Imagemaster, I don't think Monito was addressing the aesthetic qualities or preferences of DOF. The point is that DOF can be reduced to a mathematical model. And the intent of this specific thread was to point out that sensor size is a factor in the model.

Sure, no one needs to do the math, because, as some have argued like you, it's too much trouble. I'll grant that point, but understanding the factors can expand one's knowledge beyond the basics of "close down for more DOF, open up for less".

I don't know if this has been mentioned yet since I only read through page 13. Many people have mentioned that the size of the CoC is based on two things: the distance from the object and the resolution of the eye. The resolution of the eye is not measured in length but an angle. The resolution of a normal person is agreed upon to be 1 arc minute ( 60 minutes in a degree, therefore, 1/60th of a degree or 0.016667º). See here.

To determine the size of the circle of confusion all you need to know is the distance to the object. At one meter, the eye can resolve 291 microns:

CoC = (distance to object)*sin(0.01667º) or (dist to object)/3438

So for the standard value of 30 microns for any sized print, the viewing distance is 0.1 m (4 feet).

By saying that the resolution of the eye is 30 microns is just wrong! Can you resolve 30 micron features on the moon while standing on Earth? No, the distance from the Earth to the moon is 385,000 km (240,000 mi). That means that your average, unassisted Joe can resolve features as small as 112km (70 mi).

What Imagemaster is asserting is essentially correct: for most uses, the difference between FF and 1.6x with regards to DOF is negligable.

I have heard some people complain that they like the DOF they got at 85/1.8 on 35mm film and they can't duplicate that same effect precisely on 1.6x digital.

The difference between 1.6x or FF and digicams in terms of DOF is not negligable; if you want to shoot portraits with diffuse out of focus backgrounds you should not choose a digicam.

Imagemaster wrote: Actually, the joke is on you. If it makes such a HUGE difference, then why are you using a 20D

Only someone who doesn't understand circles of confusion would ask such a question in this context.

Imagemaster wrote: You don't seem to comprehend that good photography is not dependent on mathematics only. In most cases, the photographer has no control over the print-viewing distance, the eyesight of the viewer, nor the tastes of the viewer.

Exactly my point. Which is why later in my post I specifically mention bracketing apertures.

bouch wrote: What Imagemaster is asserting is essentially correct: for most uses, the difference between FF and 1.6x with regards to DOF is negligable.

Wrong. You can wave your hands and be wrong or you can do the math. Math rules.

50mm at f2.8 at 5 meters:

1.6 crop: 1.05 meters (4.53 to 5.58 m) (CoC 0.019 mm)
FF crop: 1.70 meters (4.29 to 5.99 m) (CoC 0.030 mm)

That is 62 percent more depth of field, not negligible.

Now, take an example to achieve the same angle of view (i.e. same perspective):

1.6 crop, 50mm (30 deg diag AoV), 5 meters, f2.8: 1.05 meters (4.53 to 5.58 m) (CoC 0.019 mm)
FF crop, 79.5mm (30 deg diag AoV), 5 meters, f2.8: 0.66 meters (4.69 to 5.35 m) (CoC 0.030 mm)

Now it is the 1.6 crop that has 62 percent more depth of field, not negligible.

Conclusion: Sensor size matters

Now can we put this to rest? Golly.

Ref. http://ca.geocities.com/lokejul/jlcalc.htm

Monito wrote:

That is 62 percent more depth of field, not negligible.


For most uses it's negligable. If I want to shoot a portrait with a blurred background, I can accomplish that in the 1.6x format, no problem. If I want to shoot a stopped-down landscape where everything looks sharp I can usually accomplish it in either 1.6x or FF. Rarely does a photographer specifically want 1m DOF instead on 1.6m. I outlined a possible case in my post.

I come back a couple of days later and this thing is still going.


depth of field
: the range of distances of the object in front of an image-forming device (as a camera lens) measured along the axis of the device throughout which the image has acceptable sharpness .

Webster's

I've never considered Webster's any kind of an authority for technical i.e. photographic definitions. Members will do much better to read one of the better web sites or one of the definitive text books (available from any public library) on the subject.

bouch wrote:

For most uses it's negligable. If I want to shoot a portrait with a blurred background, I can accomplish that in the 1.6x format, no problem. If I want to shoot a stopped-down landscape where everything looks sharp I can usually accomplish it in either 1.6x or FF. Rarely does a photographer specifically want 1m DOF instead on 1.6m. I outlined a possible case in my post.

Quite correct Bouch. Monito just seems to be stuck in a mathematical rut. Anyone who can't get the DOF that they require without worrying about sensor size has a problem.

Conclusion: Sensor size does not matter to most photographers.

Die thread, die!

Conclusion: Sensor size does not matter to most photographers.

This is incorrect, because of the general nature of the statement. It may not matter to Imagemaster or bouch, but to the vast majority of students involved in a professional program there is a distinct difference. Not all people want or can afford an f/2 or f/2,8 lens. Try to do portraiture with an f/4 or f/5.6 lens and getting just the tip of the noise to the top of the ear in focus. It is very difficult in some instances depending on the various factors affecting DOF (sensor size, focal length, aperture etc.).

So if sensor size is not important to the way you shoot or type of subjects you shoot fine. But don't make general assumptions about most people.

Jeff Donald wrote:
Not all people want or can afford an f/2 or f/2,8 lens. Try to do portraiture with an f/4 or f/5.6 lens and getting just the tip of the noise to the top of the ear in focus. It is very difficult in some instances depending on the various factors affecting DOF (sensor size, focal length, aperture etc.).

