More about DoF -- Depth of Field

[Daniel Browning]

Originally Posted by Mick Haensler

Most of my still work is architecture, I've invested in a few good lenses which brings me to my dilemna. I'm torn between the D300 and D700. I would love to go to a full frame camera but don't know if I can justify the price. I'm leaning towards the 300 since I already have some lenses and can use the D50 as a backup. Thoughts anyone?

If you need wide angle, thin DOF, or better low light performance, spring for the D700. Otherwise the D300 is a better purchase.

[Mick Haensler]

Thanks Daniel, that's the kind of input I'm needing.

Mick Haensler
Higher Ground Media

[Milutin Labudovic]

Originally Posted by Daniel Browning

If you need wide angle, thin DOF, or better low light performance, spring for the D700. Otherwise the D300 is a better purchase.

unfortunately this is not quite true, in DOF matter.

DOF is only influenced by lenses, and you can fit full frame lenses to d300, cropping will occur in 1.5 factor (17mm lens will come to be 25.5mm), but DOF in that crop will be the same since the lens is projecting the picture to the sensor
kind like you take picture with d3 or d700, import it to photoshop and crop it :)

so if DOF is what you after, it will not influence you in d300 if you fit it with FF lens.

[Dan Chung]

Mick,

Just noticed you shoot architecture, if you do this in the traditional way (ie not artsy abstracts) then a D700 is a far better choice than a D300. Why? because you can get ultrawides that are well corrected for the FX sensor that you can't get for DX crop cameras. Also the widest tilt shift lens Nikon does is 24mm which is not so handy on a D300 for buildings, if you ever intend to use one of these then go with the D700.

Dan

[Daniel Browning]

Originally Posted by Milutin Labudovic

unfortunately this is not quite true, in DOF matter.

DOF is only influenced by lenses, and you can fit full frame lenses to d300, cropping will occur in 1.5 factor (17mm lens will come to be 25.5mm), but DOF in that crop will be the same since the lens is projecting the picture to the sensor
kind like you take picture with d3 or d700, import it to photoshop and crop it :)

so if DOF is what you after, it will not influence you in d300 if you fit it with FF lens.

I think that is misleading because you're talking about two totally different compositions. It's not really helpful to say "throw away half your composition to get the same DOF." Yes, a DX headshot will have the same thin DOF as an FX head-and-shoulders. But they're not the same composition, so it doesn't make any sense to consider them equal in the real world. Rather, to get the same head-and-shoulders with a DX, the photographer backs up, which changes perspective (and isn't possible or desirable with some subjects, such as the moon, unless you have a rocket ship). By backing up, the focus distance becomes longer, which makes the DOF deeper. This is what gives crop formats the (deserved) reputation for having deeper DOF than FX.

It is more illuminating to think about things the following way:

For any given composition (perspective, field of view, and focus distance), and focal ratio, the D700 will have thinner DOF.

Or, an even simpler and broader version: For any given composition, the lens with a wider aperture will have thinner DOF.

The reason larger formats tend to have shallower DOF is because they require longer focal lengths to get the same field of view. The longer focal length results in a wider aperture for a given f/stop. The wider aperture is what causes thinner DOF. So, the camera format doesn't really have anything to do with it except to determine what focal length is needed to get the desired field of view.

So the D700 *does* allow thinner DOF.

[Milutin Labudovic]

Originally Posted by Daniel Browning

I think that is misleading because you're talking about two totally different compositions. It's not really helpful to say "throw away half your composition to get the same DOF." .

i might of bin misleading in the `whole` picture, and i apologies because of that. it was not intention :)
i agree that if you have a FF camera the range of lenses (tilt-shift...) can bring you different visual perspective when in 1:1 ratio is worth investing.
the price difference is around 1000$.
it would be helpful to know what kind of lenses Mick Haensler already has.
and d50 is a small d300, so in a way new d90 would be better than d300 if remained in DX.

