Depth of Field markings on old M lenses with new M8 camera

[Finder]

Technically, the DOF markings are already too conservative for modern emulsions. the perceived DOF is narrower when using film with very fine grain, and some even recommend that you should carry over the markings one stop (i.e., the DOF for f/4 is really for f/2, etc.) when examining large prints.

Film resolution has nothing to do with depth of field. It is simply the size of the circle of confusion even if the circle of confusion is larger than what the film can resolve.

 

[jaapv]

DOF is NOT related to resolution - the ability to distingush between two points. DOF is related to the circle of confusion - the size of each point. When the circle appears too large, the human visual system thinks it is out of focus. Since the DOF scales are calulated based on a 35mm frame, they will not be "correct" for the smaller sensor. Whether they are accurate enough depends on the photographer as DOF is a subjective number based on viewing a standard size print at a relative viewing distance. DOF is always in relation to the human visual system and is not an absolute physical property.

It is related to the resolution of the human eye, whether one or two photosites in the eye are excited. Nothing else. Your brain translates that to in or out of focus. If a car approaches you from a long distance in the dark you see one lightspot that suddenly transforms into two headlights. That is the resolution limit of your eye and you could say the car has reached the limit of DOF. Oviously large structures will hit more sites. The eye is remarkebly like a sensor in that way.

 

[40oz]

they are talking about shooting a ruler, for instance. As you enlarge a print of a photo of the ruler, there will be edges that looked fine at 3x5 that now are clearly blurred(!) at 8x10.

But one could say as you increase the magnification of the loupe, DOF shrinks, for the same reason. In the case of a crop sensor in a digital camera, you are cutting out the center of a 50mm lens image, for example, and blowing it up so the depth of field with that lens will actually be smaller than it would on a 35mm film camera (image magnified, so edges will appear less sharp).

If you consider the depth of field markings on a lens to be relative to some standard, say film negative on 8x10 prints, then the DOF never changes. As your usage deviates from the standard, you have to expect the DOF markings lose some of their meaning.
***
apparently i type slow - the post i was replying to is lost somewhere up there.

 

[Finder]

I R Stoopiid. How does changing the print size change the depth of field. I mean, once the image is recorded there is no resolving anything that is not already there.... Right? Maybe I'm just having problems wrapping my brain around that concept.

Does my avatar look sharp? Enlarge it in photoshop by 400%. Does it still look sharp? The resolution of the image has not changed, but it is really fuzzy when big.

Now look at a picture on the wall. At a large distance, the entire picture may look sharp. Walk up to it. When you get closer the background my separate from the subject. The closer you get, the more you will notice the difference of the object at the plane of focus and the background.

Think of a picture as a collection of circles. As an example, imagine the circle at the plane of focus is 1mm; just outside that plane, the circles are 2mm; and far from that plane in the background, the circles are 3mm. Your visual system has an angluar resolution in that at a certain angle a circle will be so small it will look like a dot. Let us say at 9m, a 3mm circle looks like a dot. The image will look completely sharp. If I move to 6m, the 2mm circle looks like a dot, but the 3mm circle will be big eough where it is bigger than a point and so the background looks soft, but the area around the subject appears sharp. Move to 3m from the print and now only the 1mm circle looks sharp so only the object at the plane of focus looks sharp and everything else looks fuzzy. The 3mm circles look really fuzzy and the picture appears to now have a narrow DOF. (At 9m, it had a large DOF.)

That is it.

 

[Finder]

It is related to the resolution of the human eye, whether one or two photosites in the eye are excited. Nothing else. Your brain translates that to in or out of focus. If a car approaches you from a long distance in the dark you see one lightspot that suddenly transforms into two headlights. That is the resolution limit of your eye and you could say the car has reached the limit of DOF. Oviously large structures will hit more sites. The eye is remarkebly like a sensor in that way.

It is related to the ANGULAR resolution of the eye, not the RESOLVING POWER of the eye. It is simply if the eye sees a circle of confusion as a disk or a point.

 

[Gabriel M.A.]

Film resolution has nothing to do with depth of field. It is simply the size of the circle of confusion even if the circle of confusion is larger than what the film can resolve.

You are correct, film resolution has nothing to do with DOF. The ability to resolve beyond the circle of confusion and printing large, upon examination, seems to make a difference.

If you want to start a little argument on Hooewe's on first and Watt's on second we can go down that path.

 

[jaapv]

We are saying the same thing. Both depend on the "pixel" density of the eye. Of course that translates into a resolving angle, but that is already one step further.

 

[Finder]

We are saying the same thing. Both depend on the "pixel" density of the eye. Of course that translates into a resolving angle, but that is already one step further.

Well, we are talking about the same thing, but then again we are not. Here again, it is not a "pixel" density problem, but it is whether you see a disk (blur) or a point. (Your eye is always moving and rod and cone size (pixel pitch) is not what determines what you can see.) I know it sounds crazy, but the ability to resolve something and the perception of sharpness is not the same. So we agree on principle, but not on mechanics.

