Potential for background blur

It is well known that the maximum amount of background blur depends on the size of the entrance pupil of the lens. The maximum blur occurs with a background at infinity and the blur size is equal to the diameter of the entrance pupil.

Just for fun, I lined up five different lenses to visually compare the entrance pupils (the hole that you see when looking into the front of the lens):


The lenses are 12mm f/2.8, 45mm f/1.8, 75mm f/1.8, 25mm f/1.4 and 9mm f/4. The entrance pupils are 4, 25, 42, 18 and 2 (all in mm).

The very large variation in the size of the entrance pupil is obvious.

Just to relate this to actual photographs, here are shots taken with the smallest and largest entrance pupils of the lenses above.




9mm f/4. Entrance pupil is 2.25mm




75mm f/1.8. Entrance pupil is 42mm

The first image has the distant parts of the background blurred by about 2mm, as measured on the same scale as the flower (the larger bloom is about 80mm diameter).

In the second image, the blurring on the distant skyline is hard to estimate precisely in size, but appears consistent with the 42mm entrance pupil.

The only way to easily measure the blur diameter is if the background contains a small point of very bright light that appears as a sharply defined blur disc in the image.

Looking into the front of your lens to see the size of the entrance pupil is a very easy way to estimate how much blur to expect (from a distant background). Of course, if you are stopping down, remember to reduce the entrance pupil size accordingly.

Great little demo

It’s one of the reason I’m not keen on equivalent aperture. The pupil size is in some ways more useful

Tom Axford said:

Tom Axford wrote:.........

Looking into the front of your lens to see the size of the entrance pupil is a very easy way to estimate how much blur to expect (from a distant background). Of course, if you are stopping down, remember to reduce the entrance pupil size accordingly.

Click to expand...

Looking at the entrance pupil (that is the aperture blades seen from the front) works very well with manual lenses.

- in a zoom, the entrance pupil size changes with the focal length setting. So when looking into the lens to compare entrance pupil, you have to set it to the FL you are intending to use.

- with our AF lenses, the aperture is normally closed when off camera. You would have to take the lens off with the camera power on. But even then, not all lenses will have the aperture fully open, because they may do the AF with slightly closed aperture and only open it fully when taking the shot.

- constant aperture zooms may electrically progressively close the aperture at the wider settings, to achieve that constant aperture. You cannot replicate that with the lens off camera with no power. If you closely watch the aperture blades of say the 12-40/2.8 pro lens with camera turned on, and rotate the zoom ring, you will see the aperture blades physically move (its not just an optical effect). This lens is really a variable aperture zoom like 12-40/2.2-2.8, that was electronically made into a constant aperture lens (that is also why it is as large as it is).

***

The entrance pupil diameter is the focal length divided by the f-number

Hence it is much easier with AF lenses to calculate the entrance pupil. For example with the 12/2.0 lens, the entrance pupil is 12 / 2 = 6mm, with the 75/1.8 it is 41.66mm

It is safer too, because taking off the lens under power is not recommended.

Interesting display of the enterance pupils.

Those two example images appear to have been taken from different positions - I would would have liked to have seen the fence in both images to see how much blurring there was for objects closer behind the flower. With the close fence absent in the second image all we see are the distant trees and so the subject isolation is exaggerated.

Peter

john Clinch said:

Great little demo

It’s one of the reason I’m not keen on equivalent aperture. The pupil size is in some ways more useful

Click to expand...

Thanks John.

Yes, equivalence works for depth of field and for diffraction, but it does not work for background blur.

Thanks for your comments.

Very conveniently, all my MFT lenses are in the fully open state when powered down and removed from the camera. I never remove a lens while the camera is switched on.

I haven't come across any MFT lenses that are stopped down when removed from the camera.

The two zooms I included are manual zooms and were set at their shortest focal length when I photographed the entrance pupils.

However, there is still an uncertainty about the exact state of the lenses in that they all have internal focussing and when they are removed from the camera, they may or may not be in the infinity focus position. If a lens is not in the infinity focus position, this may have some effect on the entrance pupil. However, in this context, it is not significant. After all, the lenses were not at infinity focus for the example shots I took.

