Photo-Editing for Beginners

The whole idea of digital cameras and their popularity seems somehow to be linked with the modern urge for instant gratification. With a film camera, you shoot a roll and maybe you remember to get it processed. If you want the pictures bad, even one-hour photo is a long time, and when you have to wait you get tied up with other things…you get the picture (no pun intended). With digital, you have the picture the instant you take your finger off the shutter button.

Well, maybe.

Certainly all digital cameras have at least an option to let you pull an image out of the camera and use it "as-is". A lot of really good photographers with lots of experience often make this "no-touch" method work for them by carefully setting the camera's controls and watching lighting and exposure critically. Unfortunately I'm not one of the people who can do that. To be kind to myself, careful preparation and controlled settings don't go well in wildlife photography, which is primarily what I do. If you're not going to be able to get everything set up perfectly, then you're going to have to fix the image after you've shot it.

Software to alter digital images is usually called "photo editor" or just "photo software". Many of the digital cameras will come with software, and Microsoft, Adobe, Corel, and a bunch of other vendors all have photo-editing products available for purchase. Sometimes a version of one of these commercial packages will be provided with a camera. With a photo-editor, you can correct many of the problems that are likely to pop up in your digital picture-taking. A good photo-editor is essential if you can't go back and take the shot over if you've messed up. It makes the difference between having a picture (maybe not the best, as we'll see) or having nothing.

Before we start, I do want to point out that there's no substitute for doing a good job when you take the picture in the first place. Digital cameras and photo-editors cannot fix some cockpit errors. Among them are out-of-focus shots, motion blur, and extreme over- or under-exposure. You've gotta get things at least close to right. So to open our tip on photo-editing, let's start out with some rules to apply before you crank up the computer and start diddling. These are rules that are applied where the rubber meets the road-when you're taking the picture in the first place.

Rule number one is to get to know your camera and what it can do. Try shots under a lot of different conditions and with different settings until you have a decent understanding of your gear. One nice thing about digital is that it doesn't cost you anything for film or processing. Burn those electrons like they were water! When you're comfortable, you can move on to taking pictures for real.

The second rule is to check the results of any shot while you're still there on the spot. All digital cameras will play back an image, and some will even show you a chart or histogram that depicts the light-and-shadow distribution in your shot. If you're taking pictures of running wolves who probably won't go back and rerun their course for you on command, take a shot of the area before you start looking for wildlife to get your exposure set. While you're doing this, check the shutter speed and lens opening (if they're accessible) to insure that you're not trying to shoot running wolves with a half-second shutter speed. Heck, at that speed you could get one of your travel companions to streak through the scene and claim the resulting blur was a wolf! A running animal moves at about 20 feet per second, so your wolf will have moved ten feet in the time the shutter was open.

Rule number three is to take multiple shots (preferably with somewhat different settings-what pros call "bracketing") of anything important where possible. If you are shooting a critical shot without a tripod and are afraid of camera shake, blow off two or three extras in the hope that at least one will be steady.

Got that? OK, let's say you've done all this stuff and you are still interested in photo-editing. How do you go about it?

Picking Your Editor

The premier photo editor is, without much question, Adobe's Photoshop. Now a part of Adobe's new "Creative Suite", Photoshop lets you do about anything you're likely to ever want (or be able) to do with a digital image. Photoshop's an expensive product, though. The retail version is about $700 bucks if you don't have an older version to upgrade, and the new CS version is licensed for one computer and registers to insure you don't use it on more than that. If you're serious about digital photography but not that serious, there are some down-home options you may want to consider.

First and best is a program called "Photoshop Elements". PE is a stripped-down version of Photoshop designed more for the typical consumer than for the professional or graphic artist. You can get PE for less than a hundred dollars; Version 4 is the current one, and it's darn good.

Another option that may be good for many is Corel's PhotoPaint. Available as part of a low-end product called "Essentials", PhotoPaint is a decent editor with features at least on a par with Photoshop Elements, and in many areas as good as Photoshop itself. Since Corel Draw, a graphics and presentation package, comes with "Essentials" too, some may want to go this route.

Other editor programs are available (Paint Shop Pro is a popular one), but I can't recommend them; I have no experience in their use. Frankly, it's very likely that if you really want to do good editing you're going to want either Photoshop or Photoshop Elements. The techniques used in each program are similar, and there's a host of good documentation and tips available for them. If you can, get one of these packages.