Talk about a general assumption. "some instances" hardly applies to most photographers. How about the general assumption that for 100 years, most portrait photographers have coped just fine with DOF without having to use an f2 or f2.8 lens, and most certainly without having a clue about sensor sizes

But don't make general assumptions about most people.

Don't misinterpret my statements. I said most photographers, not most people. Anyone who takes photos can be called a photographer. According to figures on camera sales, point & shoot cameras far outnumber DSLR's. Take a poll of one million random camera owners and ask them if sensor size matters to them. I will make the general assumption that not only do most of them not care about sensor size, but that most of them do not know that there are different sensor sizes.

bouch wrote:
Monito wrote:

That is 62 percent more depth of field, not negligible.


For most uses it's negligable.

Not always. Keep in mind that a 1.6x crop factor camera does NOT have a fixed 62% more DOF than a FF camera at equivalent FOV. It only holds true at distances much less than the hyperfocal distance of the lens/format/aperture combination. The hyperfocal distances of a 1.6x crop factor camera are closer than that of a FF camera, so the rear DOF will increase much more rapidly as object distances increase. This could spell the difference between having a blurred background and a sharp one.

Imagemaster wrote:
[Don't misinterpret my statements. I said most photographers, not most people. Anyone who takes photos can be called a photographer. According to figures on camera sales, point & shoot cameras far outnumber DSLR's. Take a poll of one million random camera owners and ask them if sensor size matters to them. I will make the general assumption that not only do most of them not care about sensor size, but that most of them do not know that there are different sensor sizes.


I'll wager that most of them don't care about narrow DOF let alone sensor size. They're perfectly happy with everything in the image sharp. So, that's the wrong demographic to be using to justify your claim.

slin100 wrote:
wrong demographic

That really is funny. So for my general assumption, I should narrow it down to a certain demographic of your choice. Yeah, sure.

Talk about a general assumption. "some instances" hardly applies to most photographers. How about the general assumption that for 100 years, most portrait photographers have coped just fine with DOF without having to use an f2 or f2.8 lens, and most certainly without having a clue about sensor sizes

For the last 100 years the majority of portrait photographers have used large and medium format cameras for their portraiture. The DOF of the larger formats are even less than 35mm and didn't require f/2 of /2.8. Why do you suppose many large format lenses stop down to f/90 and even f/128? Do you think that is to introduce more diffraction to soften the picture because it has so much DOF? It would help your argument if you had some understanding of the history of photography and a working knowledge of large format photography.

I'll wager that most of them don't care about narrow DOF let alone sensor size. They're perfectly happy with everything in the image sharp.

I'll wagger you're wrong. In my adult education classes, where the majority of students have P&S cameras, some of the most frequently asked questions are how to get better portraits and why don't their pictures look more like professional portraits. The students recognize the difference in their images, they just don't know that it's DOF and the limitations it imposes.

Imagemaster wrote: How about the general assumption that for 100 years, most portrait photographers have coped just fine with DOF without having to use an f2 or f2.8 lens, and most certainly without having a clue about sensor sizes

Of course. But they weren't as clueless as you suggest. They used large format cameras. That is the whole point of this discussion. Sensor (including film) size matters.

On a 4x5 camera, a 260mm lens has about 30 degrees angle of view, about what an 80mm has on a 35mm camera, a nice perspective for portraits. At a 3 meter distance, which gives nice framing for an upper body portrait, the depth of field at f5.6 goes from 2.94 to 3.06 m or about 118 mm (about 4 1/2 inches), a pleasing depth of field from the ears to the tip of the nose.

Sensor size matters and drastically affects depth of field, even if you and some other photographers don't think about it when photographing.

And while you are at it not thinking about it, why don't you and others drop the ad hominem attacks on those of us who do think about it, calling into question our skill as photographers. We have not questioned your skill as a photographer, only your technical understanding. If we choose to study and understand the mathematics it means we have extra tools and understanding at our disposal. It is not a "joke" and we are not "stuck in a mathematical rut". If you want to group yourself with the skill set of the mass of photographers who use point & shoot cameras and don't think about sensor size, you can. Or you can look at the facts as we have laid them out for you and others and learn that sensor size matters. For you to heap scorn on us while you choose to remain ignorant of and dismissive of the mathematics doesn't prove your case or make you look wise.

Jeff Donald wrote:

slin100 wrote:
I'll wager that most of them don't care about narrow DOF let alone sensor size. They're perfectly happy with everything in the image sharp.

I'll wagger you're wrong. In my adult education classes, where the majority of students have P&S cameras, some of the most frequently asked questions are how to get better portraits and why don't their pictures look more like professional portraits. The students recognize the difference in their images, they just don't know that it's DOF and the limitations it imposes.

What percentage of photographers have the gumption to take a class on photography? Probably in the single-digit percent range. The rest, therefore, don't know or care.

DOF is a perceived quantity - it's value is always given relative to a set of assumptions. Just as the physical aperture, and focus distance affect depth of field, so does image enlargement and viewing distance. Since image enlargement is as much a function of the sensor/film size, as it is the final print/display size, they all affect the perceived DOF. Any DOF table with fixed values for different focal lengths/f-stop/focal-distance combinations, assumes a certain print size and viewing distance - for a given medium (i.e. 35mm film).

For a detailed explanation, including formulae...
http://bobatkins.com/photography/technical/digitaldof.html