I get a lot of questions about aperture settings. At which value do we get better results? Should we open it or close it? When dealing with the subject, I look at it from a purely macro point of view. In macro, we usually focus on detail. In portrait or landscape shooting, we may not need so much detail.
As much as I would like to keep the theory short, I cannot do so without going into it. But since I have emphasized some concepts in previous articles, I will not repeat them. Don’t worry, there are not only concepts, we will also see practical results.
- Depth of field
- Aperture
- Effective aperture
Therefore, if you feel that the concepts in the list above are missing, it may be useful to first take a look at these two articles
Megapixels, sensor size, aperture, ISO (Entry-level concepts)
Why is the depth of field so narrow? What is effective aperture? (Aperture overview)
Which one is sharper? A closed aperture or an open aperture?
I often see it in photo posts and even read it in some articles. In macro photography, the aperture is closed down as much as possible to widen the sharp area. Looking at this information, we take photos at the most extreme value allowed by the lens, such as f32. We see f64 in some macro lenses and think it is a superior feature.
The f64 value comes from the “effective aperture” measurement. When a macro lens focuses to infinity, the aperture value we set is matched to the amount of light let in, i.e. it shows the correct value. However, when we change it closer to the minimum focus distance, that is, when we increase the magnification, the light we let in decreases. Up to 0.5X magnification, there is no serious effect. But when we get to 1:1 magnification, we are now at f64 in the aperture as the effective value. So even if the lens shows f32, we are actually working with f64. We need to consider this effective value when considering the aperture we will choose. Diffraction effect happens according to it.
Some bodies are smart and show the effective aperture in EXIF, some don’t. Because we usually use manual lenses, reversed lenses, the body has no idea about the lens. We’d better take over the smart behavior.
So is f32 a bad thing? I think yes! At any magnification…
We can use f32 if we have to. But our goal is “not to have to”. We need to find a better method and work without closing the aperture so much. Or what? First, theory again.
Diffraction
We need to go back to high school and remember physics lessons. Waves, whether sound waves, water waves or light, diffract and scatter as they pass through a narrow space. The narrower the hole through which light passes, the greater this effect is.
Diffraction at open aperture 
Diffraction at closed aperture
As you can see, if the aperture is open more than a particular value, the light passing through the lens can continue its path undistorted. But when the aperture is too small, the shape/direction of the light changes and becomes scattered. If we look at it from the lens point of view, the light passing through our lens is not in the form of a flat wave as in the figure, but in the form of a cone that shrinks and focuses towards the sensor. When we focus correctly, the pointed end of the cone-shaped light of the focused point hits the sensor and we get a sharp focused image.
If this focused and point-like image is diffracted due to a small aperture, the image formed on the sensor is no longer point-like. If we think of this point as a pixel of detail, instead of giving a clear, sharp detail, its image blends with other pixels around it, like the effect of a stone thrown into water. Imagine all the pixels that make up the image affecting their surroundings in this way. It’s not a situation we like.
The photo shows how a diffracted point of light spreads around.
Diffraction – Sensor size
The effect of diffraction is directly related to sensor size and megapixels. Let’s imagine that the distortion in the next photo falls on the sensor. If the distorted point of light is small enough to fit within 1 pixel, then it cannot affect neighboring pixels. But on a small sensor or a high-megapixel camera, the area where this tiny light falls has maybe 10 pixels instead of 1. Or as we reduce the aperture, we increase the diameter of the distortion. This means that the light falling on all neighboring pixels is mixed together. Even if the resolution offered by the camera has high megapixels, it is garbage because of diffraction.
Diffraction tests
Now let’s go back to the original question. Does all this talk have any practical significance?
It is necessary to decide according to the purpose and necessities. If we are going to use the photo on a web page in a small size, we will not see any diffraction effect. There is no need to calculate anything for small-scale use. Everything seems to be fine. We don’t care about the distortion. Or if we don’t have the chance to do a focus stack on an insect, even handheld, we want to increase the depth of field and we enter the diffraction zone.
We’d better test it and see the results.
I decided to test 3 different magnifications: 1X, 2X and 3X. For each of these I wanted to use a different lens system so that the effective apertures would be different. I didn’t measure the effective apertures though, measuring them is a bit of a pain, and I want to give the essence rather than confuse with numbers. We already talk enough about concepts.
In macro lens tests we should choose an object with as more detail as possible so that we can easily see the differences between the two shots. Butterfly wings are often used for this purpose. This is because of the tiny scales on them. Even with high magnification microscope lenses, you get endless detail in these scales. That’s why I used the wings of the silkworm moth and the hopper beetle. I didn’t make any edits to the sharpness, etc. I just corrected the tiny exposure differences between them because different apertures resulted in different exposure times.
1X magnification, Tamron 90mm
Our insect is a bit small for 1X, about the size of the nail of my index finger. But no matter, the point is not to show it but to see the effect of aperture on detail. If we put a series of photos taken at f3.5 – f32 side by side, we get the following view.
As I said before, when used in small sizes, the only noticeable difference between the photos seems to be the depth of field. At f3.5, the background is completely erased, whereas at f32, not only is our insect completely clear, but we can now see what’s going on in the background. Instead of focus stacking, we finish the job with a single frame at f32. Or if we don’t want to make the background too obvious, we can use f18 without much exaggeration. But is it really so? Let’s take a closer look. I’ve created the comparison below with 100% crops from these images.