[Jeff Donald]

Originally Posted by Daniel Browning

I think that is misleading because you're talking about two totally different compositions. It's not really helpful to say "throw away half your composition to get the same DOF." Yes, a DX headshot will have the same thin DOF as an FX head-and-shoulders. But they're not the same composition, so it doesn't make any sense to consider them equal in the real world. Rather, to get the same head-and-shoulders with a DX, the photographer backs up, which changes perspective (and isn't possible or desirable with some subjects, such as the moon, unless you have a rocket ship). By backing up, the focus distance becomes longer, which makes the DOF deeper. This is what gives crop formats the (deserved) reputation for having deeper DOF than FX.

It is more illuminating to think about things the following way:

For any given composition (perspective, field of view, and focus distance), and focal ratio, the D700 will have thinner DOF.

Or, an even simpler and broader version: For any given composition, the lens with a wider aperture will have thinner DOF.

The reason larger formats tend to have shallower DOF is because they require longer focal lengths to get the same field of view. The longer focal length results in a wider aperture for a given f/stop. The wider aperture is what causes thinner DOF. So, the camera format doesn't really have anything to do with it except to determine what focal length is needed to get the desired field of view.

So the D700 *does* allow thinner DOF.

That's not really correct. Please read this thread http://www.dvinfo.net/conf/open-dv-d...ld-skinny.html and take a look at this web site Depth of field

[Daniel Browning]

Originally Posted by Jeff Donald

That's not really correct.

Thanks for the correction, Jeff. I've read those references before, but I'm sure I missed something. Would you be more specific about what, exactly, I got wrong? I would like to understand DOF correctly.

[Jeff Donald]

Read the first link and you'll see what the factors are for DOF. You list perspective, field of view and distance to subject. The formula does not list perspective or field of view. Field of view is a fixed number and does not change. You may be thinking focal length of the lens, which is a factor. Perspective changes by moving the camera, which could be thought of as distance to subject, but you already used it. Van Walree's site shows excellent photos which illustrate how background blur stays the same etc. with different lenses. Please read the articles and if you still have questions post back.

[Daniel Browning]

Thanks for the response, Jeff. I will try to explain my understanding as clear as possible. In my haste to simplify DOF I may have made important mistakes. I would really appreciate it if you will, again, explain where I went wrong, so I can improve my knowledge in this area.

The context was a discussion about which sensor format is capable of thinner DOF: DX (APS-C, D300) or FX (FF, D700). My central point was the format with the larger sensor was capable of thinner DOF, because it requires a longer focal length in order to keep perspective and field of view the same.

I stated it this way:

Originally Posted by Daniel Browning

For any given composition (perspective, field of view, and focus distance), and focal ratio, the D700 will have thinner DOF.

I think it was the main point you replied to.

Originally Posted by Jeff Donald

The formula does not list perspective or field of view.

My understanding is that field of view is included in the formula in the form of "focal length" (which must change for a larger sensor to keep FOV constant). Perspective is included in the formula as "distance to the object being focused on".

Field of view is a fixed number and does not change.

Field of view does change if you keep focal length constant and vary the sensor size. This is important because it forms the central point of the discussion you replied to.

You may be thinking focal length of the lens, which is a factor.

To have the same field of view as a smaller format, you *have* to change the focal length of the lens on the larger format. That is exactly why the FX has thinner DOF, according to my understanding.

The reason I said "field of view" is because it automatically encodes the relationship between focal length, field of view, perspective, and sensor size.

Perspective changes by moving the camera, which could be thought of as distance to subject, but you already used it.

The reason I listed perspective and focus distance separately is that someone might not focus on their subject. The subject may be out of focus on purpose, such as the start of some "rack focus" shots.

Van Walree's site shows excellent photos which illustrate how background blur stays the same etc. with different lenses.

Right, but the context of the discussion is different-size sensors (DX vs FX).

So my biggest verbal shortcut is this: For any given perspective and field of view, the lens with a wider aperture (not focal ratio) will have thinner DOF.

Here's how I came up with that, please correct me if I'm wrong:

• A given perspective means subject distance and focus distance are constant.
• A given field of view means that a change in sensor size (if any) is automatically compensated by a change in focal length.
• "Wider aperture" can be attained by focal ratio, focal length, or both.