For the really and truly adventurous, there's a very powerful open source (meaning "free") image editor called GIMP. The Windows version of GIMP is available at http://www.gimp.org/~tml/gimp/win32/ and there's documentation of a sort on the homepage http://www.gimp.org. There are free plugins for GIMP, and it has incredible flexibility and capability, but the user interface is very different from the Windows standard and the documentation is best used by someone who has some technical skill in both computers and imaging. If you fit the bill, try it out. I believe in the open source movement, and you can find more on it and on other open source software at http://www.sourceforge.net. Check it out.

When you select an editor, before you even try to use tips like this further, play with it, read the manual, and get at least a nodding familiarity with its features. If you want to do serious digital image processing, buy a how-to book or two. One should cover your photo editor as a product, and others should deal with specific issues like color correction or fixing digital images. Where a choice of Photoshop or Photoshop Elements will really help you is that nearly all the good books on how to do things with digital photos are written for Photoshop. Many of the tips can be applied with Elements as well. Some may be workable with other products if you take the time to figure out the equivalent step. Whatever you decide, take the early study process seriously. If you just start fiddling, you'll get into a lot of sub-optimum habits.

This tip is a half-book as it is, so I'm afraid I can't cover how to do everything in all the editor choices. I'm going to use Photoshop as an example, with comments on the small differences in Photoshop Elements.
Workflow and You

Besides "Photoshop", the word you hear most often from pro digital photographers is "workflow". This nifty sounding term basically means the procedure you follow in processing a digital image. The idea is to establish a very structured way of approaching the task of editing a picture, so that you get consistent and good at it. Digital workflows tend to be very personal; people trade tips on them all the time and everybody generally goes back to the way they've learned to do it themselves. It's almost like the superstitious set of things that players do before a big game. Still, workflow does have its basic elements, and that's what we're going to cover here.

The first thing you need to do when processing a digital image is to save your negative! Before you go berserk and wonder why we're taking "negatives" when we left film and chemistry behind to go digital, let me explain that I'm talking about a digital equivalent of a negative. Your camera produced an image file, and you downloaded it somehow into your PC. Before you do anything at all with that image file, you need to preserve it. Digital editing is destructive. You'll learn new and better techniques every week for the first couple of years. If you ever want to go back and correct what you now know are your early stupid amateurish mistakes, you'll need something to go back to.

When you get a bunch of images loaded onto your PC, scan through them looking for ones that are clearly beyond hope. You know the type-hand in front of the lens, forgot the flash and it's all black, photographed the sun and it's all white, moved the camera or the subject moved…we could go on. Save a defenseless humanity from this stuff and delete it now. You can pretend these shots never happened. What you have left over is the stuff you have hopes for, and those are the shots you need to preserve.

When your "hopeful" shots are culled out, the first step is to rename the digital files into something at least semi-logical. Cryptic file names like "DCS0020.jpg" aren't going to help you find Auntie Maud's image two years from now when your buttering her up so she'll put you in the will. You'll need a system to name your files. The simplest is one that is based on date: "2004-01-27-001.jpg" might be the first shot taken on January 27th, 2004. If you use your camera as I do, to record trips or adventures, you can assign a leading code to identify the trip, then number sequentially by day of trip and number of picture. That's the system I use. My Churchill trip was "B" in my new naming, and the next two digits are the day of the trip, with the last four being the image number within the day. If you don't like that system, make up your own.

Renaming can be done by hand, but it's a pain in the neck. A better approach is to get an file renaming utility. I use one called "A.F.5 Rename"; it's free and you can download it at http://www.fauland.com/. Learn to use the one you get so you don't mess things up at this early stage! You can also get a product that renames while acquiring the images. Photoshop Elements does that, but some may not want to clutter up an image laptop with big software packages. Another tool is ImageIngester (http://basepath.com/ImageIngester/) which is an absolutely wonderful tool for grabbing raw files, converting them with Adobe's DNG converter to compress them, renaming them, and saving them.

Now that you have your files named, there are two ways to create a "digital negative". The best way is to save the shots, exactly as they came from the camera or compressed with Adobe's free Digital Negative Converter, on something durable like a CDROM. A second way, for use when your camera saves its images in a compressed "JPEG" or "JPG" format, is to convert the images to a lossless format like Photoshop's own .PSD files or the generally usable .TIF files, and then archive them to CDROM. In theory, there's no reason you have to convert and then save, and your files will take up a lot more space if you convert them to a lossless format. In practice, it may save you a problem later on-I'll get to that point in a minute.

