The Southampton Press, August 23, 2007

Picture This

By Cheryl Machat Dorskind

Reducing Noise for Clearer Photos

Each pixel in a camera sensor converts light into electrical signals. These signals vary over time, color temperature, and film speed (ISO). These variations are called noise. Luminance noise, visible artifacts like the grain of a film, is most visible in uniform surface areas such as a sky or a wall. Color wavelengths (RGB) create color noise which is especially apparent with compact point and shoot cameras. Noise is typically more visible in the red and blue channels, than in the green channel.

Noise increases with higher color temperature and ISO. Color temperature is based on the Kelvin Scale (K). Candle light is rated at 1900 K, a camera’s flash is 5500K, and mid day blue sky is rated at 9000 K. Noise also becomes apparent with long time exposures (“stuck noise”), such as night photography, and when scanning images especially negatives.

Moonrise-Shore Road,Westhampton Beach. The foliage creates an S-curve—a design element that enhances a composition. This photo also utilizes the “rule of thirds”—the trees are on the right and left vertical thirds and the moon lies within the top horizontal third.

Understanding noise is important when purchasing a camera. While camera manufacturers are selling the bigger mega pixel package, what you should really be concerned with is the size of the sensor, which is often compared to the size of 35 mm film.

All pixels are not created equally. The size of the sensor creates the size of the pixels. Likewise, a small sensor will create more noise than a larger size sensor. While pro cameras like the Canon 1D Mark III (28.1mmx18.7mm) or Nikon 2Xs (23.7mm x 15.7mm) have sensors close in size to a 35mm piece of film (36 x24mm) and virtually no noise, a Canon compact model, such as the Powershot SD900 (10 megapixel camera), sensor measures only 7.176 x 5.319 mm.

There are four general rules to follow to reduce noise and optimize detail with your digital camera.

If possible, turn off the “in-camera” noise reduction.
Turn off the “in-camera” sharpening – sharpen instead in the photo editing stage.
Use the camera’s flash in low light conditions when possible.
Use normal contrast in your camera settings.

In the photo editing stage, to combat and reduce noise, I use “Noise Ninja” by Picture Code (www.picturecode.com) or Define2 by Nik Software (www.niksoftware.com). Both programs can be loaded as plug-ins to Adobe Photoshop Elements 5, Adobe Photoshop CS3, or Corel Paint Shop Pro XI or as stand alones. Both software programs enable global corrections and selective noise reduction using a digital paint brush. It’s best to reduce noise early in the photo editing workflow.

The screen seize was created with “Snag It” by TechSmith. Visit their website at www.technsmith.com for a 30 day free trial.

Define 2.0 and Noise Ninja are easy to use noise reducing software. The auto profiling works well, or for digital imaging savvy, there are plenty of advanced noise reduction tasks. After opening my picture into Photoshop, I immediately reduced the noise on a separate layer using Define 2.0. Working with Define 2.0 by Nik Software, I manually determined the noise profile of this picture using the task panel on the right of the screen. Notice the amount of noise on the left side of the split screen (before noise reduction). The right portion is a preview of how the image will look once I click “OK.” After this step I will adjust the tones and colors running through a series of five basic editing steps.

Don’t let noise stand in the way of your otherwise perfect picture.

contact: Cheryl@cherylmachatdorskind.com.