To stack or not to stack

gardenersassistant wrote:

Here are a pair of images captured under more controlled conditions. I set down a line of tape on a table and mounted a light coloured plastic rule at an angle to the tape and a metal rule perpendicular to the tape, as shown below.

I captured an image with two cameras, placing each camera on the table and using the line of the tape to help me align each camera so it was as best as I could make it head on to the metal rule and therefore at the same angle to the plastic rule. The vertical angle on the scene varied a bit as the cameras and lenses were of different sizes and shapes and needed to be differently supported so as to be able to frame both rules.

Here is an uncropped image captured with a Panasonic FZ330 camera, which has a 1/2.3" sensor, with the f-number set to its maximum f-number of f/8, and with a Raynox 250 on the camera lens.

Here is an uncropped image with the same scene width of 12mm captured with a full frame Sony A7ii with a Meike 60mm macro lens with a 2X teleconverter, a 1.4X teleconverter and 68mm extension tubes. The f-number was set to f/22.

The image captured with the small sensor FZ330 using its maximum f-number of f/8 (and therefore giving the maximum achievable depth of field for that setup) has less depth of field than the image captured with the full frame sensor A7ii.

OK, that's fine. But the smaller sensor isn't causing the drop in depth of field, the limited maximum Fstop is.

Why was it necessary to add two teleconverters and extension tubes to the 60mm macro for the full frame shot? Did it really take all of that kit just to get to 3x (36mm wide sensor / 12mm)? Seems like just the TCs would have taken that lens to 3.4x.

If I'm getting the specs right the 1 2/3" sensor is 6.17mm wide, so that shot is about 1/2 life size.

gardenersassistant wrote:

mawyatt2002 wrote:

gardenersassistant wrote:

I use stacks not for the detail you can see if you zoom in, because I produce images for viewing "as is" without zooming in on them. I use stacking because I like the look of the images I can produce using it.

Nick,

Interesting comment. Can you elaborate more on the "look" stacking gives? Is this the in focus areas or the out of focus areas, bohea so to speak?

Best,

Both the in focus areas and the out of focus areas.

One of the reasons I use aperture bracketing (for non-stacked images) is that it gives me a set of images from which I can choose the one(s) which have the balance between the amount of the subject that is in focus and the rendition of the background that best pleases my eye.

For a lot of people a highly out of focus background is very pleasing, and some prefer more or less, or completely, featureless backgrounds, for example replacing a natural background with an artificial one by placing a card behind the subject. I generally prefer more going on in my backgrounds than that, and I prefer natural backgrounds, and I spend time exploring angles of attack looking for ones which produce a background that seems to me to be more or less harmonious with the subject in terms of lines, shapes, textures and/or illumination. And with single images you tend to get a smooth transition between the in focus and out of focus areas (unless the entire subject is rendered in focus against a background that is relatively far away and out of focus; with the whole subject in focus and the whole background out of focus this is not problematic).

Sometimes I can find a single-capture image of a scene that gives a combination of in focus and out of focus areas, and a transition between them, that pleases my eye. Sometimes I can't.

Sometimes when I can't get a balance that I like between subject coverage and background rendition with a single-capture image, I can get it with a stacked image. In fact, I'm finding that that is quite often the case. I haven't counted, but I would guess that at present I'm choosing to use a stack for perhaps two thirds or so of the scenes and single-capture images for the rest.

Stacked images can have a different look from single-capture images, with a rather sudden transition between the in focus and out of focus areas. Sometimes this can look similar to the subject-against-a-distant-background look of a single capture image, but sometimes there can be a rather jarring transition, for example if a stem, branch, leaf or petal jumps from being in focus to out of focus. I try to pick the rearmost element of the stack such that the transition isn't too jarring, but sometimes I can't find a suitable break point. And one of the curious things about capturing images (or mainly videos in my case) for stacking is that I can't compose with the final image in mind. With single-image captures I can either see before capturing, or immediately afterwards, what the final image will look like. For stacks I can't. It is only when I do the stacking and find out how the transition falls that the look of the image emerges, for better or worse.

And sometimes a stack that looks ok in terms of composition is spoilt by artefacts that either I can't get rid of or that need more time to fix than I'm prepared to spend on it. When I started out with stacking I would spend ages on an individual image, but now I've got a much better grasp of what will work and what won't, and I don't spend long on any single stack. Once it starts getting time-consuming I drop it and move on.

