Craft Photography and DoF

[William W]

That are two different things: how it is and how you want it to be. I only try to tell you how it is with some arguments how I got to that conclusion. . .

Actually, George: No these are NOT how it is and how others want it to be.

Depth of Field is nothing more than an illusion anyway: and ever so quite far away from the purity of Mathematics upon which you have based your arguments to come to your conclusions.

We only use Mathematics to corral DoF into a suitable package. But Mathematics does NOT DEFINE DoF. DoF is elastic, not so much as Bokeh is elastic - but similar in some fashion.

DoF is steeped in history and premised on (amongst other things): 10 x 8 plates; a man's arm's length and the ability of the average person's eye.

There is little pure mathematical accuracy in those premises.

WW

[Mark von Kanel]

Ta.

For clarity - I did NOT thank you for "helpful post" because you believed me.

I marked you post as "helpful post" because your post articulated in a different way that this is a discussion primarily about the PRACTICAL applications of theory and NOT about theory, per se.

I thanked you for "helpful post" on behalf of George in the sincere hope that he comes to understand the difference.

WW

Bugger now i AM confused!

[george013]

George,


1. The measurement is simply taken from the marking of "phi" on the camera to the Subject.
2. Then the camera format is selected and (maybe) a suitable CoC for the camera format is put into to the calculator (usually the CoC agreed by convention or is 'suitable' to the individual Photographer as s/he has determined 'suitable').
3. Then the Aperture is put in to the calculator.
4. Then the Focal Length of the lens is put into the calculator
WW

Again things are how they are and not what you want them to be.

Now look at the lens formula. As soon the object distance get smaller as the focal distance the image distance gets negative, an imaginair image.And resulting in a negative DOF.

Would it be as you say that the calculator measures from the film plane, than this would happen at a distance of 2 times the focal distance.

Pure theory. No camera can focus at those distances.

And of course in practice you don't care. But for the question asked it does.

George

[casper]

Mark, Pierre, George, John (response up to this point). Thank you all for taking the time to assist. I have revisited Mark's "Bees" and "DoF" threads, to which you all contributed, and found that very helpful - should have done that first. Pierre, I didn't understand the effects of wide angle and tilt shift at first, but the excellent links provided by John included an explanation of how these can be used. George, I had measured the "calculated" 0.75m from front of lens to centre of required field (not accounting for perspective, which comes up later) and Kuso gave me focus distance of 0.79m, which puts the origin further back as you suggest, so that has given me a good start. John, very grateful for the detailed reply, I have read all of the links - digesting it all may take a little longer. Hopefully, my understanding will increase progressively as I work my way down the thread. Regards, Noel

[casper]

The front nodal-point is nearly the optical center point. The distance scales on the lens, that's something you can't work with. http://en.wikipedia.org/wiki/Nodal_p...nts_and_planes
The dof-calculators are all based on some formules, take a look http://en.wikipedia.org/wiki/Depth_of_field somewher in the middle.
Did you understand what I meant with the negative DOF?

@Casper,


I'm not sure what you measured, but from this point of view it should be something top-front to bottom-back.

George

George, I hadn't considered the camera angle, but should have - I just measured point to point from lens to centre of subject. The camera was not a great deal higher, but I had taken an incorrect measurement. I was only concerned about the front to back focus of the glass rim and stickpins, as the intention was to crop around this section later, just below the rim of the glass. Thanks for pointing this out. Regards, Noel

[casper]

Ignoring for a moment the: questions; answers; and the few incorrect uses of technical terms about DoF and DoF Calculators and Calculations . . .

If you want to attain a GREATER DoF across the OBJECT for the general shooting scenario that you have described and illustrated, then here are four practical methods you might consider:

1. Keep the same CAMERA VIEWPOINT but use a WIDER LENS and then crop in post production to attain the FRAMING that you require.
This will keep the PERSPECTIVE the same as the image that you have in your illustration.
Rough guess, based upon experience – try about 18 ~20mm Lens as a starting point. (Kit Zoom lens would be good if you have one)


2. Use the same 50mm lens, but move the CAMERA VIEWPOINT farther away from the OBJECT and then crop in post production to attain the FRAMING that you require.
This will CHANGE the PERSPECTIVE from the image that you have in your illustration.