When your files have been archived to CDROM, you're ready to go with digital editing. That brings us to the first rule: You will always start your editing process with your digital negative, never with the result of previous edits (unless you're doing something very minor). Copy your negatives back from CD to your hard drive to work on them if you've erased them or edited them before. This is particularly true if you're using the JPG format, because that compression scheme loses small elements of detail every time you save the file, and if you continually resave a JPG you're likely to end up with visible defects in the image.

The workflow for actual digital editing is a phased process. We're not going to teach you how to do all of these phases because the technique will vary depending on what editor you have, but we will guide you through how the process should work at a high level and what each step will accomplish for you. The best sequence of actions to take in optimizing a digital image is probably this:

Examine the image carefully, looking for anything that's a defect (dirt on the lens or CCD, for exmple), an unwanted reflection or red-eye, or a small object that's intruding into the shot and distracting you (overhead wires are a favorite). You'll want to use your editor to remove them.
Use the editor's color, lightness, and contrast tools to adjust the overall image appearance to the most natural level (or to the unnatural level you like).
Increase the "sharpness" of the image using one of several sharpening tools.
Provisionally save your work at this point, if you're going to multi-purpose the image.
Size and format the image for each of the uses you plan for it-print, email, etc. You'll start each process with the version of the image you saved in Step 4 above.

We'll take each of these steps in order now, in a bit more detail. Remember, you'll need to consult the manual or help file for your editing software to find out how to perform each of these steps in your particular product.

Fixing Problems in an Image

Many otherwise good snapshots (digital and otherwise) are ruined by some little thing. A good example is a big dot on the end of the subject's nose, caused by a gnat landing on the lens (or on the subject). Others are ruined by a reflection, red-eye, or some trick of light. In the old film days, it was difficult or impossible to fix these things. With digital, it's easier.

First, we've got to separate little problems from big problems. A little problem is a spot, or dot, or blob. A big problem is a tree, a person (maybe you'd like your ex out of a shot). The reason is that you need a different approach depending on just how much of the picture is impacted.

If you have a little dot/speck problem, most photo editors will let you use a tool called a "clone stamp" or "rubber stamp". This tool works by lifting a piece of the image from one place and stamping it down in another. The trick is to find a place in the image where there's a good sample of what should be under the defect, lift it with the stamp, and deposit it where the image has a problem. You should adjust the size of the stamp to be just a bit larger than the size of the defect, or use a small stamp and take several whacks at it. In effect, you're painting good stuff over bad. You have to work carefully, both in selecting where you got the "good" sample from and how you stamp it onto the new area.

Full versions of Photoshop have a better tool called a "healing brush". It works the same way as the clone stamp, but it does a better job of blending in the new textures. Try both tools; one will work best for any given situation, but not always the same one.

For a big problem, clone stamping would take too long and it would produce a kind of unnatural texture. To fix a big problem like a tree growing out of someone's head, you have a couple of approaches that might work.

Approach number one is the "select and paint" approach. Photo editors have various tools to select a piece of the image to work on. One way is to carefully draw around the edge of what you want to select, using what's often called a "lasso" tool. Another way is to touch the area with a "magic wand" that then selects adjacent pixels of the same general color as you touched. Again, try each of these approaches a couple of times to see what works for you. When you have something selected, you'll generally need to "expand" your selection a few pixels to get rid of transition color (leaves on the tree growing out of your uncle's head, for example). When the faint dotted area that marks your selection is covering what you need covered, you're ready for the painting step. You first sample the color of the sky (using our tree example), and then paint over the selected tree area using that color. Get a sample of sky from nearby so the shades will match.

Painting works only if the background behind the object is a flat uniform color. If it isn't, you need to replace the offending object with something that looks like what would be behind it. That process also starts by selecting. The trick is to find another spot in the picture that contains suitable background, select it, copy it, and move the copy over the offending object. For example, you might find a nice piece of sky with fluffy clouds, copy it, and use the copy to obstruct the tree we've been using as an example.

The "copy" kind of fix is done by creating a copy of a piece of the image and then moving that copy so it's covering the offending object. Just doing a selection and copy doesn't actually create a copy, though. You'll need to be careful and paste the copy in (if you're using Windows, the keyboard combination "Ctrl-C" and "Ctrl-V" copies and pastes). If you forget to paste, you'll move not a copy of the area you selected but the area itself, leaving a white hole!