Sometimes I can't find a single-capture image or a stack of a subject that I like.

I've written at some length about using a combination of single-capture images and stacks for botanical subjects. You might want to have a look at this post here at dpreview, which is quite short, and mainly images. If you are interested you might want to click through from there to a more substantial set of posts on another site with a lot more discussion of various issues to do with stacking, and not stacking.

Nick,

This appears the stacking is also improving the effective resolution as well as creating a pleasing background. The stacks were using  f2.8, while the singles were f8 or f11 which explains the apparent resolution improvement. Agree, I like the stacked images better, but all are very good.

When imaging fast moving subjects like race cars, blurring the background often gives a sense of speed. But generally you don't want the subject to be blurred, so panning with the moving subject keeps the subject sharp but burs the background during the exposure. Your interesting use of stacking seems an analogy in that you are moving the lens or focus, effectively sweeping the background while the subject is sharply in focus.

Very interesting concept indeed, thanks for the reply.

Best,

John K wrote:

gardenersassistant wrote:

Here are a pair of images captured under more controlled conditions. I set down a line of tape on a table and mounted a light coloured plastic rule at an angle to the tape and a metal rule perpendicular to the tape, as shown below.

I captured an image with two cameras, placing each camera on the table and using the line of the tape to help me align each camera so it was as best as I could make it head on to the metal rule and therefore at the same angle to the plastic rule. The vertical angle on the scene varied a bit as the cameras and lenses were of different sizes and shapes and needed to be differently supported so as to be able to frame both rules.

Here is an uncropped image captured with a Panasonic FZ330 camera, which has a 1/2.3" sensor, with the f-number set to its maximum f-number of f/8, and with a Raynox 250 on the camera lens.

Here is an uncropped image with the same scene width of 12mm captured with a full frame Sony A7ii with a Meike 60mm macro lens with a 2X teleconverter, a 1.4X teleconverter and 68mm extension tubes. The f-number was set to f/22.

The image captured with the small sensor FZ330 using its maximum f-number of f/8 (and therefore giving the maximum achievable depth of field for that setup) has less depth of field than the image captured with the full frame sensor A7ii.

OK, that's fine. But the smaller sensor isn't causing the drop in depth of field, the limited maximum Fstop is.

Correct.

Why was it necessary to add two teleconverters and extension tubes to the 60mm macro for the full frame shot? Did it really take all of that kit just to get to 3x (36mm wide sensor / 12mm)? Seems like just the TCs would have taken that lens to 3.4x.

The 60mm macro was not at 1:1. The idea here was to illustrate that a full frame with extensions could have more depth of field than a small sensor camera using maximum f-number. I added all that extension so as to get a clearly visible increase in depth of field.

With the 60mm at 1:1 it looked like this, scene width around 5.5mm, around 6.5:1.

I used less magnification than this so we could see more of the mm marks on the inclined rule so as to make it easier to compare the depths of field.

If I'm getting the specs right the 1 2/3" sensor is 6.17mm wide, so that shot is about 1/2 life size.

Correct.

And around 3:1 on the full frame, which exemplifies the reason I have reservations about the 1:1 to 1:10 definition of macro. It means that the first of the inclined ruler images above is not a macro and the second one is, even though both cover (roughly, because of the different aspect ratios) the same field of view.

To avoid the confusion (and around here at least, the arguments) that can arise over the 1:1 definition I tend to talk in terms of scene widths when comparing kit, and tend to refer to my images as close-ups, even when they are of rather small subjects like springtails and barkflies. (Unlike the problems that can arise when using the term "macro", no one has ever objected to my use of "close-ups" for images they would define as macros.)

It depends on the context. Around here you sometimes have to be rather precise about things in order to (try to) avoid contention. In other places I find it works fine to talk about my invertebrate images as macros in contradistinction to my botanical images which I refer to as close-ups, even though most of my small sensor invertebrate images are not macros according to the 1:1 definition.

mawyatt2002 wrote:

gardenersassistant wrote:

mawyatt2002 wrote:

gardenersassistant wrote:

I use stacks not for the detail you can see if you zoom in, because I produce images for viewing "as is" without zooming in on them. I use stacking because I like the look of the images I can produce using it.

