are the DOF and vignetting interdependent?

Started Sep 29, 2016 | Questions
Carl Mucks Regular Member • Posts: 199
are the DOF and vignetting interdependent?

I'm speculating here, but at first glance the loss of light from the peripheral portion of the glass should contribute both to a lens vignetting and also deeper DOF.

So here are the questions.

1) Do the lenses with high vignetting have proportionally deeper DOF at the same f-stops compared to those with less vignetting?

2) Is DOF different in the center and in the corners?

3) Is pixel vignetting (microlens vignetting) affecting the light loss as well as deeper DOF?

 Carl Mucks's gear list:Carl Mucks's gear list
Canon PowerShot Pro70
ANSWER:
This question has not been answered yet.
Tom Axford Veteran Member • Posts: 7,476
Re: are the DOF and vignetting interdependent?
1

Carl Mucks wrote:

I'm speculating here, but at first glance the loss of light from the peripheral portion of the glass should contribute both to a lens vignetting and also deeper DOF.

So here are the questions.

1) Do the lenses with high vignetting have proportionally deeper DOF at the same f-stops compared to those with less vignetting?

2) Is DOF different in the center and in the corners?

3) Is pixel vignetting (microlens vignetting) affecting the light loss as well as deeper DOF?

If you are being very precise, the answer to all three questions is yes.

However, I have never seen a DoF calculator that took the lens characteristics into account to that level of detail. Depth of field isn't something that needs to be determined very precisely because an image doesn't suddenly change from being tack sharp to being blurry when you move beyond the depth of field. It's a very gradual change in sharpness.

The effect of vignetting on DoF is not going to be significant in practice (except perhaps in extreme circumstances). It isn't worth worrying about. Of course, in principle, other lens aberrations will also have an effect on DoF and they should be taken into account in theory. In practice, they are generally ignored for the same reason.

(unknown member) Senior Member • Posts: 2,180
Re: are the DOF and vignetting interdependent?
5

Tom Axford wrote:

If you are being very precise, the answer to all three questions is yes.

However, I have never seen a DoF calculator that took the lens characteristics into account to that level of detail.

DoF calculators are all based on simple geometrical proofs that do not hold up for any real lens.

blog.retrorefractions.com/?p=4

Vignetting can have a profound impact on depth of field.  Take, for example, the 24L II.  The lens is f/1.5 but if we idealize it to f/1.4 for a moment, the ratios are nicer.

The 24L II has, at its peak, a bit over two stops of vignetting.  This literally means in the corner of the frame, it is an f/2.8 lens.  Most would agree that the depth of field between f/1.4 and f/2.8 is considerably different.

J A C S
J A C S Forum Pro • Posts: 17,558
Re: are the DOF and vignetting interdependent?
3

Yes to all, as already said above. Vignetting changes the shape of the blur, not just the size.

I have comparisons of my 22/2 on M (crop) vs. the 35L at f/3.2 on FF. The DOFs in the center are very close and very different in the corners (where field curvature could be potentially responsible, as well).

Jack Hogan Veteran Member • Posts: 7,581
Localized Contrast and Spatial resolution

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner?  If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

(unknown member) Senior Member • Posts: 2,180
Re: Localized Contrast and Spatial resolution

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

The region of interest is a red herring, as it does not affect the aperture.  Depth of field has many parameters, this thread is to examine only the impact of vignetting (aperture).

Tom Axford Veteran Member • Posts: 7,476
Re: Localized Contrast and Spatial resolution
1

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

However, it is more complex than that! The out-of-focus blur changes shape with vignetting. As one moves away from the centre of the image, the blur becomes less circular. This will produce some really complex effects in the depth of field (e.g. the depth of field may be greater for parallel lines oriented tangentially than it is for parallel lines oriented radially).

The region of interest is a red herring, as it does not affect the aperture. Depth of field has many parameters, this thread is to examine only the impact of vignetting (aperture).

Jack Hogan Veteran Member • Posts: 7,581
Re: Localized Contrast and Spatial resolution

Tom Axford wrote:

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

I still don't get why light fall-off of 2 stops at the corners is equivalent to increasing f-number by 2 stops as far as spatial resolution is concerned.  I would guess a lot less than that.