Now we are beginning to realize that something serious is going on here. You can click on the photo and enlarge it to see it more clearly.
At f3.5 the depth of field is very narrow, sharpness is not bad. At f5.6 and f8 it is clear that sharpness goes to maximum. f5.6 looks a bit better. At f13 the diffraction effect starts to erase the details. f18 produces an unusable result in my opinion. f32 is a disaster, the photo looks like watercolor. High level detail can’t pass through the aperture which is getting narrower and narrower. Diffraction effect really ruined our work, as well as the transmittance of the glass.
There’s one more thing that stands out. Surprise! Our sensor dust is revealed! The sensor dust shadows are blurred and spread out like bokeh at open apertures, while they are small but solid at low aperture. I didn’t clean the dust with Photoshop so that these can be seen as well.
2X magnification, Tamron 90mm + Raynox DCR-250
These two, which I love very much, don’t actually give 2X. Our magnification value is 1.8X. Don’t think that we didn’t change the lens because we used Tamron 90mm again, we did something very strange in the optical design when we put Raynox on.
I mentioned that the effective aperture is used in the calculations of depth of field and diffraction. Actually, it is wrong to say “in the calculations” because this is not a virtual effect, it is projected exactly on the photograph. Since we place a magnifying glass in front of the lens, we let in the light that the magnifying glass collects and increases. For the same aperture value, about 3 times more light comes in. With Raynox, the effective aperture is lighter (smaller) than the lens shows.
Compared to the previous situation, the depth of field seems to have decreased. This is due to both the magnification and the switch to an effectively more open aperture.
Looking at the 100% crop comparison, we see that f2.8 and f4 are too soft. The effective aperture operates the lens at a more open aperture than it shows on the lens. The Tamron 90, which gives good results at f3.5 when used alone, is still very soft at f4 with Raynox, working as if it were f2.8. At f5.6 the lens recovers quickly. f7.1 and f10 produce the sharpest results. From f16 we see diffraction. As the aperture is closed down, the details quickly fade.
Even though Raynox increased the magnification, it gave us a brighter viewfinder, shorter exposure time and a diffraction effect that appeared later. We saw no diffraction at f10 even though we doubled the magnification. If the effective aperture had not changed, we would have seen blurring at f10. Front-mounted close-up attachments like Raynox have such an advantage. If we had used a teleconverter, the opposite would have happened as the magnification increased. The disadvantage of Raynox is that it shortens the working distance considerably. The teleconverter does not change the working distance.
3X, Componon-S 50mm
We use our lens, which has a considerable place amongst the 50mm enlarger lenses, at 3X magnification and reversed. The subject is again the wing of our hopper insect. Let’s arrange the photos side by side.
The smallest aperture of our lens is f16, so that’s as far as we can test it. But when we look at the results, we will see that even f16 is too much. Again, although the lens starts at f2.8, I started the test with f3.5, I know that at f2.8 it will give soft images and have very narrow depth of field. Let’s take a closer look, don’t forget to click and enlarge.
f3.5 might trick you. The depth of field is very narrow, barely containing the tip of a feather. Sharpness is not bad when we look carefully, but there is a better one: f5.6. This gives the sharpest result we are looking for. At f8 diffraction starts and after that it’s useless.
With Tamron + Raynox we were getting sharp results at f10 and now we are restricted at f8, why? Because the magnification has increased. The formula for the effective aperture is as follows:
Effective Aperture = Lens Aperture * (1 + Magnification / Pupil magnification) we can assume pupil magnification as 1. For detailed information, you can check the links I gave at the beginning of the page.
So if we use f10 for 1X magnification it gives, 10 * (1 + 1) = f20, and for the 3X we tested 10 * (1 + 3) = effective f40. That’s a lot of diffraction.
Conclusion
As the magnification increases, we need to use the lenses with a little more open aperture. But too open aperture doesn’t give good results either. While trying to avoid diffraction, we can also enter the region where the lens works soft. This is where the quality difference of the lenses emerges. Each lens has a so-called “sweet spot” where it works with the highest sharpness. If magnification allows, it is best to use our lens at the “sweet spot” value. But if we are working at a magnification where diffraction starts at that value, we may need to open the aperture a little more.
Lenses used for high magnification such as Componon 35mm, Rodagon 28mm give the best results at the widest aperture. It is important to know the lens and pay attention to this when choosing an aperture.
Note: There are older lenses that have very poor sharpness but aesthetically create a “dreamlike” look. They are successfully used for macro and close-ups. The aim is sometimes not detail but framing and aesthetics. For example Meyer Trioplan 100mm f2.8
Additional test
What if we are not shooting macro? Does diffraction also affect the result in daily photography?
I did a simple shooting test on a detailed object, the texture of a curtain, at a distance of 4m. f6.3 and f32 were used.
The distortion is obvious. Magnify it and the difference is better visible. This is especially important for those who want to take long exposures during the day. In order to make a long exposure in bright weather, sometimes the aperture needs to be closed too much. Instead, I think it would be best to use an ND filter of the appropriate type and not close the aperture too much.
I wish you diffraction-free shots 🙂