Nick,

Interesting comment. Can you elaborate more on the "look" stacking gives? Is this the in focus areas or the out of focus areas, bohea so to speak?

Best,

Both the in focus areas and the out of focus areas.

One of the reasons I use aperture bracketing (for non-stacked images) is that it gives me a set of images from which I can choose the one(s) which have the balance between the amount of the subject that is in focus and the rendition of the background that best pleases my eye.

For a lot of people a highly out of focus background is very pleasing, and some prefer more or less, or completely, featureless backgrounds, for example replacing a natural background with an artificial one by placing a card behind the subject. I generally prefer more going on in my backgrounds than that, and I prefer natural backgrounds, and I spend time exploring angles of attack looking for ones which produce a background that seems to me to be more or less harmonious with the subject in terms of lines, shapes, textures and/or illumination. And with single images you tend to get a smooth transition between the in focus and out of focus areas (unless the entire subject is rendered in focus against a background that is relatively far away and out of focus; with the whole subject in focus and the whole background out of focus this is not problematic).

Sometimes I can find a single-capture image of a scene that gives a combination of in focus and out of focus areas, and a transition between them, that pleases my eye. Sometimes I can't.

Sometimes when I can't get a balance that I like between subject coverage and background rendition with a single-capture image, I can get it with a stacked image. In fact, I'm finding that that is quite often the case. I haven't counted, but I would guess that at present I'm choosing to use a stack for perhaps two thirds or so of the scenes and single-capture images for the rest.

Stacked images can have a different look from single-capture images, with a rather sudden transition between the in focus and out of focus areas. Sometimes this can look similar to the subject-against-a-distant-background look of a single capture image, but sometimes there can be a rather jarring transition, for example if a stem, branch, leaf or petal jumps from being in focus to out of focus. I try to pick the rearmost element of the stack such that the transition isn't too jarring, but sometimes I can't find a suitable break point. And one of the curious things about capturing images (or mainly videos in my case) for stacking is that I can't compose with the final image in mind. With single-image captures I can either see before capturing, or immediately afterwards, what the final image will look like. For stacks I can't. It is only when I do the stacking and find out how the transition falls that the look of the image emerges, for better or worse.

And sometimes a stack that looks ok in terms of composition is spoilt by artefacts that either I can't get rid of or that need more time to fix than I'm prepared to spend on it. When I started out with stacking I would spend ages on an individual image, but now I've got a much better grasp of what will work and what won't, and I don't spend long on any single stack. Once it starts getting time-consuming I drop it and move on.

Sometimes I can't find a single-capture image or a stack of a subject that I like.

I've written at some length about using a combination of single-capture images and stacks for botanical subjects. You might want to have a look at this post here at dpreview, which is quite short, and mainly images. If you are interested you might want to click through from there to a more substantial set of posts on another site with a lot more discussion of various issues to do with stacking, and not stacking.

Nick,

This appears the stacking is also improving the effective resolution as well as creating a pleasing background. The stacks were using f2.8, while the singles were f8 or f11 which explains the apparent resolution improvement.

Yes, f/2.8 is close to optimum for sharpness with that lens. The reason I don't use the slightly better f/4 or (arguably) very slightly better f/5.6 or f/8 for stacks is to keep the shutter speed up for hand-holding while trying to stay at or as near as possible to base ISO. (I use an auto-ISO setup that handles both aspects for me so I don't have to think about it and can concentrate on composition etc.)

I do wonder if there is something else going on as well. Presumably as the camera traverses the scene there are often several images in sequence which have a particular area in focus. I wonder if the stacking makes any use of that extra information? (similar to median stacking for noise reduction, done locally on focus-overlapped areas, but for sharpness/detail rather than noise).

Agree, I like the stacked images better, but all are very good.

Thanks.

When imaging fast moving subjects like race cars, blurring the background often gives a sense of speed. But generally you don't want the subject to be blurred, so panning with the moving subject keeps the subject sharp but burs the background during the exposure. Your interesting use of stacking seems an analogy in that you are moving the lens or focus, effectively sweeping the background while the subject is sharply in focus.

That's a novel way of looking at it. Not sure I can quite get my head around it as the lens does in fact focus right to the furthest background. I just dump those images from where the focus traversal moves beyond the furthest point I want in focus.

