Noise and sharpening
Add noise with the Add Noise filter
Sometimes noise is a good thing. Not the “head next to a buzz saw” type of noise, but the colorful din of a street fair or the controlled commotion of a baseball game. Using noise in Photoshop can also produce such organic results, especially if you use noise to add realism and randomness to an image. If you aren't already using noise in Photoshop, you should be. If you are, keep reading to learn about what else you can accomplish with noise.
What is noise?
Noise, as defined by Adobe, adds “pixels with randomly distributed color levels.” Simply put, noise imposes a random value to the pixels that make up an image. However, the randomness of noise doesn't have to be a distraction. In fact, using noise in images often can help hide problem areas such as banding and awkward cloning. The noise filters in Photoshop, located in the Filter menu, help to reduce noise and eliminate the effects of overly noisy images. While these filters, which include Despeckle, Dust & Scratches, and Median, have positive attributes of their own, they exist to remove noise. We'll concentrate on the fun of introducing noise to your image using the Add Noise filter.
The Add Noise filter comes alive
The Add Noise filter was originally created to simulate the grainy texture of high-speed film. Since its introduction, it has been a staple of serious image jockeys because it can do so much more. The Add Noise dialog box, shown in Figure A, consists of the standard preview window that helps gauge the impact of your noise options. Fine-tuning the noise values is an art in itself, so use the preview as much as possible.
The Amount slider and text box in the Add Noise dialog box work on an 8-bit scale, which means it uses 256 shades of gray between black and white. Basically, the Amount slider determines how far the pixels in your image can drift from their original color. This technical mumbo-jumbo is important, but the most important thing to remember about the Amount slider is that higher numbers mean higher contrast between your pixels, and therefore, more noise.
Types of noise
There are two types of noise you can apply. The first, Uniform, creates completely random noise. This means that any color is just as likely as another to be used to produce the effect. The second option, Gaussian, creates noise along the Gaussian curve, which results in a more consistent allocation of noise values. Since colors are set to a specific range, the colors of Gaussian noise more closely resemble the colors of your image. Gaussian noise often creates a more realistic effect, while Uniform noise delivers a more distinct result. Figure B shows a sample of each of these noise options applied to a solid block of 50% gray.
The Monochromatic check box at the bottom of this dialog box controls the look of both Uniform and Gaussian noise. Selecting this option allows you to apply grayscale noise to your image. Monochromatic creates an even distribution of noise across the channel without changing the colors. If this check box isn't selected, the pixels and colors are scattered randomly throughout all the available channels. This option is a good idea when you're working on images with areas of solid color, but a bad choice if you're trying to blend contrasting areas of color together. Again, use the preview window to see if Monochromatic noise makes sense for your imaging needs.
There are many reasons to apply noise in your everyday image editing. You'll find that noise can help correct many problems caused by poor scans, blurry digital camera images, or strange rasterization. In any case, we'll take a look at one of the most common scenarios for using noise. From there, you can build upon your repertoire and turn the noise into a symphony.
As mentioned earlier, the Add Noise filter was designed to simulate film grain in a pixel-based environment. This is still one of the most useful features of this filter, and you can use it in many ways to improve the look of an image. You can add noise to blurry images to increase contrast, to black and white images to heighten the visual mood, to vector art to increase texture, or to retouched images to blend in edited areas. Our example, shown in Figure C, is missing something. To add a nice “old photo” grain, we added some noise to this grayscale image to give the image some texture and personality.
An imaging cacophony
There are many other uses for noise in Photoshop. For the time being, experiment with adding noise to your designs. You'll soon find noise all over your images, and that's a good thing. Technically, noise filters are considered destructive filters because they're used to degrade the quality of an image. But this isn't necessarily the case. Noise has corrective values as well, when used properly and subtly.