Sharpening

Image sharpening is now widely available in most digital processing software packages, and is often used to draw viewer attention to certain areas within the frame, by making them stand out or ‘pop’; sharpening can also be used to emphasise texture and detail.  Moreover, this process can be used to add a certain amount of sharpness to our images, however, over use of this filter can produce unsightly artefacts, so a certain amount of trepidation and experimentation is required during its application.  Cambridge in Colour states:

Image sharpening is a powerful tool for emphasising texture and drawing viewer focus.  It’s also required of any digital photo at some point – whether you’re aware its been applied or not.

The reason for adding sharpening to our post-processing workflow is because a digital camera’s sensor and the lenses we use will always add a degree blur, hence the reason why we need to understand the best way to address the issues that arise.

Sharpening can be divided into two separate classes; the first to correct loss of sharpness due to digital processing, and the other to optimise an images appearance in relation to the medium in which it will be displayed.

How it all works

In most cases, image sharpening is managed by applying a filter called “Unsharp Mask”, which despite the name, actually works to sharpen an image by exaggerating the brightness along the subjects edge(s).

Sharpness describes the clarity of detail contained within a photograph, but it can also be a visual aid for emphasising texture.  Even though post-processing can go a long way in improving our images, image sharpness is actually limited by camera equipment, image magnification and viewing distance.  There are two fundamental factors that contribute towards the perceived sharpness of an image; resolution, which describes the amount of pixels (or dots) per inch, and acutance, which is the transition of brightness levels around the edges within the frame.  It should be noted that an Unsharp Mask cannot create additional, or recover lost detail, but by applying sharpening to an image, it can create a better appearance of pronounced edges, thus making it appear sharper.

Concept

Taken directly from the Cambridge in Colour tutorial on Sharpening using the Unsharp Mask, the theory behind this filter works thus so:

Concept behind the Unsharp Mask Filter

Concept behind the Unsharp Mask Filter

The sharpening process works by utilising a slightly blurred version of the original image.  This is then subtracted away from the original to detect the presence of edges, creating the unsharp mask (effectively a high-pass filter).  Contrast is then selectively increased along the edges using this mask – leaving a sharper final image:

Note how it does not transform the edges of the letter into an ideal ‘step’, but instead exaggerates the light and dark edges of the transition.  An unsharp mask improves sharpness by increasing acutance, although resolution remains the same.

In practice

As is often the case in life, too much of a good thing can have disastrous consequences, and this can also be true when over using the Unsharp Mask.

Due to the manipulation of pixel details within our images, if over-sharpening occurs, artefacts known as sharpening halos will begin to appear around our subject, which is very unsightly and an unwanted presence within our work:

You can see from the images above that the third shot is very stark and that halos have appeared around the edges of the dragon, this has been caused by over zealous sharpening on my part.

However, as you can see when comparing the first and second images, when used correctly and within limits that show marked improvement, the unsharp mask is a great tool to add to any post-processing arsenal.

Other complications to be aware of when using the unsharp mask filter are possible shifts in colour.  This is due to the change in RGB pixel values applied during the sharpening effect and more details of this can be found at the following link: http://www.cambridgeincolour.com/tutorials/unsharp-mask.htm

Within the Unsharp Mask filter there are three settings to be aware of:

Unsharp Mask Filter Settings

Unsharp Mask Filter Settings

Playing around with each of these settings will enable you to get the sharpness of your images right, and as we will see below, certain combinations of settings work better for specific output sharpening.

Image Noise

Dependent on your tolerance for noise in images, any noise issues you have should be dealt with prior to using the Unsharp Mask.  This is so because image noise can influence our perception of sharpness; there may be times when much noise is present in an image, but due to a very high acutance, we may be tricked into thinking that the detail present within the shot is sharp, which is probably not the case.

A good place for dealing with image noise is in your software’s Camera RAW packages, such as Adobe Camera RAW, which is part of the Adobe Creative Suites Package.

Sharpening Workflow

It is agreed by most that effective sharpening occurs when it is applied more than once during any given post-processing workflow, and that the unsharp mask filter should become a permanent part of a photographers overall post-processing workflow.

There are three areas where sharpening should be considered:

Capture Sharpening: aimed to address blurring at the source, whilst taking image noise and detail into consideration.  Within our camera, blurring can be caused by the camera sensor’s anti-aliasing filter and demosaicing process; it can also be caused by our choice of lens. The camera applies sharpening automatically as it captures JPEG images, ensuring that images respond well to subsequent sharpening during post-processing.  However, RAW data files require manual manipulation in a RAW processing package on a computer.  It should be noted that images taken in landscape mode are usually much sharper than those taken in portrait.

Creative Sharpening: applied selectively, based on image content or artistic intent.  This type of sharpening is mainly used to selectively sharpen specific areas within an image, thus making the subject standout within the frame.  Key to this type of sharpening is the use of layer masks, which is a defined way of specifying the where, why and how of creative sharpening to be applied.

Output Sharpening: used at the very end of the image editing workflow, output sharpening concentrates on the amount of sharpening needed for optimum display of your image in any given media.  Once an image has been sharpened, its appearance should look nice and pin-sharp on screen, however, this is not usually sharp enough to produce a pin-sharp print.  Output sharpening requires a large leap of faith as it is nearly impossible to judge whether an image on screen has been sharpened correctly for print, in fact, output sharpening can make an on screen image look brittle, harsh and quiet mechanical.