For constant sensor size, focal ratio is the only way to increase aperture for a given perspective/FOV.

For variable sensor size, the larger sensor will require longer focal lengths, which results in wider apertures if the focal ratio is the same. But focal ratio is not always the same in lenses of different formats, and the smaller format can even have wider apertures.

For example, 35mm is capable of thinner DOF than Medium Format, despite having much smaller sensors. The reason is that no one makes medium format lenses with apertures (not focal ratio) as wide as 35mm. 35mm lenses have lenses such as 135mm f/2.0, 85mm f/1.2, and 50mm f/1.2 in wide availability. For the same field of view, Medium Format has 210mm f/4.0, 135 f/2.8, and 80mm f/2.0. Not to mention the cost, which is triple or quadruple for similar filed of view and the closest focal ratio.

I applied the same method to conclude that FX (35mm format) is capable of thinner DOF than DX (APS-C).

I'm looking forward to your comments.

[Jeff Donald]

i wish you would have read the links I posted.

1) Focal length can change, such as by changing lenses, or zooming a lens. Focal length is generally measured in millimeters and in the formula for DOF, it is fixed. Meaning you can't have two different focal lengths in the same formula.

2) Field of view is a fixed measurement related to the focal length of the lens. So if you want to measure the DOF of two different chips (sensors) the number of variables need to be eliminated so you are only measuring the differences of the chip. Hence, the lens stays the same (focal length) and accordingly the field of view.

3) DOF is the discussion of the "in focus" attributes of an image. A rack focus shot may or may not have parts of the image in focus and are irrelevant to a discussion of DOF (the in focus parts of an image).

4) Perspective only changes when the position of the camera changes. In discussions of DOF changing the camera (distance to subject) will change the perspective. So, if you want to compare the DOF of two different size sensors, several variables must be eliminated. The position of the camera must not change, the focal length of the lens must not change and the aperture (F/number) must not change.

5) F/number is a part of the DOF formula. The term Focal Ratio, which you seem to prefer, means the same as F/number, aperture etc. in any standard photographic reference. Basically, the f/number is derived by dividing the size of the entrance pupil by the focal length of the lens. In other words a 50mm lens with a 50 millimeter opening would have an f/number of f/1. A 50mm lens with an entrance pupil of 25mm would have an f/number of f/2

[Daniel Browning]

Originally Posted by Jeff Donald

i wish you would have read the links I posted.

As I stated before, I have read them, several times. I didn't understand some of it and I'm sure that I have much more to learn about the topic, so I appreciate your help.

1) Focal length can change, such as by changing lenses, or zooming a lens. Focal length is generally measured in millimeters and in the formula for DOF, it is fixed.

I must have made a very ignorant mistake for you to think I don't already understand such rudimentary facts.

Meaning you can't have two different focal lengths in the same formula.

That's also quite obvious, but the part that isn't obvious to me is where I made the mistake to include two different focal lengths in the same formula. My understanding is that I used focal length correctly.

2) Field of view is a fixed measurement related to the focal length of the lens. So if you want to measure the DOF of two different chips (sensors) the number of variables need to be eliminated so you are only measuring the differences of the chip. Hence, the lens stays the same (focal length) and accordingly the field of view.

But the field of view does not stay the same. Larger chips will have a larger field of view. The discussion you replied to was about the D700 and D300. Any lens that works on the D700 will have a narrower field of view when used on the D300.

3) DOF is the discussion of the "in focus" attributes of an image. A rack focus shot may or may not have parts of the image in focus and are irrelevant to a discussion of DOF (the in focus parts of an image).

You said it: it may have parts of the image in focus. Some rack focus shots will start with a long shot where the plane of focus that contains nothing of interest. The subject is intended to be the moving bokeh, sometimes an obvious subject (that's Joe, but is he crying or laughing?), other times less obvious (is that Joe or Jane walking toward the camera?), other times completely blurred (what is that white blob moving around?). In any case, the subject is not the plane of focus until the rack is started.