Jack

Tom Axford Veteran Member • Posts: 7,476
Re: Localized Contrast and Spatial resolution

Jack Hogan wrote:

Tom Axford wrote:

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

I still don't get why light fall-off of 2 stops at the corners is equivalent to increasing f-number by 2 stops as far as spatial resolution is concerned. I would guess a lot less than that.

Jack

I would be inclined to agree with you, Jack. The blur disc is not going to be simply half the diameter as it will no longer be circular. On average it may be something like three quarters the diameter, but I'm not sure how one should define the "average" in this context.

(unknown member) Senior Member • Posts: 2,180
Re: Localized Contrast and Spatial resolution

Tom Axford wrote:

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

However, it is more complex than that! The out-of-focus blur changes shape with vignetting. As one moves away from the centre of the image, the blur becomes less circular. This will produce some really complex effects in the depth of field (e.g. the depth of field may be greater for parallel lines oriented tangentially than it is for parallel lines oriented radially).

That is not the definition of vignetting.  It is not even the definition of Relative Illumination.  Vignetting means obstructing rays with an insufficient aperture.  If the aperture stop is reduced in size, it produces on-axis vignetting.

The region of interest is a red herring, as it does not affect the aperture. Depth of field has many parameters, this thread is to examine only the impact of vignetting (aperture).

Tom Axford Veteran Member • Posts: 7,476
Re: Localized Contrast and Spatial resolution
1

AiryDiscus wrote:

Tom Axford wrote:

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

However, it is more complex than that! The out-of-focus blur changes shape with vignetting. As one moves away from the centre of the image, the blur becomes less circular. This will produce some really complex effects in the depth of field (e.g. the depth of field may be greater for parallel lines oriented tangentially than it is for parallel lines oriented radially).

That is not the definition of vignetting. It is not even the definition of Relative Illumination. Vignetting means obstructing rays with an insufficient aperture. If the aperture stop is reduced in size, it produces on-axis vignetting.

I have never heard that definition of vignetting before! Have you any references for it?

Here is an extract from Wikipedia, which accords with all usage I have seen in textbooks and scientific papers:

(unknown member) Senior Member • Posts: 2,498
Re: are the DOF and vignetting interdependent?

Carl Mucks wrote:

I'm speculating here, but at first glance the loss of light from the peripheral portion of the glass should contribute both to a lens vignetting and also deeper DOF.

So here are the questions.

1) Do the lenses with high vignetting have proportionally deeper DOF at the same f-stops compared to those with less vignetting?

2) Is DOF different in the center and in the corners?

3) Is pixel vignetting (microlens vignetting) affecting the light loss as well as deeper DOF?

Are you mixing up vignetting with apodizing? Placing an apodizing filter in the lens plane to darken the periphery of the lens will change the DoF, but doesn't correspond to vignetting the image.

Microlens vignetting can limit the effective maximum aperture of large aperture lenses - the f-stop blues.

Joe

(unknown member) Senior Member • Posts: 2,180
Re: Localized Contrast and Spatial resolution
1

Tom Axford wrote:

AiryDiscus wrote:

Tom Axford wrote:

AiryDiscus wrote:

Jack Hogan wrote:

AiryDiscus wrote:

The 24L II has, at its peak, a bit over two stops of vignetting. This literally means in the corner of the frame, it is an f/2.8 lens.

I am trying to understand: why do two stops of vignetting mean two stops lower effective f-number in the corner? If the intensity in the field of view were uniformly lower by 2 stops, would that change the effective f-number?

Wouldn't the size and consequent uniformity of the region of interest (i.e. magnification) also have a lot to do with perceived spatial resolution, hence DOF?

Jack

If the lens elements' diameters were lowered enough to vignette on-axis by two stops, that is exactly the same as stopping the lens down to stops.

By definition, vignetting does not occur on-axis. The loss of light increases with the distance from the centre of the image. So the depth of field would increase with the distance from the centre.

However, it is more complex than that! The out-of-focus blur changes shape with vignetting. As one moves away from the centre of the image, the blur becomes less circular. This will produce some really complex effects in the depth of field (e.g. the depth of field may be greater for parallel lines oriented tangentially than it is for parallel lines oriented radially).

