info@centerfordigitalarts.com | Gift Certificates

Single Blog Title

This is a single blog caption
noise reduction example
19 Nov 2014

Noise Reduction

Posted By

Noise Reduction

What Is It, How to Remove It, and Should I?

Students in our Adobe and Art Photography classes often experience a big-time wakeup call when it comes to image noise while printing images at some size, say 13X19 or larger. Suddenly they see noise, white speckles, or sometimes color pixel lights, which we call artifacts, and can occur with JPEG images. Today we want to take a closer look at this issue and how noise reduction will help all of your work.

Image noise is not a simple subject, and it is a built in issue for digital camera sensors called Couple Charged Devices (CCD). A CCD employs many tiny pixels to measure the light and form a matrix of pixels to depict the image but none of the sensor devices are noise free and they can contribute more noise to an image when providing greater electronic amplification such as high ISO settings, long exposures, or both.

However, most lighting situations provide more light then the camera’s inherent CCD. On the other hand, dark scenes demand longer exposures or higher ISO settings and these generate a wider signal to noise ratio. In these cases, the noise data may exceed the light data, which creates random white dots, which is luminance noise, or color dots which is color noise.

Another noise generator occurs when you apply software to open the shadows or the blacks in an effort to recover lost detail. The software does not make more noise; rather it reveals noise previously hidden in the lost details.

When printing images at larger dimensions noise can become quite apparent and appear as white dots scattered randomly on the photo. More often than not, we will see noise in dark landscape skies or in the shadows of portraits.

One thing for certain is that todays’ cameras generate far less noise than their predecessors, even as recently as 5 years ago and camera noise issues continually improve which means photographers are seeing less and less noise in their images even at ISO numbers of 10,000 plus.

You can reduce noise with your camera’s built in noise reduction programs or in post-production with software application built into Lightroom, Photoshop, or with a third party program.

All of these approaches can produce negative side effects because they work on a principle that involves averaging out pixel values to remove the “random” pixel values that represent noise. However, when we average out color values to reduce the appearance of color noise, overall saturation is reduced and colored halos may appear in some areas of the photo. With luminance noise you can easily experience a significant loss of sharpness and detail

As with all things in photography, there are tradeoffs, or a bit of ying and yang. The Lightroom noise filter is a powerful tool for noise reduction and provides several controls to help minimize the side effect of noise reduction. You should always check your on screen images at 100 % magnification when examining the image for noise and do all noise corrections at this magnification.

In the example photo, one can see significant noise, because this image was shot at ISO 10,000 and then we added more exposure and shadow detail in Lightroom, which reveals significant noise at 100% magnification. If we navigate to the detail panel and then choose the noise reduction controls, we can do very good job of cleaning up the noise.

Moving the noise slider to 50 eliminates the luminance noise, but it has softened the hair too much. This is where the detail slider can help. By moving that to the right, we are able to recover some of the detail lost to the noise contraction and then by adding contrast we can recover more. Adjusting the sliders a few times in order of structure will allow us to arrive at a final image that is much cleaner and we have essentially salvaged a very underexposed, noisy image.

Noise reduction technology at the in-camera level as well as at the post-processing step has improved significantly and will continue to improve. Further camera technology has also reduced the amount of noise generated by the systems, even at high ISO’s, so noise is less of a problem with each new technology iteration. However, it still occurs and Lightoom offers a great place to adjust the impact.