It is tempting to think of the background as being what is in the furthest frame that I keep, but it is more complicated than that. Depending on the method and parameters being used backgrounds can get mucked around with (noise, colour shifts, desaturation, posterisation, artefacts). Hmmmm, I'm wondering if you might ask for examples - but I don't recall seeing much of that recently. I wonder why not. Perhaps I'll play and see if I can generate some nastiness to remind myself about it.)

Very interesting concept indeed, thanks for the reply.

You are very welcome. I like rambling on about this stuff.

Best,

Jim B (MSP) wrote:

gardenersassistant wrote:

mawyatt2002 wrote:

gardenersassistant wrote:

I use stacks not for the detail you can see if you zoom in, because I produce images for viewing "as is" without zooming in on them. I use stacking because I like the look of the images I can produce using it.

Nick,

Interesting comment. Can you elaborate more on the "look" stacking gives? Is this the in focus areas or the out of focus areas, bohea so to speak?

Best,

Both the in focus areas and the out of focus areas.

One of the reasons I use aperture bracketing (for non-stacked images) is that it gives me a set of images from which I can choose the one(s) which have the balance between the amount of the subject that is in focus and the rendition of the background that best pleases my eye.

I need to experiment more in this area. We sometimes forget that we want an image that is pleasing overall.

I've felt for quite a while now that "making pretty pictures" is what appeals to me most about photography. Well, pretty to my eye anyway. It's a rather inconsequential aim but it keeps me amused.

....

Sometimes I can find a single-capture image of a scene that gives a combination of in focus and out of focus areas, and a transition between them, that pleases my eye. Sometimes I can't.

......

Stacked images can have a different look from single-capture images, with a rather sudden transition between the in focus and out of focus areas.

I have experimenting with hand held stacked images. I generally find that Photoshop does a better job of stacking than does Helicon, as PS better accounts for my shake perpendicular to the focus plane.

Interesting. I've never tried PS for focus stacking. I'm periodically surprised by the amount of movement that Helicon can tolerate. Halos are what get in my way most.

.....

I've written at some length about using a combination of single-capture images and stacks for botanical subjects. You might want to have a look at this post here at dpreview, which is quite short, and mainly images.

I had forgotten about this post. This is an excellent comparison.

Thanks.

....

gardenersassistant wrote:

Pixel Pooper wrote:

With the same field of view, and the same entrance pupil diameter, you get the same DOF (and the same diffraction) regardless of magnification or sensor size.

I am a bit puzzled by this. Perhaps you can correct my thinking here please.

You got me. The statement above is true at normal distances, but since we are talking about macro I should have said exit pupil rather than entrance pupil.

Below 1:10 magnification we use the f/stop which depends on the entrance pupil to calculate DOF, but above 1:10 magnification we have to use the effective f/stop which depends on the exit pupil.

Suppose I photograph a scene at 1:1 with a full frame camera using a 100mm macro lens set to f/8. (To avoid a complication later, let's assume the f-number is set physically on the lens.) The aperture is focal length / f-number, which is 12.5mm.

Suppose that I now add a 2X teleconverter, without changing the f-number setting on the lens, and move away from the subject so as to get the same field of view as before, and hence the same magnification of 1:1. So the field of view is the same as before, and the entrance pupil diameter is the same because I haven't changed the f-number setting. But isn't the depth of field different now, roughly twice what it was before because the effective f-number is two stops larger, and the effective aperture two stops smaller? If that is so, depth of field does not depend solely on field of view and entrance pupil diameter.

In this example the field of view is not the same. The magnification at the focal plane is the same, but the lens without the teleconverter has a wider angle of view so the background has a wider FOV. In this case my previous statement applies:

"For a given sensor size, DOF depends only on magnification and effective f/stop but this doesn't work across different formats."

gardenersassistant wrote:

To avoid the confusion (and around here at least, the arguments) that can arise over the 1:1 definition I tend to talk in terms of scene widths when comparing kit, and tend to refer to my images as close-ups, even when they are of rather small subjects like springtails and barkflies. (Unlike the problems that can arise when using the term "macro", no one has ever objected to my use of "close-ups" for images they would define as macros.)

It depends on the context. Around here you sometimes have to be rather precise about things in order to (try to) avoid contention. In other places I find it works fine to talk about my invertebrate images as macros in contradistinction to my botanical images which I refer to as close-ups, even though most of my small sensor invertebrate images are not macros according to the 1:1 definition.

