Tuesday, 11 November 2014

Sharpening

Couldn't help noticing that some of the pictures entered
into our club competitions don't look sharp enough, and in some cases look
over-sharpened. It helps a lot to get sharpening right and I thought a bit of a
meandering discussion on this blog might help those members who struggle with
it.

Blogs work best in bite-sized chunks, so there will be
several episodes to this saga - unless somebody stops me. But when I've
finished, I hope you will understand what digital sharpening actually does and
how to get it right in the various circumstances you might use it.

Like many things humans arrogantly take credit for inventing
(like flying machines), nature has been using the principles of sharpening for
advertisement and deception for at least a hundred millions years.

But before I say anything about image sharpening (which I
will have to keep for the next blog) there is something I have to get out of
the way first - image resolution.

Resolution is about how much detail a system enables us to
discern.

You could be forgiven for thinking that the more detail
there is in a picture the sharper it will look. This is not the case. When we
were choosing a projector for the club, we compared side-by-side an HD
projector (1920 x 1080 pixels) with an XGA projector (1024 x 768 pixels), the
one we finished up buying. The same images looked significantly sharper on the
lower resolution projector, one of the reasons we chose it.

So chasing more and more pixels does not give you sharper
pictures. Image resolution and image sharpness need to be considered as
separate entities.

Most of the parameters we use in photography can be given a
precise value. Think of shutter speed (seconds), aperture (f number), digital
sensor and film sensitivity (ISO). The resolving power of a system can also be
measured and given a value. This is true whether it applies to a digital
sensor, a film, a screen, a projector, a lens, or the eye.

This can be done by observing an image of closely spaced
lines. The higher the resolution, the closer the spacing can be before the
lines cannot be distinguished from each other.

  


 

            Test charts in high and low resolution   

 

Sharpness is something different and I am not aware of any
standardised way of measuring it. Why is that I wonder? There is a subjectivity
about sharpness; yet we all know when a picture looks sharp and when it
doesn't, don't we?

The next blog will explain what sharpness actually is, and
why exploiting the principle was so important a factor driving the evolution of
life in the struggle for survival.

(I'm a biologist, OK... get over it!)


Art is what you can get away
with.

Andy Warhol

No comments:

Post a Comment