding “Alt”
Click to enlarge.

Again, with the white slider, we stopped just before any serious color clipping occurred.

Here are the results that we achieved by simply adjusting the black and white sliders:

 
Original image
 
Image after levels adjustment
Click on above images to enlarge.

Original histogram

Histogram after levels adjustment

You can clearly see how we’ve “stretched” the histogram to cover more of the brightness range. This has increased the contrast and given the photo more “pop”. Below is a final comparison where we pulled the white slider just a bit further to the left. Although this causes some slight color clipping, it is hardly noticeable and the effect is just what we wanted.


Hold mouse over image.





Red-eye Reduction

Given the time and patience required, everyone could remove red-eye by hand using a paint brush in Photoshop or Elements. The point of this section is to show you how to do it as quickly as possible. Since the technique is a bit different between the two programs, we'll explain the process with Photoshop and Elements separately.

Here is the original image we will be working on.

Photoshop (CS)

1) Open the image with the red-eye problem and set the magnification to "Actual Pixels" so that you can get a good look at those red eyes. In our example image, the resolution of the original file is only 1 megapixel, so we are going to magnify the image by 300%. Next, hold the left mouse button down on the "Healing Brush Tool" and select the "Color Replacement Tool" from the flyout menu.


Color Replacement Tool

2) Press "D" to reset the colors to their defaults and make sure that Saturation mode is selected from the Options palette.

3) Select a brush that is about the size of the red-eye area and click once in the center.

4) Repeat for all other eyes in the image.

5) As an optional final touch, we like to get the Burn tool from the tools palette to darken the pupils a bit.

Photoshop Elements 3

1) Select the Red-eye Removal Tool from the tools palette.


Red-eye Removal Tool

2) Click and hold the left mouse button and draw a box around the red eye area. When you release the mouse button, the red-eye will be replaced with black. We found the default settings of "Pupil Size: 50%" and "Darken Amount: 50%" to work just fine for us.


Drawing a box around the red-eye area

In the end, we couldn't see much of a difference in effectiveness between the two programs. However, Photoshop Elements was a bit more efficient by darkening the pupils for us.

 
Photoshop CS
 
Photoshop Elements 3
Click images to enlarge.




Sharpening

If you didn't know already, one of the best sharpening tools available in Photoshop and Elements is called Unsharp Mask (under the Filter\Sharpen menu). It should always be one of the very last things that you do to an image after editing and resizing. Because sharpening is such an important part of making a good image look great, we are going to give you some basic input boundaries for the Unsharp Mask (USM) dialog. Also, because we often see over-sharpened images on the web, we are going to provide some examples of what over-sharpened images look like.

First off, do all your exposure adjustments, color correction, etc. Resize your image for whatever application for which you intend the image to be used. Our example below is going to be used for the web, so we have resized it to 535x800 pixels (the thumbnail is 300x449 pixels).


Original image
Click to enlarge.

1) Set your image magnification to Actual Pixels. To do this, double-click on the Zoom tool in the tools palette. This is very important so that you can see the effect on your image as you adjust the parameters.

2) Open up the USM dialog by going into the Filter\Sharpen menu.


PUSM dialog

You'll see three adjustment sliders: Amount, Radius, and Threshold. Unfortunately, there's no way for us to tell you what to put in these fields. The best settings will depend on the detail in the photo, original sharpness, and your subjective sharpness tolerance. However, we can tell you what the adjustments will do. Think of the Amount slider as a gain control; it will determine how much sharpening to apply. Since Unsharp Mask is essentially sharpening the edges in your image, the Radius will determine how many pixels from the edge to sharpen. Finally, think of the Threshold as a tolerance slider. If the Threshold is lower, more pixels will be considered part of the edge. If the Threshold is higher, fewer pixels will be considered part of the edge. In other words, a higher Threshold will result in less sharpening. Now, let's look at a real example.

