Circle of Least Confusion

[FOOXX]

me am trying to find out more abt hyperfocal diztance .....

came acrozz thiz term called Circle of Leazt Confuzion .....

can anybody tell me wat iz that ..... ?? .....

[ST1100]

A point is focused on film. When it is out of focus, the point appears as a circle with a certain diameter. When it goes into focus, the image on film gets smaller and smaller, approaching the 'point' size it is supposed to be. The COC is the largest allowable diameter of the image of the point for that point to be considered 'in focus'. The smaller the COC, the sharper the image. For 35mm format, COC of 0.03mm is the norm.

[jimtong]

Originally Posted by ST1100 A point is focused on film. When it is out of focus, the point appears as a circle with a certain diameter. When it goes into focus, the image on film gets smaller and smaller, approaching the 'point' size it is supposed to be. The COC is the largest allowable diameter of the image of the point for that point to be considered 'in focus'. The smaller the COC, the sharper the image. For 35mm format, COC of 0.03mm is the norm.

So does a DLSR consider as 35mm in this aspect?

[ST1100]

A dSLR with a crop factor will need a stricter COC for the same size print, bcoz the degree of magnification is higher from sensor to print.

However, i don't think the lens is aware of that. The DOF scales on the lens are calibrated for 35mm and is not aware that the body has a crop factor. The dSLR body, which does the focussing, can theoriticaly be calibrated to a stricter COC but i don't think it's done.

[Jed]

Originally Posted by ST1100 A dSLR with a crop factor will need a stricter COC for the same size print, bcoz the degree of magnification is higher from sensor to print. However, i don't think the lens is aware of that. The DOF scales on the lens are calibrated for 35mm and is not aware that the body has a crop factor. The dSLR body, which does the focussing, can theoriticaly be calibrated to a stricter COC but i don't think it's done.

Errr. Lost me.

[TME]

yeah me also lost in translation ....... speak English dude! Thanks!

[StreetShooter]

Actually I thought that ST1100 gave a very good explanation. Just think a bit about what he said. This is my understanding of it:

Let's say we draw an infinitely small black dot (or, in real terms, say 0.001mm diameter) on a white piece of paper, and try to focus that image on your camera. When the image of that black dot is out of focus, it appears like a grey circle. As you gradually bring it into focus, it becomes smaller and smaller as it approaches the size of the black dot. The circle of confusion is the size of that black dot on the film when it is CONSIDERED to be in focus ie when your eye can't tell the difference even if it becomes more sharply focused after that. In the case of 35mm film, this is usually ARBITRARILY determined to be 0.03mm diameter. In other words, even if you manage to focus until that dot is 0.01mm diameter on your film, it won't APPEAR any sharper.

For a DSLR, because of the crop factor, the focused image of that dot should be smaller (eg 0.02mm) ie the circle of confusion should be tighter. This is because the DSLR sensor is smaller (0.6X crop factor). If you blow up both the 35mm film image and the DSLR sensor image to, say, 8R size, the DSLR sensor image will be more highly magnified. That 0.03mm dot will be magnified MORE than the 0.03mm dot on the film negative. If the image is considered in focus when the circle of confusion is 0.03 for a DSLR image, then the DSLR image will appear less sharp. In other words, there needs to be a tighter tolerance for focusing with a DSLR, to achieve the same degree of sharpness for any particular print size.

The DOF scale on a 35mm SLR lens indicates which area will be considered "in focus" ACCORDING TO 35MM FILM FORMAT, ie the circle of confusion is arbitrarily taken to be 0.03mm in calculating the DOF achieved at any particular aperture. This is a looser tolerance than a DSLR requires, and therefore may not be accurate for a DSLR. Why are we talking about a DOF scale? Because the original question was about hyperfocal focussing, which presumes the use of a DOF scale.

In PRACTICE, I doubt that the difference matters very much at all. Just use the old rule of thumb: f8 and be there. If you want to be a bit more kiasu, use f11. But it better be very bright, or you better be using a tripod.

[jimtong]

Originally Posted by ST1100 A dSLR with a crop factor will need a stricter COC for the same size print, bcoz the degree of magnification is higher from sensor to print. However, i don't think the lens is aware of that. The DOF scales on the lens are calibrated for 35mm and is not aware that the body has a crop factor. The dSLR body, which does the focussing, can theoriticaly be calibrated to a stricter COC but i don't think it's done.