Obscuring junk with a copy of something good works best if the copy is about the size of the junk you want to remove. In some photo editors, you can accomplish this by first using the lasso tool to select a rough area that includes the object you want to remove, and then using a command (in Photoshop it's "Select>Transform Selection") to move the selection outline around on the image till you find a spot where it would select the right kind of background, then copying that area and using the copy to blot out the distracting object.

With any of this type of fix, you have to be sure the edges of your new material blend with the old. That's easiest if you use soft edges on the stuff you're copying. When using the stamp, softening the edge of the area that's cut out for pasting later is done by manipulating how hard the "brush" that forms the stamp is. Use a softer brush to get softer edges. With the copy-and-move method, you'll need to use a command called "Feather" after you've selected the area you want to copy. Feather lets you soften the edge over a range of pixels-pick a smallish number for most applications-between about 4 and 10. You can also use a "soft" brush for the eraser tool to blur the edges of what you're pasting, but be careful not to remove so much the bad spot shows again.

When you do a copy-and-paste kind of fix, you'll notice that the pasted material seems to float above the rest of the image. That's because Photoshop puts it on a different layer. Layers are useful tools in Photoshop, but a little beyond beginner level. The only thing you should remember for now is that when you're done covering your offending object with a copy of something else, go to the "Layers" menu, and select (toward the bottom) "Flatten Image".

Making the Color Right

It's very common for people who have taken a digital picture to think ,when viewing it on the screen or on a print, that there's something wrong with the color or exposure. You hear things like "It's too dark", or "it's washed out", or maybe it's too blue, green, purple, or yellow. Sometimes that's not just a reflection of your personal style (sometimes it is, of course). Digital images can be over- or under-exposed and can also be shifted in color. Photoshop can't fix all of these problems, but it can often help.

There are a lot of ways to fix exposure and color, some of which take quite a bit of time. There are a few shortcuts that may help you out, so it's worthwhile to try them first.

If you're using Photoshop, you go to the Image menu and select the "Adjustments". If you are using Elements, go the "Enhance" menu. When you're there, you'll see some of the following options (how many will depend on which version of Photoshop you're using): Auto Levels, Auto Contrast, Auto Color. What you want to do is to click the "Auto" choices, in order, from the top to the bottom. After you click one, see if you think the image is better. Remember how much better your choice looked, then go to "Edit" in the menu bar and select "Undo". Then move on to the next "Auto" command, until you've tried each option available to you. If none of these made you even slightly happy, you'll have to try the harder way. If you think one option is just what you're looking for, you're done-pick that one and go on to the next section. If you think one of the results is better but still not good enough, go back and redo that one, then continue in this section to do more detailed adjustment.

Let's start with exposure. A scene may look too dark or too light, but you have to be sure that the image isn't really dark or light. Look for little symptoms, like areas of light that seem completely white and without detail, or areas of dark that are featureless black. Major exposure problems are pretty much impossible to fix with Photoshop or anything else, but it's usually possible to make shots that are just a tad off look better.

The first step in exposure correction is to learn to love a Photoshop tool called "Levels". In the full version of Photoshop, this command is found in the Image>Adjustments>Levels menu. In Photoshop Elements, you can find it in Enhance>Brightness/Contrast>Levels Either way, you click the command and you'll see a control box with a bunch of buttons and what looks like a black graph. That's called a "Histogram", and what it represents is a graph of the number of pixels in your image that have an overall brightness level ranging from pure black on the left to pure white on the right. The histogram shows a bulge everywhere that there are at least some pixels, and the bigger the bulge the more pixels are there. Where the graph moves downward toward the bottom and becomes a very thin line, or winks out completely, it means that there are no pixels at all with that lightness value in the image.

Some books (and people) will tell you that a "perfect" histogram is a kind of hump square in the middle of the graph, tapering off evenly to zero at each end. I've rarely been lucky enough to see one like this, and you probably won't either. Don't worry about how a histogram should look in the ideal sense, worry about how to make yours look better. To do that, you start by looking at the top and bottom ends of the histogram. At the left, you'll see a black button/slider at the bottom of the graph, and at the right you'll see a white one. The first question you have to ask is whether there's an area to the right of the black slider or to the left of the white one that shows no thickness, meaning that there are no pixels in that range. If you have that kind of "blank" area, move each slider toward the middle until the slider us underneath an area of the graph were there are at least a few pixels shown. When you've done that, you'll see the image itself will change in lightness and contrast. What you've done is to narrow the range of pixels that Photoshop has to display (moving the two ends closer means that the space between them is less), so Photoshop can display more fine detail on the part that's left. You've also changed the value of overall brightness. If you like the image now, you can quit and close the Levels control box.