When sharpening an image for print, image size and viewing distance needs to be taken into consideration, these, and other requirements will have bearing on the amounts of sharpening needed in post-processing.  Output sharpening therefore relies on mathematical formula to calculate the estimated sharpening radius; this is based on a) viewing distance, b) resolution of print (in DPI/PPI), c) printer used and d) type of paper.  As a general rule, larger viewing distances require larger sharpening radiuses, therefore, the key is having the radius small enough for the eye to see, but large enough that visibly improves sharpness.

When sharpening an image for use in digital media, especially if the image has been reduced in size to less than 50% of its original size, all existing sharpening halos are removed; therefore additional sharpening is needed to offset this effect.  The unsharp mask settings need to be reduced to a radius of 0.2-0.3 and an amount of 200%-400% these seem to work well, and although at such small values sharpening halos do not exist, other artefacts such as aliasing/pixilation and moiré could cause new issues.

The following points should also be taken into consideration before sharpening commences:

  • Sharpening is irreversible, so it is advisable to save all unsharpened originals whenever possible.
  • Due to their unprocessed properties, RAW & TIFF files respond much better to sharpening than JPEG files; sharpening my largely increase JPEG compression artefacts.
  • Blurring due to motion or camera shake may require advanced techniques such as deconvolution or Photoshop’s “smart sharpen” tool.
  • Some camera lenses do not blur objects equally in all directions, see the following Cambridge in Colour Tutorial on Lens Quality: http://www.cambridgeincolour.com/tutorials/lens-quality-mtf-resolution.htm#astigmatism
  • Images will often appear sharper if you also remove chromatic aberrations during RAW development. This option can be found under the “lens corrections” menu in Adobe Camera RAW, although most recent photo editing software offers a similar feature.
  • Grossly over-sharpened images can sometimes be partially recovered in Photoshop by (a) duplicating the layer, (b) applying a Gaussian blur of 0.2-0.5 pixels to this layer 2-5 times, (c) setting the blending mode of this top layer to “darken” and (d) potentially decreasing the layer’s opacity to reduce the effect.
  • The light sharpening halos are often more objectionable than the dark ones; advanced sharpening techniques sometimes get away with more aggressive sharpening by reducing the prominence of the former.
  • Don’t get too caught up with scrutinizing all the fine detail. Better photos (and more fun) can usually be achieved if this time is spent elsewhere.

So, there you have it.  Now I know the reasons for lens blur and the workings of Adobe’s unsharp mask, it is now a case of putting my learning’s into practice for the Sharpening for Print exercise that follows.

Source:

Reference:

Cambridge in Colour.  (n.d.) Guide to Image Sharpening [online article].  Available at: http://www.cambridgeincolour.com/tutorials/image-sharpening.htm [Accessed 8 November 2013].

Bibliography:

Cambridge in Colour.  (n.d.) Guide to Image Sharpening [online article].  Available at: http://www.cambridgeincolour.com/tutorials/image-sharpening.htm [Accessed 8 November 2013].

Cambridge in Colour.  (n.d.) Image Resizing for the Web & Email [online article].  Available at: http://www.cambridgeincolour.com/tutorials/image-resize-for-web.htm [Accessed 11 November 2013].

Cambridge in Colour.  (n.d.) Lens Quality: MTF, Resolution & Contrast [online article].  Available at: http://www.cambridgeincolour.com/tutorials/lens-quality-mtf-resolution.htm#astigmatism [Accessed 12 November 2013].

Cambridge in Colour.  (n.d.) Sharpening: Unsharp Mask [online article].  Available at: http://www.cambridgeincolour.com/tutorials/unsharp-mask.htm [Accessed 11 November 2013].

Cambridge in Colour.  (n.d.) Tutorials: Sharpness [online article].  Available at: http://www.cambridgeincolour.com/tutorials/sharpness.htm [Accessed 11 November 3013].

McNally, J.  (2013) Pro Tips On How To Hold Your Camera for Super Sharp Images [online video].  The Photo Argus.  Available at: http://www.thephotoargus.com/tips/pro-tips-hold-camera-super-sharp-images/ [Accessed 11 November 2013].

Reichmann, M.  (n.d.) Understanding Sharpness [online article].  Available at:  http://www.luminous-landscape.com/tutorials/sharpness.shtml [Accessed 11 November 2013].

Reichmann, M.  (n.d.) Understanding Resolution [online article].  Available at: http://www.luminous-landscape.com/tutorials/understanding-series/und_resolution.shtml [Accessed 11 November 2013].

Walsh, J.  (2010) Photoshop CS5: Michigan’s MI Learning, 0902 PS CS5 (Unsharpen Mask) [iTunes U podcast].  Available from the iTunes store.

Walsh, J.  (2010) Photoshop CS5: Michigan’s MI Learning, 0902 PS CS5 (Noise Reduction) [iTunes U podcast].  Available from the iTunes store.

Advertisements
This entry was posted in The Final Image ~ Learning Log and tagged , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s