3. Use a combination of suggestions 1 & 2 above.
This is to preserve the PERSPECTIVE as close to the illustarion as much as possible by NOT moving the CAMERA VIEWPOINT back too much, from the OBJECT.
Rough guess, based upon experience, move back a bit and use about a 30mm lens, a 35 prime if you have one or the Kit Zoom Lens at about 30mm.


4. Use a Tilt Shift Lens. For the size OBJECT the TS-E 45 or the TS-E 24 would be suitable, the TS-E 45 would be my first choice.
However to attain the full benefits of using a TS-E lens you should also use a 135 format camera (aka ‘full frame’).
Note this option is expensive.

*

Here is a simple example of how moving the CAMERA VIEWPOINT farther away, and then cropping in Post Production back benefits a greater DoF for photographing a small OBJECT and you can see how little area the OBJECT occupies of the original framing in the camera:



WW

hi Bill, thanks for helping out. In my example, I think I could use 1. 2. or 3. 2. seems easiest as I didn't have anything in front or behind the subject to worry about, e.g. a background that I wanted OOF - I only needed to make the depth of field comfortably larger than the subject. I have a 12-24mm (+1.6 crop factor) that may work for option 1. or 3. Regards, Noel

[casper]

It surprise me that nobody is reading my arguments. A main formula is the lens formula.
1/focal distance= 1/object distance + 1/image distance
The CoC gives you a minimal and a maximal image distance. By calculating the according object distances and subtracting them, you get the DOF.
Now look at the lens formula. As soon the object distance get smaller as the focal distance the image distance gets negative, an imaginair image.And resulting in a negative DOF.

Would it be as you say that the calculator measures from the film plane, than this would happen at a distance of 2 times the focal distance.

Pure theory. No camera can focus at those distances.

By using a wider angel lens and then crop the image, you should also use another CoC. I think the result will be the same.

Casper didn't react anymore, so I am guessing.

I think that should be more.

George

hi George,
I was getting there. I was probably sleeping and working during most of the exchange above. Subject (glass and contents) was 0.075m in depth, but DoF could be anything greater than that - I was only limited by my uneducated choice of lens. Thanks for your input, I appreciate it. Regards, Noel

[casper]

Hi Noel,

From the information you have given you state the subject distance was 75mm 'front to back'. The question arises as to how you determined this, whether on the true horizontal plane OR the subject on the same plane as your camera which in the photos clearly shows to be angled down.

If you measured the 75mm on the horizontal plane the subject distance 'front to back' is going to be greater with your camera set as it is.

Grahame

hi Grahame,
Guilty. I measured the depth of the subject on the horizontal, and the depth to the subject from the angled lens. Thanks for putting me straight. Regards, Noel

[casper]

Thank you everyone. I appreciate all of your efforts to help with my query. At my current level, the generic responses were quite sufficient, but when that all becomes way too easy I can refer back to the more advanced and robust viewpoints put forward here. Regards, Noel.

[William W]

Again things are how they are and not what you want them to be . . . .And of course in practice you don't care. But for the question asked it does. George

George,

Not so – it is not how things are and not what I want them to be.

Rather we are discussing the world of Practical Photography and not the world of Theoretical Mathematics.

***

Now on the issue of personal comment that you have made -

I do care and that is specifically why I took the time and effort to systematically respond to your direct conversation to me.

Frankly, it occurs simply impolite and arrogant of you to even begin to assume that you know either my motives, or my raison d'être.

William

[george013]

When I say "and of course in practice you don't care" in the sense I used it, I mean "nobody cares". I didn't mean anything personal.

George

[William W]

When I say "and of course in practice you don't care" in the sense I used it, I mean "nobody cares". I didn't mean anything personal.

George

OK. No problem.

Then it was just a misunderstanding based upon the words that you chose to use.

Specifically that you chose to use the personal pronoun "you" and you related "you" to the quote of my words specifically.

Thank you for clearing up you intent.