Unfortunately, modern lenses are impossible to control (i.e. set focus and aperture) when removed from the camera unless you can replicate the camera's electrical signals to the lens.

Messier Object said:

Interesting display of the enterance pupils.

Those two example images appear to have been taken from different positions - I would would have liked to have seen the fence in both images to see how much blurring there was for objects closer behind the flower. With the close fence absent in the second image all we see are the distant trees and so the subject isolation is exaggerated.

Peter

Click to expand...

Thanks for your comments.

The example I included was far from ideal. The huge difference in angle of view between 9mm and 75mm made it very difficult to get a suitable background for the second shot. I think a better example would have objects at a variety of distances away.

Out of pure laziness, I shot the hollyhock flowers because they were close at hand (and they look nice).

Sigma 56mm 1.4 has also got massive pupil, and really good background blur.

Mait said:

Sigma 56mm 1.4 has also got massive pupil, and really good background blur.

Click to expand...

Yes, at 40mm entrance pupil it is almost exactly the same as the 75.

Not always sharp but I love the blur and bokeh this ancient combo creates.

--
Stupidity is far more fascinating than intelligence. Intelligence has its limits...

Tom Axford said:

Thanks for your comments.

Very conveniently, all my MFT lenses are in the fully open state when powered down and removed from the camera. I never remove a lens while the camera is switched on.

I haven't come across any MFT lenses that are stopped down when removed from the camera.

Click to expand...

Yes, many m43 lenses are fully open when powered down and removed.

Other m43 lenses behave differently. There seems not to be a standard for this.

For example, among my personal lens collection about 2/3 power off with closed aperture, but 1/3 power off with either fully or partially closed aperture:

- the Oly 12/2.0 closes the aperture about one stop when powering down and removing, (however, the closely related Oly 17/1.8 powers off with open aperture).

- the Lumix 25/1.7 fully closes the aperture

- the Oly 12-40/2.8 fully closes the aperture

- the Oly 40-150/2.8 almost fully closes (hard to say, it closes to about 5mm entry pupil which may well be the maximum it can close)

Maybe you check your own lens collection to make sure yours really all power off fully open. You may be surprised.

Tom Axford said:

The two zooms I included are manual zooms and were set at their shortest focal length when I photographed the entrance pupils.

However, there is still an uncertainty about the exact state of the lenses in that they all have internal focussing and when they are removed from the camera, they may or may not be in the infinity focus position. If a lens is not in the infinity focus position, this may have some effect on the entrance pupil. However, in this context, it is not significant. After all, the lenses were not at infinity focus for the example shots I took.

Unfortunately, modern lenses are impossible to control (i.e. set focus and aperture) when removed from the camera unless you can replicate the camera's electrical signals to the lens.

Click to expand...

As someone considering m4/3 I find this really interesting as one of the downsides frequently quoted is subject separation. I'd be keen to see more examples with different lenses...

--
Stupidity is far more fascinating than intelligence. Intelligence has its limits...

cba_melbourne said:

Tom Axford said:

Thanks for your comments.

Very conveniently, all my MFT lenses are in the fully open state when powered down and removed from the camera. I never remove a lens while the camera is switched on.

I haven't come across any MFT lenses that are stopped down when removed from the camera.

Click to expand...

Yes, many m43 lenses are fully open when powered down and removed.

Other m43 lenses behave differently. There seems not to be a standard for this.

For example, among my personal lens collection about 2/3 power off with closed aperture, but 1/3 power off with either fully or partially closed aperture:

- the Oly 12/2.0 closes the aperture about one stop when powering down and removing, (however, the closely related Oly 17/1.8 powers off with open aperture).

- the Lumix 25/1.7 fully closes the aperture

- the Oly 12-40/2.8 fully closes the aperture

- the Oly 40-150/2.8 almost fully closes (hard to say, it closes to about 5mm entry pupil which may well be the maximum it can close)

Maybe you check your own lens collection to make sure yours really all power off fully open. You may be surprised.