If you still think the image is too light or too dark, try one more thing. In between the black slider on the left and the white one on the right you'll see a gray slider. Twiddle this back and forth a little bit. Moving it to the right will make the image darker overall, and moving it to the left will make it lighter overall. By diddling with the middle slider you can adjust the overall lightness of the picture to a value that looks best.

When you have things as good as you can get them, take another look at the image. If you're happy, move on to the next section. If not, there's one more thing you can try.

Go to the Image>Adjustments>Hue/Saturation menu item. For the moment, ignore the Hue slider on top. Try moving the saturation slider a little to the right to make the colors stand out more. You can also move the Lightness slider just a little either way to balance the brightness of the image. If this doesn't work, your problems are beyond a beginner tutorial. Buy a book on Photoshop color!

Improving Sharpness

Before I start, let's get one thing straight. If you shoot a picture that's out of focus to any significant degree or that has a motion blur because you shook the camera or somebody moved, it's pretty much beyond hope. Fixing sharpness means bringing out detail better, not inventing it!

There are three basic ways to sharpen up pictures. One is using an oddly named Photoshop filter called "Unsharp Mask", and the other is using an equally weird-named filter called "High Pass". Only the last, "Sharpen Edges" seems logical. You'll want to experiment here, so remember to use the Edit>Undo to go back to the original state unless you've gotten the effect you want.

It's best to start off with unsharp mask, which is usually abbreviated as USM. This filter is good for increasing the detail/contrast that tends to get a bit washed out in digital photography. A little of this goes a long way. Go to Filters>Sharpen>Unsharp Mask, and Photoshop will open a dialog box. There you'll find three settings: Amount, Radius, and Threshold. Start with an amount of 70%, a radius of 1.3 to 1.5 (you can fiddle) and a Threshold of 0 to .2. See if you like the effect. If not, you may want to try increasing the amount (to as much as about 150%) or reducing it if you got too sharp. When you've got the best effect here, you can move to the next idea, but try to avoid going beyond an amount of about 70% if you're going to use other techniques along with this.

Sometimes you want more edge contrast without making the whole image look…well…hashy. For that, try Filter>Sharpen>Sharpen Edges. This will make the edges of things (distinct differences in color/brightness) harder. There's no adjustment for it, so give it a try. If you now like the image, you're done.

If you want really strong sharpening for something, you've probably got to move on to "High Pass" and this is complicated, so follow these steps carefully:

On the right side of your Photoshop/Elements screen, you should see a "Layers" pallet docked. If you don't see a layer saying "background", click on the Layers tab till you do.
At the bottom of that pallet is a series of icons. The rightmost one is a trash can, and next to it is a weird little double-box. If you put your cursor over it, you'll see that it's the "create a new layer" button. Don't push it! Just remember where it is.
Click on the background layer and drag it to that button. This should create a layer called "background copy". If you get an empty "Layer 1" you pushed the button instead. Right-click on Layer 1 and delete it, then redo this step.
Make sure that your background copy layer shows the "eye" and the "brush" to the left of the image, meaning that it's selected for action. If not, click on it.
Go to the menu bar and select Filter>Other>High Pass. Wait till the high-pass window opens.
Select a Radius value of 1 to start. Your image will be essentially a gray blob with little detail showing. You should just be able to make out the edges of objects, looking like they're embossed. If you see a lot of detail, decrease the radius by .2 or so. If you don't see anything at all, increase it a little, but don't go above 1.5.
Push the OK button. Don't panic here because your image looks lost!
Go back to the Layers pallet, and you'll see under the tab a pull-down menu that's preset to "Normal". Click on the arrow and you'll see a list of choices. These are the layer blending options. Part-way down, you'll see a choice "Overlay". Select it. Your image will clear up, and it should also look sharper.
If you want to increase this effect, drag the background copy layer to the "create a new layer" button one or more times, and you'll create additional layers that intensify the effect. Be careful here; a little goes a long way in sharpening.

