 |
| See text below for explanation. By the way this photo was made on a Canon EOS M50 with a budget 15-45mm lens which actually does a pretty good job right across its focal length range. |
The current Covid-19 lockdown is pretty much bad news for everyone but can have some little upsides.
One of these is that I have been investigating the behaviour of my camera lenses in ways which I have not done previously in any systematic fashion.
This has led me to realise that lenses do not all behave the same way when the aperture diaphragm is closed down. This is of particular relevance to landscape and similar situations where depth of field is an important element of the photograph.
A lens at its widest aperture (smallest f-stop) produces a shallow depth of acceptable sharpness (a.k.a. depth of field, a.k.a DOF) in the photo.
As the lens aperture is decreased (f-number increased) the DOF increases. This DOF is distributed both away from the sharply in focus plane (represented in the viewfinder as up) and towards the camera from that plane (represented in the viewfinder as down).
If you consult a photography information source such as cambridgeincolour.comor dofmaster.com you can read more about this.
But these sources can be misleading or unhelpful as they do not take account of the actual characteristic behaviour of different lenses and they do not provide a handy guide to optimal focussing which can be applied right in the viewfinder.
The next problem is modern high pixel density cameras. As pixel counts climb sharpness loss due to diffraction at the lens aperture becomes a more pressing issue. Gone are the days when we could just stop down to f22 and hope for the best.
Small sensor cameras can be diffraction limited starting at f4. High pixel count full frame cameras can start to show the effects of diffraction at the aperture diaphragm at f11.
In practice this means that you want to know rather precisely what is the largest aperture (smallest f-number) which will expand the zone of acceptable sharpness to the whole frame but no further and will bring the background and foreground into acceptable sharpness simultaneously.
This post describes a process by which you can do that.
The first step is to characterise the behaviour of the lens as the aperture is closed down.
Every lens is different. They do NOT all behave the same way.
For instance I have a Canon RF 24-105mm f4 L and a Canon RF 24-105mm f4-7.1 STM. Each lens has the same focal length range, is made by Canon and was tested on the same camera. The way each responds to progressive closure of the lens aperture is very different.
The zone of acceptable sharpness from the f4L lens moves down the frame more than up the frame.
The zone of acceptable sharpness from the STM lens moves up the frame much more than down the frame.
This means I need to place the focus box in a different part of the picture frame for each lens.
See the photo in portrait (vertical) orientation.
For the test procedure I use portrait orientation as this is more demanding on near-far focus depth than landscape orientation.
Start with the focus box in the middle of the viewfinder. Frame up a scene with foreground elements close to the camera and background elements far from the camera. Place a focus target (I just use two crossed sticks) so it locates in the center of the frame, covered by the focus box indicator in the viewfinder.
Now make a set of photos starting at the smallest f number then in whole f-stops to about f16.
For the test photo here I used a Canon RF 35mm f1.8 starting at f1.8, then f2, 2.8, 4, 5.6, 8, 11, 16.
Check out the resulting photos on screen. You will see the “sharp enough” zone expanding up (away from the camera) and down (towards the camera).
With the RF 35mm f1.8 the DOF expands up the frame and down the frame at the same rate.
This means that when I use this lens for landscape I will place the AF box in the center of the frame or over a suitable subject element very close to the center of the frame.
But when I repeat this test with the 24-105mm STM lens I find that if I start with the AF box over a focus target at the center of the frame in the viewfinder then as I close down the aperture sharpness extends up (towards the background) much more than down (to the foreground) .
So when I use this lens for landscape I have to place the AF box significantly lowerthan the center in the viewfinder frame.
When I switch to the 24-105mm f4L the opposite occurs. As the aperture is closed the sharpness moves down (towards the camera) more than up (towards the background).
So for this lens I need to start with the AF box over a suitable part of the subject higher in the frame than the center.
The headline photo was made with a Canon EF-M 15-45mm lens on a Canon EOS M50.
I knew from previous testing that this lens brings the foreground into sharpness more than the background as the lens aperture is closed down.
Therefore I placed the focus box where you see it, higher in the frame than the centre. I made exposures at f5.6, 8, 9, 10, 11 and found that with this subject f8 gave me acceptable sharpness in the background and foreground. I find with this camera that sharpness starts to go off a bit from f11 so f8 is ideal for a landscape type subject such as this one.
Summary
Every lens has its own characteristic depth of field distribution as the aperture is closed down. This can be observed quite easily.
This affects optimal location of the focus box in the frame particularly with a landscape type subject.
My practice is to make a little note [F.Near] or [F.Far] on stickyback paper and attach this to each lens to remind me which way to bias focus for optimal results.
With modern high pixel count cameras closing the aperture more than necessary could result in loss of sharpness across the whole frame due to diffraction.