That is not the definition of vignetting. It is not even the definition of Relative Illumination. Vignetting means obstructing rays with an insufficient aperture. If the aperture stop is reduced in size, it produces on-axis vignetting.

I have never heard that definition of vignetting before! Have you any references for it?

Here is an extract from Wikipedia, which accords with all usage I have seen in textbooks and scientific papers:

The photographic definition of vignetting is not the "true" or "scientific" definition of vignetting.  In optical design circles, vignetting is a verb used to describe the blocking of rays.

Because of the cos^4 law, vignetting is not truly the reduction of an image's brightness.  What your wikipedia page describes is relative illumination.

OP Carl Mucks Regular Member • Posts: 199
Re: are the DOF and vignetting interdependent?

Joe Pineapples II wrote:

Carl Mucks wrote:

I'm speculating here, but at first glance the loss of light from the peripheral portion of the glass should contribute both to a lens vignetting and also deeper DOF.

So here are the questions.

1) Do the lenses with high vignetting have proportionally deeper DOF at the same f-stops compared to those with less vignetting?

2) Is DOF different in the center and in the corners?

3) Is pixel vignetting (microlens vignetting) affecting the light loss as well as deeper DOF?

Are you mixing up vignetting with apodizing?

No.

Placing an apodizing filter in the lens plane to darken the periphery of the lens will change the DoF, but doesn't correspond to vignetting the image.

Microlens vignetting can limit the effective maximum aperture of large aperture lenses - the f-stop blues.

I'm quite aware of DXO article, but their analysis is quite superficial, that's why I asked. Reviewers posted contradictory information on whether microlens vignetting affects only loss of light or DOF as well. Some say "yes", some say "no", but nobody posted any rationale.

I appreciate all answers that provided rationale here, though "yes" or "no" responses aren't really helpful, as I'm not taking a poll.

 Carl Mucks's gear list:Carl Mucks's gear list
Canon PowerShot Pro70
(unknown member) Senior Member • Posts: 2,498
Re: are the DOF and vignetting interdependent?

Carl Mucks wrote:

Joe Pineapples II wrote:

Carl Mucks wrote:

I'm speculating here, but at first glance the loss of light from the peripheral portion of the glass should contribute both to a lens vignetting and also deeper DOF.

So here are the questions.

1) Do the lenses with high vignetting have proportionally deeper DOF at the same f-stops compared to those with less vignetting?

2) Is DOF different in the center and in the corners?

3) Is pixel vignetting (microlens vignetting) affecting the light loss as well as deeper DOF?

Are you mixing up vignetting with apodizing?

No.

Placing an apodizing filter in the lens plane to darken the periphery of the lens will change the DoF, but doesn't correspond to vignetting the image.

Microlens vignetting can limit the effective maximum aperture of large aperture lenses - the f-stop blues.

I'm quite aware of DXO article, but their analysis is quite superficial, that's why I asked. Reviewers posted contradictory information on whether microlens vignetting affects only loss of light or DOF as well. Some say "yes", some say "no", but nobody posted any rationale.

I appreciate all answers that provided rationale here, though "yes" or "no" responses aren't really helpful, as I'm not taking a poll.

Loss of light due to attenuation of high-angle rays would certainly affect DoF as well. You can see that by drawing a ray diagram.

Joe

J A C S
J A C S Forum Pro • Posts: 17,558
Re: are the DOF and vignetting interdependent?
2

Here are comparisons taken from this review:

Sigma 24/1.4 @f/1.8

Nikon 24/1.8 @f/1.8

In the center, the blur radius of the slower lens (Nikon) is smaller but not by much. Near the borders, the radius shrinks but interestingly enough, the shape does not change much. The Sigma presumably vignettes less at f/1.8 and it preserves the size of the blur better near the borders but the shape changed more!

Another interesting observation, see the widget there, is that stopping down the Sigma does not change the blur near the borders much but it does change it in the center area. I have noticed this with my lenses. One can say that the Sigma is f/1.4 in the center but f/2 or so near the borders, and that is not restricted to the Sigma only.

dethis2 Regular Member • Posts: 373
Re: are the DOF and vignetting interdependent?
Keyboard shortcuts:
FForum MMy threads