My lessons learnt from this exercise:
- It’s interesting to observe the effect that the change of ISO settings causes on histograms of the images.
- Similarly to highlight clipping, the three RGB channels do not seem to be affected by the ISO changes in the same way so it is sometimes helpful to differentiate noise reduction per channel.
- It appears that the degree of noise might differ from one camera to the other. Whilst with the film cameras, the level of noise usually originates from the grain in the film emulsion (hence, the various ISO settings), it seems that digital noise depends on quality of a camera sensor.
- Although highlight and shadow clipping usually represents a problem, it’s also possible to use it creatively to achieve special and unusual effects.
- Noise might appear when different ISO setting are used or it might be related to exposure, for example when an image is underexposed.
- Noise may cause colour cast, loss of detail, apparent loss of sharpness, appearance of colour specks.
- Shadow clipping warning is a very useful tool to be used on a location and when processing images on a computer.
- Noise reduction can be carried out at different stages of the image processing. For example, some noise reduction usually takes place when an image is captured. Most cameras have facilities that allow some noise reduction. It’s important to understand how these facilities work and the effect they have and the best way to understand them is to practice using different settings and lighting conditions.
- It might be possible to recover some detail in shadow and reduce noise by using: 1) Sharpening tool; 2) Fill Light and Blacks tools;3) Noise reduction tool. Noise reduction can be applied to individual channels for best results.
- There are noise reduction tools in the Camera Raw, Photoshop and Lightroom programmes. Camera Raw, for example, offers the luminance and colour noise reduction. Photoshop also has a Noise reduction filter. Colour noise option deals with the problem of ‘colour specs’.
- I tried to apply different noise reduction options to the image used in this exercise. I found that although it is possible to reduce noise successfully, too much noise reduction may actually cause colour cast, loss of quality and sharpness and colour bleed. When the level of noise is high (for example at the higher end of the ISO setting), noise reduction might cause more damage to the image.
- I found a lot of useful information in Mark Evening’s manuals. For example, Mark advises that the blue channel is ‘the noisiest’ and that too much colour noise reduction might soften ‘the blue edge detail’. I’ve also learnt from Mark Evening’s manual that it is helpful to use greyscale sharpening previews as they show the effect that the sharpening tool has on the image.
- Using image magnification (1:1 or higher) is helpful as it allows to assess the levels of noise in the image.
