Sharpening
Are your pictures sharp as a tack? Get this important aspect of digital photography right with our essential advice
Virtually all digital cameras have a piece of glass in front of their sensor called an anti-aliasing filter. Its aim is too reduce artefacts (sometimes called ‘jaggies') that would other wise be present in the picture, but unfortunately it does this by blurring them very slightly. This means that straight out of the camera a digital image is slightly soft and will need sharpening.
This shouldn't be news to you, but if it is and you are worried you've been doing something wrong, then panic not - you're ok. If you've been shooting in JPEG mode then your DSLR will almost certainly have been sharpening pictures for you by applying a mathematical algorithm to the entire scene, thereby bringing back some ‘bite'. If you are a Raw shooter on the other hand, you will have been used to seeing slightly soft images already, as Raw images are free from this kind of processing.
Conventional wisdom has it that sharpening is best left until the end of your digital workflow, because different methods of output (inkjet print, magazine repro, online use) require different levels of sharpening, but more and more industry practitioners are coming to terms with a concept called pre-sharpening: an initial sharpen carried out early on in the workflow. This pre-sharpen should not be confused with the final sharpening stages you may already be employing - it is far less severe - and as long as no artefacts are introduced, then further editing should not be a problem.
Until the arrival of Photoshop CS2, sharpening was always carried out by the Unsharp Mask filter, or ‘old reliable' as many photographers know it. Nowadays, though, this is old technology and the Smart Sharpen filter is what is needed if artefacts are to be avoided at the pre-sharpening stage. Smart Sharpen performs well for two reasons. Firstly it uses the Lens Blur filter to determine what should be sharpened and what shouldn't; Unsharp Mask uses Gaussian Blur. The main advantage of this is that Smart Sharpen is better at finding the parts of an image that need sharpening (edges) and leaving alone those areas that don't (expanses of blue sky, for example). Secondly, in More Accurate mode, Smart Sharpen can recognise and reduce halos brought on by the sharpening process. Users can specify how broad a range of tones should be considered as halos and how much they should be reduced by.
Because we don't want to go over the top sharpening at this stage, it's worth experimenting with some settings to find a point where the bite just begins to come back a little, and stop there. Smart Sharpen's interface even allows for these settings to be saved as a preset, so they can be called up and employed time and time again.
BEFORE