If you like something in between say two and three copies, adjust the opacity of the top one or two down a little from 100% to reduce the effect.

When you're done and you're sure you like the result, go to the menu bar and select Layers>Flatten Image. You're then down to one layer, and you're done with the fixing and tweaking part.

Saving Your Image

If you've put a lot of work into the steps above and you have a couple of things you'll want to do with the picture, it's time to save your work. You want to be sure to save it in a "lossless" form, without a compression technique that will give you an approximation of the image only. To do that, you'll either want to save in the native format for your editor (my recommendation; for Photoshop it's the PSD extension) or in the universal TIF extension. When you save the image be sure to add something to the filename telling you this is a fixed image. For example, if your original file was "C12_0034", you might want to save as "C12_0034_Fix021504.PSD" to tell yourself that this is the version of the image you fixed on Feb 15th 2004.

Don't save images that you're not going to reuse. You can go back to the "digital negative" you saved earlier if you need to, and your skills and tools will certainly improve over time. I don't save anything in "fixed" form any more unless I did a heck of a lot of fixing.

Preparing Your Image for Printing, Email, etc.

Most digital cameras these days are at least two megapixels. The average size of an image on a computer screen is about a half megapixel. Even if you wanted to do nothing with your pictures but look at them on your own computer, it wouldn't make sense to store them in the "edit" or "negative" form for that purpose, and it would take longer to display them. If you want to use an image, that probably starts with "re-purposing" it for its intended use, and that probably starts with resizing it.

If you open the Image menu item, you'll see (in Photoshop) the command "Image Size", or (in Elements) Resize>Image Size. Either will give you this complicated window divided into two sections. The top section gives the image dimensions in pixels and the bottom the document size and resolution. There are also some check-boxes at the bottom. Before we get messin' with them, let's go over what we plan to do.

Each image use will have an ideal "size", and what you need to do at this point is to get your image to that size. Here are some basic rules (to which, as you get comfortable, you're likely to find a bunch of exceptions):

Images you plan to send as attachments in emails or put on a personal web page should be sized to between 600 and 800 pixels in the largest dimension.
Images you plan to view on your computer in a slide show should be sized so that the largest dimension of the picture is equal to the resolution of your monitor in that dimension. That means that if you have an image that's 2000x1000 pixels and your monitor is 1024x768, you'll want to change the width to 1024. If your image is in portrait mode (higher than wide) you'd size the vertical dimension to 768.
Images you plan to print should be sized to between 150 and 300 pixels for each inch of print height/width. Use the longest dimension only in your calculations; a 4x6 print would mean 900 pixels in width, for example. Use the lower pixel-per-inch number for snapshot-like prints and the larger one if you want quality. Be wary if you end up with a number that's greater than the current value of width or height-you're trying to make a print larger than your image resolution will support and you may be disappointed in the results.

Whatever you decide for your image size, enter it in the appropriate box. You need to size only one dimension under these rules, because you want to be sure that the little chain-link button to the right of the width and height is set to lock the two in ratio. Set one value and the other one changes in proportion. If you accidentally uncheck this, by clearing the check in the "Constrain Proportions" box, you'll get a really distorted image.

If you're sizing for print, you want to enter the number of pixels per inch that you selected in step 3 above into the "Resolution" box in the Canvas area. The physical dimensions of the image will now change to reflect that resolution, and you'll see how big your print will be. See below if you are trying to get a particular size.

The other little box, "resample image" isn't important for this basic tutorial. Leave it checked with "Bicubic" as the selection for normal resizing, and go ahead and press the OK button.

When you resize your image with this little window, you DO NOT want to save it back under the same file name! You'll eradicate your original with a low-resolution copy. So before you go any further, you'll need to save this resized version, so you go to File>Save As and not just "Save". For your filename, use the "negative" name (C12_0034 in our example) with a suffix of something like "_600x400" to tell you what you did to the picture. But don't do it yet because you've got to decide what file format to use.

There are two basic ways to save a "for-use" image like this. First and most often used is as a compressed file with the JPG extension. The second way is in a lossless format, using either your editor's native format (PSD for Photoshop) or TIF. You'll probably want to use the latter only for stuff you're going to print at a large size.

