ESPARSHOT - Glossary

Noise Reduction

For the past 4 years I have spent hundreds of hours researching methods to reduce noise from digital camera images. The key to noise reduction is to reduce or eliminate the noise without deteriorating other aspects of the image. Many freeware and even paid solutions negatively affect image sharpness, introduce wavy patterns in uniform surfaces and/or make them look "too uniform" (a bit like in a water painting).

These crops below out of a prosumer image illustrates the problem of edge sharpness and wavy patterns typical to a lot of noise reduction methods and compares the results with methods described in my e-book. The results are shown both for the color image and in the red channel. The areas indicated by the red squares are 4 times enlarged in the row below. On some monitors, the noise may not be very visible in the original. In that case, look at the red channel crops instead.

Original	Bad Noise Reduction	Good Noise Reduction (123di)
Original crops (1X) and enlarged red squares (4X) below


Notice the red color noise in the blue sky of the original, more visible in the red channel (*).	Bad noise reduction methods remove noise but blur the edge as shown in the 4X crop.	Good noise reduction methods remove noise but preserve edge sharpness.
(*) This type of noise is also more visible in the ISO 800 example in the sensitivity topic.
Original	Bad Noise Reduction	Good Noise Reduction (123di)
Red channel - original crops (1X) and enlarged red squares (4X) below


The noise in the blue sky of the original is very visible in the red channel.	Bad noise reduction methods replace the noise by a wavy pattern in uniform surfaces, visible in the 1X crop.	Good noise reduction methods do not introduce wavy patterns but at the same time preserve some natural "grain" and image sharpness.

JPEG Compression and Noise Reduction

JPEG compression squares are normally hard to notice in uniform surfaces at high quality levels. Since noise introduces (unwanted) detail, the JPEG squares will become more visible which further deteriorates the image. Working in RAW overcomes this problem. However, as stated in my Photoshop CS review and on my personal website, the appearance of noise can vary depending on which software you used to open the image.

Long Exposure ("stuck pixel") Noise Reduction


Original image	Dark frame	Manually cleaned image

The effect of long exposure stuck pixels can be reduced to a great extent by taking a "dark frame" (with lens cap on) either before or after the main shot and subtracting this from the original shot, as explained in my e-book. Many newer digital cameras have built-in long exposure noise reduction and take a "dark frame" with the shutter closed for the same amount of time as the main image. This dark frame is then used to identify and subtract the "stuck pixels". But even with noise reduction OFF, newer cameras will show fewer stuck pixels than in the above example which was taken with an older generation digital camera.