WW

[DanK]

Well, I use mathematics to pay my bills, so I feel a bit odd saying this, but there is a much simpler practical answer to the question that avoids the entire argument posted here, and adds a fifth option to Bill's four: Casper, simply stack your images. You can do this with quite a number of different software packages, including Photoshop, but if you are doing to do it often, it is worth buying Zerene. It then becomes trivially easy to obtain any DOF you want. You simply take a series of images, starting focused on the front and moving back, and then tell the software to stack them. In many cases, it is as simple as that. In complex images, more work is sometimes required--particularly if there is a large distance front-to-back between an edge of one surface and the next surface behind it. Even then, it may be less trouble than what you are going through now. One advantage is that you don't need to crop at all, so you retain full detail. The image below is of a milkweed pod that was far too deep for a single shot. If I recall, it required no cleanup at all.

[ajohnw]

Mark, Pierre, George, John (response up to this point). Thank you all for taking the time to assist. I have revisited Mark's "Bees" and "DoF" threads, to which you all contributed, and found that very helpful - should have done that first. Pierre, I didn't understand the effects of wide angle and tilt shift at first, but the excellent links provided by John included an explanation of how these can be used. George, I had measured the "calculated" 0.75m from front of lens to centre of required field (not accounting for perspective, which comes up later) and Kuso gave me focus distance of 0.79m, which puts the origin further back as you suggest, so that has given me a good start. John, very grateful for the detailed reply, I have read all of the links - digesting it all may take a little longer. Hopefully, my understanding will increase progressively as I work my way down the thread. Regards, Noel

The important thing Noel is to realise that it's all an approximation and while the web might give the impression that it's a be all and end all and must be used precisely in real terms it's just a method of grasping principles. When used it's wise to close down another stop especially at shorter distances and also take care where the camera is actually focusing.

The point of focus is often more important. Focus on the front of something and the front region of acceptable focus isn't being used. A common problem with AF. When cameras were manually focused best practice was to focus onto the subject and then slightly into it. Given the aids in viewfinders at that time the into it aspect wasn't too difficult to do.

The other thing to realise is that the calculators assume a certain final size of image viewed from a certain distance. This was 10x8in and 10ins. It's more usually quoted at 12x8in these days due to the shape of a 35mm frame - full format sensor. If you double the size of this 12x8 to 24x16in the circled of confusion would have to be 1/2'd or the viewing distance increased to 20in. Conversely if the image was reduced to 6x4in then the circle of confusion could be doubled providing it's still viewed from 10in. It can also be double if a 10x8 is viewed from 20in.

These sizes and distances come from figures optical people have used for years assuming the viewer has good acute eyesight. Nothing wrong with them. The problem is the approximations used in photography. The other complication is what happens to images that have been captured on a digital sensor when we actually get to see them. No one goes there other than by pure guess work. I'd say the effects tend to be favourable. Feeling bored one day I had a look at what the sensor actually captures via 2 different makes of lens at the same focal length.

There are other and I feel often better methods of getting at depth of field at some settings. Both are mentioned in this thread

Depth of field 2/3 behind and 1/3 in front???

Bill's framing aspect in particular but his numbers are for a full frame camera. The chart I posted is still some what TBD for me but I can state that it does work when say a wide angle lens is placed close to the ground so that very near objects are in the view. I chose to over resolve the detail close up by a factor of 3. Some time or the other I will try 1 and 2. Odd thing about this method is that the nearer detail will be resolved to the same level where ever the camera is actually focuses - even if focused at infinity.

What I need to determine is what happens when the lower edge of the framing is part way through the the distance to which the camera is focused. Say the view showed an area 1/2 way before the point the camera is focused at and 1/2 past it. The chart suggests that the size of detail that will be resolved will be 1/2 of the size of what the calculation comes up with. On the other hand assuming that the resolution was at the calculated figure at "start of the view" wouldn't cause much of a problem. Just means stopping the lens down more based on the assumed size of the detail seen. This method uses essentially the same maths as the usual DOF but from a how well does it need to be resolved point of view. Makes a lot more sense to me and the maths can be done in the head. It can also be used to makes sure distant detail is a sharp as it can be.

The framing method Bill mentions in the post removes focal length from the complications. It just need a guesstimate of the actual size of the view where the camera is focussed. Ideal for zoom lenses. What the web really needs is not loads of hypertripe which it current has but one that goes into determining this in a manner suitable for all format sizes and final images sizes. The chart is more suitable for typical scene / landscape work though.

John
-