There is real benefit to systematizing your approach to digital photo production. Performing image editing tasks in a repeatable way can save you time and frustration. This workflow tutorial is not intended to provide how-to detail for individual steps, but rather, an overview of the whole process. See individual tutorials for details on workflow tasks.
This workflow applies to photographic images, and assumes that you are working with Photoshop 6 or later. Other image editing programs and earlier versions of Photoshop have similar functions to those mentioned here. The biggest difference will probably be the ability to work in high-bit mode for many more corrective steps than in alternative image editing packages.
Consider the benefits of setting up a project directory for each new project. Within that directory you could include additional subdirectories to hold your original image, various stages of your editing, and final flattened images ready for output to one or more devices. There is great benefit in saving an untouched original file; doing this provides you with the benefit of starting fresh if all else fails. There is also much to recommend saving a file after various critical editing points, particularly those that were time consuming, such as extensive spotting and retouching, or technically difficult.
Your workflow begins with acquiring an image. The image might be a scanned image from a flatbed scanner, desktop film scanner, or a drum scan from a service bureau. Try to acquire an image with as much resolution and bit depth as possible. You can always reduce resolution and bit depth if necessary to accommodate your computer's memory size or processor speed, but you can never really add back resolution or bit depth.
Your image might be from a digital camera, ranging from an inexpensive consumer model to a professional scanning back. Whatever the source, once again opt for the highest resolution and maximum bit depth. Prosumer and professional digital cameras can provide files at higher bit depths and you should use them. Lower bit depth and resolution can be fine for images destined for the web, but you retain more flexibility if your original image is of the highest quality possible.
Make sure that you have saved a copy of your original undisturbed image, then correct the black and white points of the image and correct any overall color casts. If you are working with high-bit images, once you have achieved the results you want, use the Save function in the Curves or Levels dialog to save your settings. With this technique, you can always return to your original uncorrected image and apply the same corrections by loading the correction file into the appropriate adjustment dialog.
If you are working with an 8 bit file, make your changes in adjustment layers. Adjustment layers, which are not available in high-bit mode, allow you to return to your corrections as often as you want to adjust them or even turn them off. Changes made with multiple adjustment layers in your image are applied as one adjustment when you flatten your image for output. They do not degrade your image as much as making the same changes one at a time not using adjustment layers.
At this stage, perform your main image spotting and retouching. The benefit of doing the bulk of your retouching work at this point is having an image with what can be some of the most time consuming work completed before setting image size and resolution for output. If you wait until later in the workflow to do your touchup, you will have to repeat your retouching every time you create a new image size for output.
I generally leave my file in high-bit mode during this phase since many retouching tools are now available in this mode in Photoshop 6, but this is not absolutely necessary. You may need to reduce your file to 8 bits per channel to access more sophisticated touchup or image repair tools. Also, reducing a file to 8 bits per color channel can substantially reduce the size of the file if your system does not work well with larger files.
Before sharpening and fine tuning your corrections, set your image at its final size and resolution. If your image was acquired through a drum scan at a service bureau, it may already be at the correct output size and resolution. On the other hand, a desktop scan or digital camera image will probably have to be resized. You may simply do that in Photoshop, or use a third party utility like Genuine Fractals Print Pro if you are making a substantial resize of your image. As a general rule, wait to do your sharpening until you have a file set at your output image size and resolution.
You may need more than one image at this stage if you are going to multiple output devices, say both the web and photo quality output to an inkjet. Your best bet is to create separate saved files for each output requirement, since each will need a different resolution to maximize quality. You will need to proceed through the rest of the workflow for each file.
Once image size and resolution are set, sharpen the image using your preferred method of sharpening. The Unsharp Mask filter is perfectly suitable for many images. There are a number of other sharpening techniques for special situations.
If your image is from a digital camera, and particularly one using a high ISO setting, chances are you are going to have some digital noise to deal with. If your image has sharp demarcations between elements, you may have a color fringe that needs to be removed.
After sharpening, you will sometimes notice areas that need spotting or retouching that you missed during the first image touchup. This is the time to fine tune your retouching.
If you have been working in high-bit mode, now is the time to change to 8 bit mode. This requirement may change over time as output devices that work with high bit files are starting to appear. Now is also the time to change to the appropriate output color mode.
If you are working with a color image and have a color profile for your printer-ink-paper combination, now is the time to soft proof your image. If you are printing to matte or watercolor paper, make sure to check Simulate Paper White in the Proof Setup dialog box. If you are working in grayscale, you can achieve the equivalent of a soft proof by using a custom dot gain curve.
At this point, you can use a Curves Adjustment Layer to compensate for any subtle changes noted in the soft proof.
If you have been working in multiple layers, flatten your image before saving a version for output. A large file made up of multiple layers can choke some output devices. If your output is going to the web, use Photoshop's File/Save For Web... dialog to prepare your file to be used on the World Wide Web.
Copyright 2002 Michael W. Rollins