I always stay with the actual definition of macro: Projecting a 1:1 scale image of the subject onto the sensor. You hit on some of the reasons why it's important, but the main reason for me is the light. As the magnification increases the available surface area that's reflecting light back into the lens drops. So getting enough good light onto the subject becomes difficult. In addition the 1/focal length rule for hand holding a lens breaks down the closer the sensor is to the subject. I've gotten into too many arguments over light and macro motion blur with someone only to realize that they've never taken a photo that's even close to 1x.

Pixel Pooper wrote:

gardenersassistant wrote:

Pixel Pooper wrote:

With the same field of view, and the same entrance pupil diameter, you get the same DOF (and the same diffraction) regardless of magnification or sensor size.

I am a bit puzzled by this. Perhaps you can correct my thinking here please.

You got me. The statement above is true at normal distances, but since we are talking about macro I should have said exit pupil rather than entrance pupil.

Below 1:10 magnification we use the f/stop which depends on the entrance pupil to calculate DOF, but above 1:10 magnification we have to use the effective f/stop which depends on the exit pupil.

So that's how it works. That is useful to know that. Thanks.

Suppose I photograph a scene at 1:1 with a full frame camera using a 100mm macro lens set to f/8. (To avoid a complication later, let's assume the f-number is set physically on the lens.) The aperture is focal length / f-number, which is 12.5mm.

Suppose that I now add a 2X teleconverter, without changing the f-number setting on the lens, and move away from the subject so as to get the same field of view as before, and hence the same magnification of 1:1. So the field of view is the same as before, and the entrance pupil diameter is the same because I haven't changed the f-number setting. But isn't the depth of field different now, roughly twice what it was before because the effective f-number is two stops larger, and the effective aperture two stops smaller? If that is so, depth of field does not depend solely on field of view and entrance pupil diameter.

In this example the field of view is not the same. The magnification at the focal plane is the same, but the lens without the teleconverter has a wider angle of view so the background has a wider FOV.

I can see that the lens without the teleconverter has a wider angle of view, and that the background has a wider field of view. But I'm scratching my head a bit about the implications of this regarding the DOF. The field of view at the focal plane is the same, so Isn't the difference in DOF being caused solely by the difference in exit pupil rather than the difference in angle of view or the difference in field of view in the background?

In this case my previous statement applies:

"For a given sensor size, DOF depends only on magnification and effective f/stop

OK.

but this doesn't work across different formats."

gardenersassistant wrote:

I can see that the lens without the teleconverter has a wider angle of view, and that the background has a wider field of view. But I'm scratching my head a bit about the implications of this regarding the DOF. The field of view at the focal plane is the same, so Isn't the difference in DOF being caused solely by the difference in exit pupil rather than the difference in angle of view or the difference in field of view in the background?

Depth of field depends on the angle of the cones of light from the subject passing through the aperture.

To keep the same magnification with double the focal length, you must double the distance to the subject, so the light cone becomes narrower unless you also double the aperture diameter, which gives you the same f-number.

To keep the same magnification with the same angle of view, your subject distance is the same, so the light cone has the same angle at the same aperture diameter. With a smaller sensor, the same angle of view requires a shorter focal length, which gives you smaller f/number.

Pixel Pooper wrote:

gardenersassistant wrote:

I can see that the lens without the teleconverter has a wider angle of view, and that the background has a wider field of view. But I'm scratching my head a bit about the implications of this regarding the DOF. The field of view at the focal plane is the same, so Isn't the difference in DOF being caused solely by the difference in exit pupil rather than the difference in angle of view or the difference in field of view in the background?

Depth of field depends on the angle of the cones of light from the subject passing through the aperture.

Ah! Another little gem. Thank you.

To keep the same magnification with double the focal length, you must double the distance to the subject, so the light cone becomes narrower unless you also double the aperture diameter, which gives you the same f-number.

So if, with the same magnification, you double the distance to the subject, you halve the effective aperture diameter? (Or the exit pupil diameter?) Thus doubling the f-number?

To keep the same magnification with the same angle of view, your subject distance is the same, so the light cone has the same angle at the same aperture diameter. With a smaller sensor, the same angle of view requires a shorter focal length, which gives you smaller f/number.