I've shot timelapse where there is a sharp in-focus object that is still, and behind it there is all sorts of out of focus subjects moving through the frame. My intention as the photographer is that the viewer's eye is drawn to the movement in the background, where one can make out the general subjects (cars, people, etc.) despite the blur, while still enjoying the movement of light, contrast, and Bokeh. In this case the subject is not the plane of focus.

Another example would be the "waking up in the hospital" type scene, where the focus is on the lights or cieling, but the doctor's faces are much closer and out of focus. The faces, which are the subject, are still discernable (to whatever degree intended by the photographer), but out of focus. Here again, the focus distance and subject distance are different. Using "subject distance" to calculate DOF would be incorrect, as the plane of focus is on the distant cieling, not the much closer faces.

These types of examples are the reason why I made the distinction between focus distance and subject distance. This is really a minor and salient point which wasn't central at all to my post.

4) Perspective only changes when the position of the camera changes. In discussions of DOF changing the camera (distance to subject) will change the perspective.

Here again I must have made another stupendous blunder to give you the impression I didn't already know these basic principles. I don't yet understand where the error is, though.

So, if you want to compare the DOF of two different size sensors, several variables must be eliminated. The position of the camera must not change, the focal length of the lens must not change and the aperture (F/number) must not change.

I think that is a very useless "test" because it bears no similarity to the real world. Especially in the case of this thread, that you replied to, where the discussion is about two specific cameras, the D700 and D300, and which ones is capable of thinner DOF. In the real world, you don't force yourself to use the same lens on both cameras, which would mean forcing yourself to shoot wider angles of view with the D700.

Specifically, if you shot a 35mm lens on the D300, you are not forced to continue using the 35mm on the D700, which is now a much, much wider angle of view (a totally different composition). Instead, one might switch to a 50mm lens, and continue composing shots at the same distance and with the same angle of view as the 35mm, but with a thinner DOF.

This is exactly why I said Milutin Labudovic's statement was useless in the real world: a DX headshot will have the same thin DOF as an FX head-and-shoulders. But they're not the same composition, so it doesn't make any sense to consider them equal in the real world. Photographers don't change the way the compose and shoot just because they're using a larger sensor.

In any case, if someone did carry out the test you describe, and both sensors had the same resolution/area, post processing, and output display, the DOF would be the same, but the field of view would vary. In the case of the D300/D700 (the subject of this thread), you'd have to account for the different OLPF, resolution/area, and processing.

That is why I think a much more useful test, when comparing the DOF capabilities of the D300/D700, is to vary the focal length so that field of view stays the same.

5) F/number is a part of the DOF formula.

Here is another plain fact. I don't realize yet where I contradicted it.

The term Focal Ratio, which you seem to prefer

Thank you for pointing that out. I will immediately switch from the term Focal Ratio and start using the more commonly accepted F/number.

The term Focal Ratio...means the same as F/number, aperture etc. in any standard photographic reference.

F/number can be used correctly when the focal length is unknown, aperture *cannot*, so they do not mean the same thing. For example, it is correct to say "300mm f/2.8 and 30mm f/2.8 have the same F/number". It is not correct to say "300mm f/2.8 and 30mm f/2.8 have the same aperture." The widely prevalent conflation of these terms (F/number and aperture) is the source of many misunderstandings. Many other fields, such as Astronomy, do not confuse these as commonly as in photography.

Aperture, by itself, is the diameter of the opening, which can only be known when the focal length of the lens is also known (and not just fixed, as in the case of F/number).

I used "aperture" correctly in my posts, and I even used "aperture (not focal ratio)" to make it more clear that I was not referring to the incorrect use of aperture.

I hope, again, that I have explained myself clearly. My intention is to make it easy for you to see where I am incorrect and be able to point out the flaw in my understanding.

With thanks,
--
Daniel

[Jeff Donald]

Originally Posted by Daniel Browning

I think that is a very useless "test" because it bears no similarity to the real world. Especially in the case of this thread, that you replied to, where the discussion is about two specific cameras, the D700 and D300, and which ones is capable of thinner DOF. In the real world, you don't force yourself to use the same lens on both cameras, which would mean forcing yourself to shoot wider angles of view with the D700.

