Very Sharp Practice

- Page 1

Don't Try This Unless You Are Serious!

Sharpening remains contentious. We wrote on the topic in June last year and issues are still rumbling around. On a personal level only pre-sharpening of RAW files has been employed, followed by letting the Epson/Photoshop drivers and interface perform scaling of the image. This, we realise, is not 'best practice' but no complaints have been heard from any judges for images printed at 16"x20". We follow a similar practice for Professional Imagemaker pages and images. If a photographer provides images we usually convert them to CMYK, change the resolution to 300ppi and then save a TIFF file without any additional sharpening. This is placed into InDesign and scaled to fit the layout, allowing the InDesign engine to transform the size 'on-the-fly'. Again no issues have been raised, other than a some pre-flight warning for 'low resolution'.

 

Low-resolution warnings on pre-flight are flagged when the effective resolution of the image on the page has fallen (because of scaling) to less than a set value (usually 250ppi). This almost invariably occurs with adverts provided by companies who do not employ professional design studios to prepare their advertisements (in the last issue, a value of 76dpi was used at one point).

Sometimes the scaling of an image to a double page spread causes a low-resolution warning. In this situation we will either obtain a higher resolution file, use a different image or perform careful scaling (using Genuine Fractals) sometimes followed by careful sharpening and, if required, proofing at full size. This, best practice, we usually employ for the cover shot.

Martin Evening and Jeff Schewe have written extensively (and very well) on the topic. Their 'best practice' is to pre-sharpen from RAW and then apply a second, 'output' sharpening according to the device, and size of print. This is indeed good practice for 'best' work, but may not be needed for run-of-the-mill album and wall-portraiture work. There is a difference to what you do for this and what you might do for your 16"x20" Convention entry. This is not to encourage sloppy work, it is about fitness for purpose; you may not want to sharpen an image just after you have softened it to a dreamy, mood shot!

Aside from these issues there is the perennial question, 'what is the correct level of sharpening'. This is an unanswerable question, it depends upon lots of factors, including some, or all, of the following: the viewer's eyesight*, the viewing distance, the print size, the paper surface, the illumination level, the subject properties (eg a bird's feathers or a model's skin) and finally the taste of the viewer. A neutral response to an image is probably the best measure of success. If a group of judges are viewing an image and sharpening is mentioned it might well mean 'too sharp'; it could also mean unsharp! By a neutral response we mean that none of the judges says anything, suggesting everybody is concentrating on the image and issues of sharpening are not under consideration. Obviously a grossly over-sharpened image with haloes around everything is flawed and will probably be viewed as such by all present.

 

 

 

One viewer said that they could not detect any difference in a test matrix, whereupon we suggested they use the reading glasses they had suspended from their neck. What was interesting was the fact that this viewer did not seem to consider it important to wear their reading spectacles to complete the task, whereas we assumed that they would – in many ways both of us were at fault!

Judging sharpness can only be truly carried out on a print, for that is a tangible thing. A screen image is a variable, very evident if the same file is viewed on a number of different monitors. When carrying out the tests for this feature we had images which looked soft on the CRT but massively over-sharpened on the LCD alongside – the same image simply dragged across from one screen to the other. This is an issue for digitally projected or screen-viewed competitions, over-sharpening is a common complaint from judges but under-sharpening also crops up quite regularly. Much of it is outside the control of the photographer!

As far as we can tell from a limited Google search nobody has done a correctly balanced trial to discover the 'preferred' amount of sharpening. We found a number of esoteric articles on Google Scholar but all requested money ($30 to find we had no interest in sharpening goofy images from outer space was unattractive) , so we decided to do one ourselves (admittedly a slightly flawed piece of experimental design, but an attempt all the same).

Page 1  -  Page 2

Nature, Landscape and Wildlife Photographers join today and receive a fantastic free gift:

Apply Here £99 a Year
or
Apply Here £10 a Month