Click to expand...

Yes, I checked the rest of my lenses and one is closed when powered down and/or removed from the camera. It is the 25mm f/1.7 which I rarely use as I also have the 25/1.4. It certainly makes it almost impossible to do any tests on the lens while it is off the camera!

Thanks for drawing attention to this!

Lovely shots and examples! Thank you!

This is a fantastic comparison! Thanks for sharing!

Tom Axford said:

john Clinch said:

Great little demo

It’s one of the reason I’m not keen on equivalent aperture. The pupil size is in some ways more useful

Click to expand...

Thanks John.

Yes, equivalence works for depth of field and for diffraction, but it does not work for background blur.

Click to expand...

As mentioned the size of the entrance pupil affects background blur. Also, as mentioned, the size of the entrance pupil can be calculated as the focal length divided by the F stop. SO...

You take a theoretical 50mm f4 normal lens on full frame. It will have a entrance pupil of 50mm / 4 which gives 12.5 mm.

Now you to get an "equivalent" micro four thirds normal lens you would need a 25mm f2. It will have an entrance pupil of 25mm / 2 which gives you the same 12.5mm!

Since they have the same entrance pupil size, shouldn't the blur amount be the same? Or am I not understanding something here?

Chris

ChrisPCrunch said:

Tom Axford said:

john Clinch said:

Great little demo

It’s one of the reason I’m not keen on equivalent aperture. The pupil size is in some ways more useful

Click to expand...

Thanks John.

Yes, equivalence works for depth of field and for diffraction, but it does not work for background blur.

Click to expand...

As mentioned the size of the entrance pupil affects background blur. Also, as mentioned, the size of the entrance pupil can be calculated as the focal length divided by the F stop. SO...

You take a theoretical 50mm f4 normal lens on full frame. It will have a entrance pupil of 50mm / 4 which gives 12.5 mm.

Now you to get an "equivalent" micro four thirds normal lens you would need a 25mm f2. It will have an entrance pupil of 25mm / 2 which gives you the same 12.5mm!

Since they have the same entrance pupil size, shouldn't the blur amount be the same? Or am I not understanding something here?

Chris

Click to expand...

You are absolutely right! My statement that "equivalence works for ..." was incorrectly worded and wrong. I was thinking of relative aperture (f-number), which I think is what John Clinch meant when he said "equivalent" aperture.

I should have said "relative aperture (usually called f-number) works for depth of field and for diffraction, but entrance pupil size is required for background blur".

The term "aperture" has been the source of some confusion for a very long time in photography. For working out exposure as well as many other things in photography, the relative aperture or f-number is much more convenient. So much so that the term "aperture" is generally used to mean the relative aperture or f-number. However, for some things the actual aperture (or, more correctly, the entrance pupil) is needed.

as stated by others already, you can easily calculate the entrance pupil = blur potential by dividing the focal length by the aperture.

for reference heres some popular lenses used commonly for isolating subjects (professional portraits) and their blur potential (FF lenses)

35/1.4 = 25

50/1.4 = 36

85/1.4 = 61

135/1.8 = 75

70-200/2.8 = 25-71

compared to the blur potential of m43 lenses

17/1.2 = 14

25/1.4 = 18

42.5/1.2 = 35

75/1.8 = 42

its apparent that you can get more subject isolation with shorter lenses (usually 1/2 the fl) with FF lenses.

But if you can keep enough distance to frame your subject isolating the subject from the background is possible, especially with the 75mmf/1.8,

I am not sure what you think is too complicated? All you need to know is the size of the entrance pupil which (as you say) can be got by dividing the focal length by the f-number.

The entrance pupil determines the size of the blur irrespective of the sensor size. Bigger sensors typically use bigger lenses with bigger entrance pupils for the same angle of view.

A good example of this is this recent article about a DIY camera using an effective sensor size of nearly 500mm with a 432mm f/5 lens.