Specifically, if you shot a 35mm lens on the D300, you are not forced to continue using the 35mm on the D700, which is now a much, much wider angle of view (a totally different composition). Instead, one might switch to a 50mm lens, and continue composing shots at the same distance and with the same angle of view as the 35mm, but with a thinner DOF.

This is exactly why I said Milutin Labudovic's statement was useless in the real world: a DX headshot will have the same thin DOF as an FX head-and-shoulders. But they're not the same composition, so it doesn't make any sense to consider them equal in the real world. Photographers don't change the way the compose and shoot just because they're using a larger sensor.

In any case, if someone did carry out the test you describe, and both sensors had the same resolution/area, post processing, and output display, the DOF would be the same, but the field of view would vary. In the case of the D300/D700 (the subject of this thread), you'd have to account for the different OLPF, resolution/area, and processing.

That is why I think a much more useful test, when comparing the DOF capabilities of the D300/D700, is to vary the focal length so that field of view stays the same.

OK, I think here is where the major confusion takes place. If you change a factor in the DOF formula (f/number, distance to subject, focal length of lens) the DOF will change, pretty simple. Distance to subject also seems to be an area of confusion as well. When discussing DOF the subject is always the object at which the lens is focused, not an object of interest the lens is not focused on. Thus your fascination with the OOF (Out of Focus) portions of the image are not relevant to discussions of DOF (in focus portions of the image).

If the statement is made that different size sensors have different DOF you must keep keep all variable the same (except the size of the sensors) and measure any difference in DOF. The DOF of different sensors is the same. The DOF changes when you choose to change the focal length of the lens to keep the subject the same size. An example of this when you want the subject to remain a fixed size in the finished print, for example a subjects face. If the variables remain constant, the enlargement of the image (subject face) from smaller chip will be greater. Thus to get the subjects face the same size on the prints from the two different chips, either the focal length of the lens must change or the distance to the subject must change (or a combination of both). The DOF of the "chips" don't change, the DOF changes as a result of changing a variable in the formula to keep the same subject size.

Is there a "real world" example for this? Yes, in wildlife photography I take a picture of a bird with two different camera bodies, one large sensor and the other small sensor. I shoot an image of the bird on each camera body. The distance to the subject stays the same (bird didn't move and I couldn't get closer because the bird might fly). I use the longest lens I have on each body (500mm tele). F/number is the same on each image as well. Thus, in this example all the variables in the DOF formula stay the same. I enlarge each file to the same size print, 8x10. The bird is smaller in the print from the larger sensor camera. So I enlarge the file from the camera with the large chip so the bird is the same size in each 8x10 print. The resultant prints have the same DOF. Van Walree has excellent examples of this on his site.

[Daniel Browning]

Originally Posted by Jeff Donald

When discussing DOF the subject is always the object at which the lens is focused, not an object of interest the lens is not focused on.

Point conceded.

If the statement is made that different size sensors have different DOF

I should have been more clear that the lens availability (and not sensor size) is what causes the FX format to be capable of thinner DOF. For any given field of view, there are FX lenses with apertures (not F/number) much wider than any equivalent field of view on DX, except the longest focal lengths that the photographer can afford, can stand to carry, or is available.

Essentially, FX enables the photographer to use lenses with wider apertures (not F/numbers). If they're already using the widest aperture (e.g. 500mm f/4) made for their desired field of view, then FX will not make them capable of thinner DOF. I still think it's fair to say that FX is capable of thinner DOF than DX.

If the statement is made that different size sensors have different DOF you must keep keep all variable the same (except the size of the sensors) and measure any difference in DOF. [...] The DOF of different sensors is the same.

Agreed.

On a different note, to compare entire camera systems, and not just sensor sizes, one would take into account the avialability of lenses. To make such a comparison of FX and DX I think it would be useful to let focal length vary as needed to keep field of view constant. That would tell you which system is capable of thinner DOF for a given field of view (e.g. wide angle, normal, telephoto, super tele) and you could evaluate other aspects as well, such as price.

