Amateur Photographer

Depth of field and sharpness

Do we trade off sharpness for deeper depth of field? We need to understand this relation for better compositio­n

- Bob Newman is currently Professor of Computer Science at the University of Wolverhamp­ton. He has been working with the design and developmen­t of high-technology equipment for 35 years and two of his products have won innovation awards. Bob is also a camer

In the seemingly endless discussion­s concerning the merits of different sensor sizes, the issue of depth of field is a recurring topic. The position which is often put forward is that smaller sensors offer more depth of field, and more depth of field provides sharper images. Apart from the fact that simply stopping down allows a largeforma­t camera to achieve deep depth of field, the problem with this argument is that deep depth of field inevitably comes at the cost of image sharpness.

Irrespecti­ve of how deep or shallow the depth of field is, it is only at the plane of focus where the full sharpness that the lens can offer is achieved. Everywhere, apart from that plane, the image is to some extent unsharp. The familiar depth- of-field equations are based around a statement of how much unsharpnes­s the photograph­er is willing to accept as ‘sharp’. This parameter is called the ‘circle of confusion’ or COC.

There are several convention­s for deciding on the COC, but the most common one is what is called ‘d/1500’, which means that the defocus blur counted as ‘sharp’, should be smaller than the image diagonal divided by 1,500. To think how ‘sharp’ this is, consider that to render a line pair, we need a fully rendered point plus an empty space of similar dimension beside it. So, ‘d/1500’ turns out to be 750 line pairs per image diagonal, or for a 3:2 aspect ratio frame, 500 line pairs per image height. Or, we could frame this in terms of pixel count: 1,000 pixels by 1,500, or 1.5megapixel­s. This resolution would equate to a very mediocre lens indeed. It’s interestin­g to speculate how many people would knowingly accept such a lens as the price for deep depth of field.

Sacrificin­g one for the other

The other factor that downgrades image quality with deep depth of field is diffractio­n. Suppose that using a standard lens, one wanted depth of field from 2.6 metres to infinity. The standard DOF calculatio­ns suggest that this requires a hyperfocal distance of 5.26m and f/16 on full frame, f/11 on APS- C or f/8 on Micro Four Thirds. At this aperture, diffractio­n limits the resolution to 1,100 line pairs per picture height. This is approximat­ely twice that of the worst- case blur within the depth of field, but applies even to the plane of focus. However, in terms of the resolution expected from a quality lens, this is still very modest.

So, in the end, deep depth of field always comes at the cost of absolute image sharpness, and loss of sharpness away from the plane of focus. That is not to say that it should never be used; just that it is worthwhile knowing the tradeoff. It’s also worthwhile not using more depth of field than is necessary for your compositio­n, if you value sharpness.

 ??  ?? Sharpness is always sacrificed for deep depth of field. Worthwhile knowing, if this is a compositio­nal aim
Sharpness is always sacrificed for deep depth of field. Worthwhile knowing, if this is a compositio­nal aim
 ??  ??

Newspapers in English

Newspapers from United Kingdom