[TME]

Originally Posted by StreetShooter Actually I thought that ST1100 gave a very good explanation. Just think a bit about what he said. This is my understanding of it: Let's say we draw an infinitely small black dot (or, in real terms, say 0.001mm diameter) on a white piece of paper, and try to focus that image on your camera. When the image of that black dot is out of focus, it appears like a grey circle. As you gradually bring it into focus, it becomes smaller and smaller as it approaches the size of the black dot. The circle of confusion is the size of that black dot on the film when it is CONSIDERED to be in focus ie when your eye can't tell the difference even if it becomes more sharply focused after that. In the case of 35mm film, this is usually ARBITRARILY determined to be 0.03mm diameter. In other words, even if you manage to focus until that dot is 0.01mm diameter on your film, it won't APPEAR any sharper. For a DSLR, because of the crop factor, the focused image of that dot should be smaller (eg 0.02mm) ie the circle of confusion should be tighter. This is because the DSLR sensor is smaller (0.6X crop factor). If you blow up both the 35mm film image and the DSLR sensor image to, say, 8R size, the DSLR sensor image will be more highly magnified. That 0.03mm dot will be magnified MORE than the 0.03mm dot on the film negative. If the image is considered in focus when the circle of confusion is 0.03 for a DSLR image, then the DSLR image will appear less sharp. In other words, there needs to be a tighter tolerance for focusing with a DSLR, to achieve the same degree of sharpness for any particular print size. The DOF scale on a 35mm SLR lens indicates which area will be considered "in focus" ACCORDING TO 35MM FILM FORMAT, ie the circle of confusion is arbitrarily taken to be 0.03mm in calculating the DOF achieved at any particular aperture. This is a looser tolerance than a DSLR requires, and therefore may not be accurate for a DSLR. Why are we talking about a DOF scale? Because the original question was about hyperfocal focussing, which presumes the use of a DOF scale. In PRACTICE, I doubt that the difference matters very much at all. Just use the old rule of thumb: f8 and be there. If you want to be a bit more kiasu, use f11. But it better be very bright, or you better be using a tripod.

Muah...... thanks....... I really ddin't understand the 2nd part of ST's explanation, the DSLR part.... but as I dun use hyperfocal distances (I dun really understand how it is use and only my 100-400 zoom has it and I seldom use it) it does not help either....

In any case, a quick clinic if u please, I used f8 for outdoor (under a huge raintree) on a 17-35mm, and while the shot was sharp (edges were soft due to lens), it was a little dark. It was 8 in the morning with the sun coming from my right (perpendicular to camera lens). I was on A mode with flash (with 17mm diffuser as I was at the widest). What I want to ask is would it have been better if I had not trusted my meter and overexposed or remove the flash diffuser? I'm not sure the latter would help but the former certainly would. Only that I'm not too sure about this lens as I also seldom use it. Using a Dynax 7, Sigma 17-35mm and 5600HS flash. ISO400 film.

Thanks!

[ST1100]

Apologise for the poor explanation, and thanx StreetShooter for clarifying the mess.

The last bit i said about the dSLR "being calibrated for stricter COC" - i meant the dSLR can be tuned for a tighter tolerance for what is considered 'in focus'. i got this idea from all the 10Ds being 'within tolerance' when they actually are NOT in focus on the main subject - that the tolerance can be adjusted.

Hope i didn't mess it up further...

[Minoxman]

Ok, but what's the point?

[Castlesinthesky]

so in short. the COC is just a rule of thumb on saying what is the maximum point for the image to be in focus eh?

[Ian]

Er guys,

Strictly speaking the Circle of Confusion is the diameter at which the average human eye can gain no further detail from an image at a given distance, which is based on the size of print being viewed. It's typicaly 0.02 to 0.03mm at 30cm or so for a 4x6" print. It equates to around 6 line pairs per mm.

[StreetShooter]

Originally Posted by Ian Strictly speaking the Circle of Confusion is the diameter at which the average human eye can gain no further detail from an image at a given distance, which is based on the size of print being viewed.

Yes, that's correct.

Originally Posted by Ian It's typicaly 0.02 to 0.03mm at 30cm or so for a 4x6" print. It equates to around 6 line pairs per mm.

I'll have to take issue with that. I believe your 0.02 to 0.03mm refers to the circle of confusion on a 35mm negative, not on the 4x6" print. The CoC depends on several factors - the negative format, the viewing distance and the size of the final print. A 0.03mm dot on a negative blows up to a much larger dot on an 8x10" print. A 0.03mm dot on a DSLR sensor is magnified a further 1.6X more, leading to a less sharp image. But of course, if you view the DSLR print from a larger distance away than the 35mm negative print, the sharpness would appear equal.

This url gives a GREAT explanation of circle of confusion:

http://www.northnet.org/jimbullard/CoC.htm

And this gives you a nice table of the (approximate) CoC for various formats:

http://www.nikonlinks.com/unklbil/dof.htm

If I'm wrong, please let me know.

[ST1100]

Originally Posted by Minoxman Ok, but what's the point?

The original poster asked about COC wrt to hyperfocal distance.

i think it's good to understand all this math (follow FOXX's link) bcoz you would know how sharp you can expect your image to be in enlargements, and more importantly, if you need to recalculate the DOF with a different COC for a particulare size enlargement. The default value COC may not satisfy some folks' quest for sharpness, esp landscape photogs who print big.

[Ian]

Originally Posted by StreetShooter Yes, that's correct. I'll have to take issue with that. I believe your 0.02 to 0.03mm refers to the circle of confusion on a 35mm negative, not on the 4x6" print.

My error for not making it clearer, the COC is a physical limitation of the eye and it's a relative scale based on print size and accepted viewing distances. That will teach me to post on 3hrs sleep and only one cup of coffee

[FOOXX]

thankz to all for the explainationz .....

[Larry]

my brain just exploded.