...in wildlife photography...I use the longest lens I have on each body (500mm tele)...

Agreed. That's why I emphasized aperture (not F/number); FX is only capable of thinner DOF when used with wider apertures. It follows that wildlife is not affected, because aperture is already at the maximum.

In the same way, if you need the field of view of 200mm f/2 on DX, then it wouldn't give you any thinner DOF to get 300mm f/3.0 on FX, because they both have the same 100mm aperture (and DOF). You'd have to get a much heavier, more expensive, (and non-existant) 300mm f/2.0 to get thinner DOF for the same FOV/perspective.

Kind regards,
--
Daniel

[Jeff Donald]

Some basic optical definitions are in order

Aperture - The clear opening in the lens through which light passes to reach the film (chip). Its size determines the amount of light that the lens will transmit.

Effective Aperture - The diameter of the clear aperture in a lens, expressed as a dimension. It's size is determined by the iris diaphragm.

Relative Aperture - Ratio of the focal length to the diameter of the effective aperture. Traditionally written as a fraction "f/ " followed by a numerical value. Example: 12.5/25=f/2

Diaphragm Opening - The marked "f/number." In the currently used set of f/numbers, each represents an effective aperture which permits 1/2 the light passing capability of the proceeding figure or setting (f/1, f/1.4, f/2.0, f/4.0, f/5.6, f/8, f/11, f/16 etc.) To determine the relative values of two f/numbers square the figures and compare numbers (1.4 x 1.4 =1.96, 2 x 2=4 or twice (2x) the former.

Effective f/number - This is the corrected f/number that takes into consideration the increased lens to film distance, when the lens is focused on an object much closer than infinity, such as close-up or macro photography. This is dependent on both effective aperture and lens-film distance.

T-stops - The physical diaphragm setting (numbers) determined by the amount of light actually transmitted by the lens. Takes into consideration the loss of light in the lens system by absorption and reflection.

I'm not sure what purpose your differentiation between aperture and f/number serves. In the formula, the f/number is used to determine DOF. The effective aperture is the f/number and part of the DOF formula. What do you see as the difference? I think this might be where an area of confusion.

Angle of View - also referred to as Field of View. The angle formed by lines from the rear nodal point to the two opposite sides of the film (chip) located one focal length from the rear nodal point, with the lens focused at infinity.

Angle of Coverage - The angle formed by lines from the rear nodal point to the opposite sides of the circle of good definition. The maximum angle over which the lens is capable of forming an acceptable image. Not dependent on focusing distance or film size.

Some manufactures use the term Field of View Equivalency Factor (FOVEF). I think this may be where some confusion may come from. This takes into account the so-called crop factor of different size chips (or film formats). The idea comes from trying to account for different degrees of field of view between the same focal length lens on different chip size cameras. For example a 300mm lens on the 4/3 format would have a FOVEF half that of a 300mm lens on a full frame 35mm camera. But the DOF doesn't change. The formula uses the focal length of the lens not the angle of view. This can be demonstrated by shooting the same lens on different chip size camera bodies (such as my wildlife example above.)

In some real world situations, such as landscape photography, the DOF may be different for the same composition. A wider lens may be required to get the same composition on two different chip size cameras. This would result in a difference in DOF, not because of chip size, but a change in focal length of the lenses used. In the wildlife example used above, the DOF stays the same because the same lens, distance to subject and f/number are used.

So to recap, their are five factors or variables in the DOF formula. I listed them in my DOF post I linked to, but to aid in this discussion I will repost them here. These are the five factors that affect DOF.

Depth of Field (DoF) is dependent upon the following variations:
a. The focal length of the lens.

b. The diaphragm opening (f/number).

c. The distance from the lens to the object that is focused on.

d. The distance from which the image is viewed. This a perceived or relative apparent DOF.

e. The viewer's personal standard of the permissible degree of sharpness (or unsharpness). Sometimes referred to as Circle of Confusion (COC).