 

[Ben Z]

You changed two parameters, sensor size and focal length. The relationship is not linear.

I changed the sensor size because that's what we're talking about, different sensor size! I changed the focal length in order to keep the FOV constant. I needed the FOV constant because that actually has an application in practical use of these cameras for actually taking pictures. Keeping the focal length constant and only changing the sensor size might make a nice physics experiment but has little or no relevance to which sensor-size will give greater or lesser DOF for a given composition. That's what I think most people would be concerned with if they are actually shooting pictures. You've got a scene in front of you with a specific composition, requiring a particular FOV. That will require different focal length lenses depending on the sensor size. The smallest sensor will require the shortest focal length, and the DOF will be greatest.

 

[jaapv]

Most people will be using the lenses they already have on th M8, Ben, not 7.5 mm focal length or similar....

 

[Ben Z]

Yes, but if they are taking pictures not studying physics they will be comparing the DOF based on the composition not the focal length. Someone will compare the same shot taken with his M7 and his M8, same size print and find that the M8 shot has more DOF ...because to get that same shot with the M8 he had to use a shorter focal length lens. How the DOF of a 50mm lens compares from one sensor to another is really not relevant to actual photography since using a 50mm on 2 different sized sensors won't give you the same shot. Capiche?

 

[anselwannab]

I always thought "Circle of Confusion" would be a good name for a band. CoC for short.

Rock On!

I'd like to see a plot with distance on the x axis, circle of confusion on the y and I assume bell shaped(?) type curves for different F-stops. That way you could know that some CoC is needed for a certian sensor size and enlargement.

I assume if the x were fashioned correctly, you could just slide it back and forth and get the curves for any distance. Plus you could get an idea of how close you were getting to OOF.

An idea, I don't claim that it is a good one.

Mark

 

[jaapv]

Yes, but if they are taking pictures not studying physics they will be comparing the DOF based on the composition not the focal length. Someone will compare the same shot taken with his M7 and his M8, same size print and find that the M8 shot has more DOF ...because to get that same shot with the M8 he had to use a shorter focal length lens. How the DOF of a 50mm lens compares from one sensor to another is really not relevant to actual photography since using a 50mm on 2 different sized sensors won't give you the same shot. Capiche?

No, Ben , I do not capiche, for three reasons:
1. If the theory behind DOF is not understood one gets reactions like it is a law of physics, or it does not change, or it is a property of the lens etc. See this thread
2. The difference in DOF between film and the M8 is just over 1/2 a stop, so hardly relevant and it is not very illuminating to compare film to minisensors with the attendant focal lengths.
3. RD1 users report this to be a total non-issue.

 

[jaapv]

Well, we are talking about the same thing, but then again we are not. Here again, it is not a "pixel" density problem, but it is whether you see a disk (blur) or a point. (Your eye is always moving and rod and cone size (pixel pitch) is not what determines what you can see.) I know it sounds crazy, but the ability to resolve something and the perception of sharpness is not the same. So we agree on principle, but not on mechanics.

So if I had Delirium Tremens I would have more DOF? 😀😀

 

[Ben Z]

I don't know how the RD-1 people consider it a non-issue because that's a 1.5x and on my 1.6x 20D there's a very perceptible difference in DOF for shots of equal FOV compared to full frame. The same shot I use an 80mm lens for on a 5D, I must use a 50mm lens on my 20D and I get clearly, obviously and significantly more DOF (same size final printl). If there's only a half a stop difference between full frame and the 1.33x M8 I'll be a very happy camper...that would be terrific.

 

[jaapv]

I'm sure we'll both be happy🙂🙂 It is about 2/3 of a stop, actually.(on the same lens, though)

 

[Ben Z]

I'm sure we'll both be happy🙂🙂 It is about 2/3 of a stop, actually.(on the same lens, though)

The shot I'd use a 50mm for on the M8 would require only a (closest approximation) 65mm on an M6, not an 80mm as it would on a 20D, so I'd expect less of a DOF difference. In some cases, like most outdoor photography, the slightly expanded DOF will probably be more an asset than a liability. Only with portraiture will it be a hindrance and fortunately it's easy to blur a background in Photoshop whereas sharpening it would probably require compositing 2 images.

 

[jaapv]

Consider this as well: It may not always be necessary to move down in focal length for the same FOV. The use of a chimping screen will enable users to compose more tightly with confidence for many subjects. I don't know about others, but I often find I've left a "security margin" around my shots, resulting in a mild crop for the final composition.

 

[Ben Z]

Who was it who said "If they give you lemons, make lemonade" ? 😀

 

[Socke]

The difference between a 35 on a M8 and a 50 on a film M will be a couple of inches at best. I don't think the markings on the lens are acurate enough to see it now.
IMHO they can't be because Leica doesn't know what I find acceptably sharp at what enlargement.