3) In the image below, we indicated a moderate Amount of 150%. By choosing a Radius of 1, we were able to bring some definition into the eyes. For the Threshold, we specified 4 to keep the background grain from being sharpened too much.<


Click to download sharpened image.


(Amount 255%, Radius 2, Threshold 6)
Hold mouse over image.

As you can see, with a Radius of 2, we start to lose significant detail to halos at the edges. The only way to become experienced with Unsharp Mask is to experiment with it until you know what works for different types of images. As we said before, images destined for print can be sharpened more than images for the screen or web. Just create some samples and compare the printed results.




Color Space

Color space is a topic that can and does fill entire web sites and books. In fact, it can easily give some people headaches and hours of frustration while trying to get their images to look the same on different mediums (print, screen, etc.). At the beginning of this guide, we recommended setting the color space to Adobe RGB 1998 if your images were shot in the Adobe RGB color space. Let’s say that you started out editing an image in Adobe RGB, but now you want to upload that same picture to your web site. The most common mistake that people make is to just upload the image without changing the color space to sRGB. Without giving you a headache, we are going to show you how to get your images to look right on the web.

First of all, you should know that these color space issues mostly arise from the ability of some cameras (mostly higher-end) to shoot in the Adobe RGB color space. If your camera only shoots images in the sRGB color space, you should just leave Photoshop’s color space as sRGB and be done with it. By doing so, you won’t need to worry about conversion for the web.

Whenever you save an image in Photoshop or Elements, the dialog box will have an option to tag the file with the ICC (International Color Consortium) profile. The screenshot below is from Photoshop CS, but it may not be available in some earlier version of Photoshop.


“Save for Web” dialog showing ICC option

You might be thinking, “Great, so I just need to check that box?” Unfortunately, it’s not that easy. The problem is that most web browsers are not “color aware”. This means that they totally ignore the embedded ICC tags and assume that all images are sRGB. On PCs, IE, Firefox, and almost all other browsers, they are not color aware. On Macs, Safari will recognize the embedded profile and display the image appropriately. So, what happens when you save your Adobe RGB image and throw it on the web without converting it to sRGB? Below is an example of an untagged Adobe RGB image compared to an untagged sRGB image.


Original image is tagged Adobe RGB
Rollover image is tagged sRGB
Hold mouse over image.

Assuming that you are viewing this in a non-color aware browser, you can see that the tags make no difference. The browser does not even process the embedded color profiles and just assumes that all images are sRGB. This works fine for the sRGB images, but incorrectly displays the Adobe RGB images. If you take the tagged files above and open them in a color-aware application, they will both display properly because the application will recognize the embedded profiles.

Given that most browsers are not color aware, the only way to win in this color struggle is to make sure that all your images are sRGB before uploading them to the web. Fortunately, this process is very simple provided that you are not using Elements. For some reason, Adobe decided not to include the option to convert color profiles in Photoshop Elements. For that reason, the steps below will only work for the standard version of Photoshop.

1) Open up your Adobe RGB image and select “Image\Mode\Convert to Profile”.


“Convert to Profile” dialog box

2) Select “sRGB IEC61966-2.1” in the “Destination Space” section. Then, select OK.

That’s it! Now, you can save your image and post it on the web. If you use the “Save for Web” option (to optimize the file size), you can check the “ICC Profile” box to embed the sRGB profile. This will add a little bit to the file size, but this way, you can be sure that people with color aware applications will view the image with the appropriate colors.

Hopefully, this color space thing didn’t give you too much of a headache. After familiarizing yourself with the concepts, it will become much easier to think about color space. Unfortunately, there’s no way around learning this stuff if you want to have consistent color in your images. Now that you’ve read through this guide, we are certain that you can apply some of these ideas to your own images to make them better. Even if you just apply the Levels method, you will be on your way to much more impressive results. In future guides, we have plans to cover more Photoshop techniques and tricks. As always, let us know if there is something in particular that you’d like us to cover. Happy Photoshop-ing!

Log in

Don't have an account? Sign up now