When you save in JPG format using Photoshop or most other editors, you'll get a window that asks you to set the JPG compression options. For now, leave the "Format Options" at "Baseline". Look at the Image Options area of the window and you'll see a Quality level with a text decoding of what that number means, and also a slider. This is where you either make friends or enemies in creating email pictures, so pay attention; we'll assume that your use is either to email the image or to put it on a personal web page. Start with a number of 5 in the Quality box and see what it tells you for Size at the bottom of the window. You're looking for something no more than 60 KB, so if your file is bigger than that move the slider a tad to the left, and if it's smaller a tad to the right. When you've gotten it at about 60 KB, push "OK".

Other quality settings are best for the other choices. If you're compressing your file for viewing on your own PC, you can select a 9 to 12 (the maximum) quality if you have a good monitor and graphics card, and a lot of space to store your images. Otherwise set it at about 7. If you want to print the image on your own printer, my recommendation would be to set the quality level at 9 or higher, or save in a non-lossy format. If you're taking or sending the image to a store or lab for printing, ask them what format to use.

The "Wrong Sized Print" Problem

A typical digital camera image has about a 4:3 ratio between width and height. A 4x6 print has a 3:2 ratio, and an 8x10 has a 5:4 ratio. The moral is that if you simply "print" a digital image you're likely to end up with something that's not a standard size and won't fit in an album or frame. So what do you do? It depends on how complicated you want to get. There are a lot of ways to print just a part of the image, which is called "cropping". Some can get you into trouble in print quality. I'll give you an explanation of a pretty fail-safe way with some limitations; if you want more than this, consult our tip on resolution.

Obviously, the first step is deciding how big your print will be and whether it's displayed in landscape or portrait format. For the easy method, we'll assume that your image is shot in the same mode (landscape/portrait) as it will be displayed. Now, divide your minimum print dimension (which would be 4 inches in a 4x6, for example) into your maximum. Save that number. Do the same with your image size (lets use 2000x1500 as our example). In these cases, you'd get a ratio of 1.5 for the print and 1.33 for the image. If you were doing an 8x10, you'd have a ratio of 1.25 for the print. Write both numbers down!

OK, here's the complicated part of this simple strategy. If the ratio for your camera image is greater than the ratio of the print image (as it would be if we wanted an 8x10 in our example), that means that your camera image is more oblong than the print you want, and you'll have to lose some of the image in the long dimension. If the ratio for the camera image is less than that of the print (our 4x6 example), the image is "fatter" than the print and you'll lose some in the short dimension when you print. That means that when you resize your image for printing, you'll want to size images the way we specified above (using the long dimension as the guide) for images that are "fatter" than the print, and based on the short dimension when images are "longer" than the print. When you run through the resizing as described above (using the right dimension as your guide), you'll end up with an image whose "Canvas size" is equal to your print size in one dimension and a bit larger than your print size in the other. Got it?

OK, now for the next step. Go to the File>New menu and create a new document. You want a transparent background, and a canvas size that's equal to your desired print size. Set the resolution to the same value for pixels per inch that you used in your sizing calculation. You'll end up with a little window with what looks like a checkerboard in it. Minimize this for now to get it out of the way, and go back to your resized image. Click on the title bar of the window to get the image selected and then go to Edit>Select All (or just press Ctl-A). Then do an Edit>Copy (or Ctl-C). This will give you a copy of your image. Now, go back to the new "checkerboard" window (if you minimized it, click on the second-from-right button to get it back) and either select Edit>Paste or Ctl-V. Your image will now be pasted onto that new window as a new layer. But here's the good part. If you click on the "Arrow" tool in the toolbox, you can now slide the new layer around on the canvas! That means that you can pick what part of the image is going to get cut off. When you have the composition the way you like it, flatten the image (Layers>Flatten Image) and go ahead and print it.

Don't Forget!

There are no photo-tips that can substitute for learning your own photo-editor's features. As I said when I opened this tip, you have absolutely positively got to make a commitment to learn to use these products that's proportional to what you expect to get from them. Buy a book. Play with your pictures and try different techniques. Haunt the Internet for other sources of information. That's why its important to get a product that has a lot of diverse information written about it. Buy something weird and you're on your own.

The biggest other thing to remember is be consistent. Digital workflow is a process that you have to design yourself to protect yourself and your images from mistakes. Save your digital negatives. Remember to undo something you don't like. Make your file names logical. You are going to make some truly astonishing mistakes with your editing. If you're smart, you'll recover, learn from them, and get better. If you're not smart and systematic, you'll get discouraged. Your choice!