So the effective aperture and the angle of view change in lockstep, so you can use either in assessing the amount of/changes in DOF?

gardenersassistant wrote:

Pixel Pooper wrote:

Depth of field depends on the angle of the cones of light from the subject passing through the aperture.

Ah! Another little gem. Thank you.

The picture below shows how the aperture size and distance from the sensor determines the angle of the light cone, and how this angle affects the size of the circle of confusion and thus the DOF.

To keep the same magnification with double the focal length, you must double the distance to the subject, so the light cone becomes narrower unless you also double the aperture diameter, which gives you the same f-number.

So if, with the same magnification, you double the distance to the subject, you halve the effective aperture diameter? (Or the exit pupil diameter?) Thus doubling the f-number?

There is also a corresponding cone of light between the aperture and the subject. This picture makes it easy to visualise how doubling the subject distance without increasing the aperture causes the cone to narrow.

To keep the same magnification with the same angle of view, your subject distance is the same, so the light cone has the same angle at the same aperture diameter. With a smaller sensor, the same angle of view requires a shorter focal length, which gives you smaller f/number.

So the effective aperture and the angle of view change in lockstep, so you can use either in assessing the amount of/changes in DOF?

At infinity focus the entrance and exit pupil create the same angle so we can use the f/number which is the focal length divided by the entrance pupil as shown in the upper diagram below.

When we focus closer the exit pupil moves away from the sensor causing the cone to narrow which reduces the effective aperture. Imagine the exit pupil moving away from the sensor in the lower picture below and you can see how the cone becomes narrower.

The f/number is really just a simple way of expressing the angle of the cone of light. It is this angle that determines DOF, diffraction, and exposure which is why the effective aperture is the true aperture, and why for the same DOF we get the same diffraction and the same total light.

Pixel Pooper wrote:

gardenersassistant wrote:

Pixel Pooper wrote:

Depth of field depends on the angle of the cones of light from the subject passing through the aperture.

Ah! Another little gem. Thank you.

The picture below shows how the aperture size and distance from the sensor determines the angle of the light cone, and how this angle affects the size of the circle of confusion and thus the DOF.

Brilliant! So simple. So obvious, seeing it like that. Thank you so much.

To keep the same magnification with double the focal length, you must double the distance to the subject, so the light cone becomes narrower unless you also double the aperture diameter, which gives you the same f-number.

So if, with the same magnification, you double the distance to the subject, you halve the effective aperture diameter? (Or the exit pupil diameter?) Thus doubling the f-number?

There is also a corresponding cone of light between the aperture and the subject. This picture makes it easy to visualise how doubling the subject distance without increasing the aperture causes the cone to narrow.

To keep the same magnification with the same angle of view, your subject distance is the same, so the light cone has the same angle at the same aperture diameter. With a smaller sensor, the same angle of view requires a shorter focal length, which gives you smaller f/number.

So the effective aperture and the angle of view change in lockstep, so you can use either in assessing the amount of/changes in DOF?

At infinity focus the entrance and exit pupil create the same angle so we can use the f/number which is the focal length divided by the entrance pupil as shown in the upper diagram below.

When we focus closer the exit pupil moves away from the sensor causing the cone to narrow which reduces the effective aperture. Imagine the exit pupil moving away from the sensor in the lower picture below and you can see how the cone becomes narrower.

The f/number is really just a simple way of expressing the angle of the cone of light. It is this angle that determines DOF, diffraction, and exposure which is why the effective aperture is the true aperture, and why for the same DOF we get the same diffraction and the same total light.

That is all very informative. You have been very helpful indeed. Thank you!

Pixel Pooper wrote:

gardenersassistant wrote:

Pixel Pooper wrote:

Depth of field depends on the angle of the cones of light from the subject passing through the aperture.

Ah! Another little gem. Thank you.

The picture below shows how the aperture size and distance from the sensor determines the angle of the light cone, and how this angle affects the size of the circle of confusion and thus the DOF.

To keep the same magnification with double the focal length, you must double the distance to the subject, so the light cone becomes narrower unless you also double the aperture diameter, which gives you the same f-number.

So if, with the same magnification, you double the distance to the subject, you halve the effective aperture diameter? (Or the exit pupil diameter?) Thus doubling the f-number?

