Digitizing Images for Multimedia Using Photoshop
Collected here you'll find a number of steps that you can use to digitize images that come from flatbed scanners, or digital cameras. These are guidelines and will not be exactly right for every situation, but they are a good place to start for developing high quality images in many situations.
Please note that ".psd" means that the file is saved out as a "PhotoShop Document", or the standard Photoshop file format.
To the right you will see a little cue card that you can use to summarize the steps, and below you can find the more detailed descriptions of each step.
Please note that these steps work equally well with older versions of Photoshop all the way back to version 2.5. Many of these steps are generic enough to be used with CorelDraw, CorelPaint, Painter, and other image-manipulation tools.
Steps for Digitizing Images
- Scan your image or download it as necessary
Make sure your scan is large enough to cover all the possible uses that you might need it for. Just because you plan on using for a web site, you might later want to use it in a printed marketing piece. Got resolution?
- Save filename_raw.psd and save as filename_color.psd
It's important to save your files at each major step that you take, if you have the room on your drive. Depending on the amount of time each step takes, you can save yourself hours of re-work. It can also be important to save the first raw scan in case the original becomes unavailable. After you've saved the _raw version, Save As... the file as the _color version so you're working on the right file.
- Crop and/or Resize
More often than not, I find that the guess that I make when I scan in an image is way off as far as scaling. I usually bring the image a little closer to the final output. Not all the way down, just within the ballpark.
- Levels or Curves
This is the first real step in preping an image. Here we set up the tonal range of the image. Usually this compensates for scanners or cameras that don't actually make it all the way to true black or true white.
For this step, the only thing we're really interested in is the Saturation slider, which will allow us to bring up the strength of the colors. Poor lighting can often cause images to loose their tonal qualities. This is the first step to rectifying that.
- Color Balance
This is the second step in clearing up colors. The key here is to make small changes, and after you hit the 'Okay' button, use your Command/Ctrl-Z to switch back and forth between the change you just made. Did you improve the image or make it worse? Often times, the first try will only make it worse. As you use the Color Balance command more often you will become more skilled at using it.
- Unsharp Mask
Using the Unsharp Mask is the better way of working out some of the artifacts and 'fuzziness' that comes about in images from the scanning process and from resizing. Every time you resize the image from this point on, you should consider using an Unsharp Mask filter on it. The secret to using the Unsharp Mask, as with many of the commands in Photoshop, is to understand that a little goes a long way. Watch the edges of areas of color, and lines in your images to see if you start getting strange artifacts from over-sharpening the image.
- Save filename_color.psd and Save As filename_retouch.psd
Now that you've spent all this time color balancing and getting your image ready for the retouching and manipulation stage, it's a good idea to save this version, in case something happens and you want to get back to using the image just like it was when it was first cleaned up. Once you've saved the _color version, save your file as the _retouch version to begin working on the Retouching stage.
- Retouch, manipulate and perform your artistic magic.
This is the point where you actually get to do all the cool things... rotating the image, giving a color cast, or running cool filters on it, moving people's heads, giving them extra arms, etc. Go wild, but be sure to save out different versions as you're working along.
- Save filename_retouch.psd and save as filename_final.psd
Now that your image has reached photo nirvana, go ahead and save out the image in it's fully layered, fully filtered large size. This will probably be the main file that you refer to if you need to make a type change or put in this year's favorite color from the Spiegel catalog.
- Resize and/or Crop
Now get the piece into position. Resize it and crop it so that it's ready for it's proper placement. This is only the second time where you are thinking about your file method of outputting the image. Up till this point, you were only supposed to focus on creating the best, highest-quality image you could using the machine you've got.
- (If you resized, use Unsharp Mask again.)
Remember, each time you resize, you should consider running an Unsharp Mask. If you've reduced your image to any less than 80% of your previous version, I highly recommend running the Unsharp Mask.
- Save filename_final.psd
Just to be sure, save out one last version with your layers.
- Export image as the final output format (JPEG, GIF, etc.)
Do the final export to JPEG, GIF or whatever format and you're on your way. Photoshop's Save For Web command is great for this.
Make good note of that Unsharp If Resized Below 80% rule of thumb. It's really made a difference with the image quality in my projects.
Make ample use of the Undo/Redo command to switch back and forth between each step that you take. After each step, hit Command-Z or Ctrl-Z a few times in row. Rapidly switching between the view just before and just after each step will help you accurately determine if the action or filter that you just used actually improved the image or not.
And there you go. These steps have been developed from my personal career, but they rely heavily on various pieces of training and tutoring that I received from Creative Multimedia, in the mid 90's. I've updated some of the terms, but these steps are largely the same as the ones we used while developing CD-ROMs. The main goals are A) to keep revisions of the images, B) keep the amount of information (pixels, bits, etc.) high enough to make a good image, and C) only then optimize the images as you come to the end of the process.
Comments and questions about this article are encouraged. Contact me through http://www.rossolson.com/.