There is also a corresponding cone of light between the aperture and the subject. This picture makes it easy to visualise how doubling the subject distance without increasing the aperture causes the cone to narrow.

To keep the same magnification with the same angle of view, your subject distance is the same, so the light cone has the same angle at the same aperture diameter. With a smaller sensor, the same angle of view requires a shorter focal length, which gives you smaller f/number.

So the effective aperture and the angle of view change in lockstep, so you can use either in assessing the amount of/changes in DOF?

At infinity focus the entrance and exit pupil create the same angle so we can use the f/number which is the focal length divided by the entrance pupil as shown in the upper diagram below.

When we focus closer the exit pupil moves away from the sensor causing the cone to narrow which reduces the effective aperture. Imagine the exit pupil moving away from the sensor in the lower picture below and you can see how the cone becomes narrower.

The f/number is really just a simple way of expressing the angle of the cone of light. It is this angle that determines DOF, diffraction, and exposure which is why the effective aperture is the true aperture, and why for the same DOF we get the same diffraction and the same total light.

Nice explanation. The illustrations really help with the understanding.

"The picture below shows how the aperture size and distance from the sensor determines the angle of the light cone, and how this angle affects the size of the circle of confusion and thus the DOF."

Just a note, don't think the Circle of Confusion is affected by the angle, it's just an arbitrary level of focus acceptance (and confusion) that folks have decided to use. In Fig 17 the CoC is constant but then cone angle is quite different, and thus the DoF is much different.

Thanks for posting.

Best,

mawyatt2002 wrote:

Just a note, don't think the Circle of Confusion is affected by the angle, it's just an arbitrary level of focus acceptance (and confusion) that folks have decided to use. In Fig 17 the CoC is constant but then cone angle is quite different, and thus the DoF is much different.

The circle of confusion is the blur circle created by the cone of light. The wider the angle of the cone, the larger the circle of confusion at any given distance from the focal plane.

Pixel Pooper wrote:

mawyatt2002 wrote:

Just a note, don't think the Circle of Confusion is affected by the angle, it's just an arbitrary level of focus acceptance (and confusion) that folks have decided to use. In Fig 17 the CoC is constant but then cone angle is quite different, and thus the DoF is much different.

The circle of confusion is the blur circle created by the cone of light. The wider the angle of the cone, the larger the circle of confusion at any given distance from the focal plane.

My point being that CoC is a value (~0.03mm for 36mm FF, ~0.018mm for APC )  determined by other factors, the cone creates a blur circle the CoC defines the acceptable limit of such.

From Wiki, "In photography, the circle of confusion (CoC) is used to determine the depth of field, the part of an image that is acceptably sharp. A standard value of CoC is often associated with each image format, but the most appropriate value depends on visual acuity, viewing conditions, and the amount of enlargement. Usages in context include maximum permissible circle of confusion, circle of confusion diameter limit, and the circle of confusion criterion."

Best,

mawyatt2002 wrote:

Pixel Pooper wrote:

mawyatt2002 wrote:

Just a note, don't think the Circle of Confusion is affected by the angle, it's just an arbitrary level of focus acceptance (and confusion) that folks have decided to use. In Fig 17 the CoC is constant but then cone angle is quite different, and thus the DoF is much different.

The circle of confusion is the blur circle created by the cone of light. The wider the angle of the cone, the larger the circle of confusion at any given distance from the focal plane.

My point being that CoC is a value (~0.03mm for 36mm FF, ~0.018mm for APC ) determined by other factors, the cone creates a blur circle the CoC defines the acceptable limit of such.

The circle of confusion is the blur circle created by the cone of light. The value you are taking about is the maximum permissable circle of confusion which is the diameter of the largest COC that is still perceived as a point in the displayed image.

From Wiki, "In photography, the circle of confusion (CoC) is used to determine the depth of field, the part of an image that is acceptably sharp. A standard value of CoC is often associated with each image format, but the most appropriate value depends on visual acuity, viewing conditions, and the amount of enlargement. Usages in context include maximum permissible circle of confusion, circle of confusion diameter limit, and the circle of confusion criterion."

Best,

The first paragraph of that Wikipedia page says:

"In optics, a circle of confusion is an optical spot caused by a cone of light rays from a lens not coming to a perfect focus when imaging a point source. It is also known as disk of confusion, circle of indistinctness, blur circle, or blur spot."