f/5.6, optimal focus | f/11, 3 cm defocus at 3m (100 mm lens) |
(http://echophoto.dnsalias.net/ekr/images/DoF2/A55_100Macro_small1-5.jpg) | (http://echophoto.dnsalias.net/ekr/images/DoF2/A55_100Macro_small1-6.jpg) |
F/5.6, optimal focus | f/11, optimal focus | f/16, optimal focus |
(http://echophoto.dnsalias.net/ekr/images/DoF2/A55_100Macro_small1-5.jpg) | (http://echophoto.dnsalias.net/ekr/images/DoF2/A55_100Macro_small1-13.jpg) | (http://echophoto.dnsalias.net/ekr/images/DoF2/A55_100Macro_small1-17.jpg) |
No I did not miss that. I have shot large format for years and know perfectly well about the use of hyperfocal focus setting.
Reportage involves a lot more than shooting acceptably sharp at infinity with the camera set to hyperfocal point.
Intgeresting, and I'm not going to argue with or question any of that. An alternative view...
CoC gets a lot of time on the net but personally I think that for many it's a non issue or at best a red herring. For it to be crucial I suppose we'd need to start by knowing the image size and the viewing distance but these days many people never actually print the vast majority of their images and I bet rarely view at the image size and distance to make the numbers as pure and correct and vital as the internet would have us all believe that they are.
Personally I solve these issues by ignoring the whole CoC thing. To me it's an irrelevance or at least has only been an issue once in my life, when producing artwork for a stage play. When doing something like that it does matter (people sat at the back wont see small detail at the back of the stage) and the CoC could be into inches but it's never ever mattered to me in my decades of taking pictures with cameras. All I care about is that the hardware I use is capable of producing the detail that I want at an image size bigger than I'll ever print. My current gear does that. My 5Dc and lenses seem to be ok, they were designed to work together so that's no surprise, and my G1 is ok too even when ignoring the new sharp lenses designed for MFT (x2 crop, smaller CoC, sharper lenses, more image magnification etc) and going with old manual Rokkor and Zuiko lenses. I'm sure that throws the numbers out but it doesn't seem to matter, not to me anyway.
Anyway, front to back DoF. Unfortunately I think that it probably does as much damage as good. Personally I think that being selective with focus and letting it drift rather than aiming for a fudge from front to back produces a more natural looking result, but of course we do have to keep in mind some guesstimation of what the final image size and viewing circumstance will be... but I'm not going to get tied up in CoC calculations. Plus there are many who seem to strive for near to infinity "sharpness" when their scene doesn't need it. This is another thing that we see on the net IMVHO, people using setting gleamed from DoF tables that result in a DoF that exceeds their scene. Then there's the Merklinger method, easier to deal with if you don't want to either memorise or carry DoF tables.
Having got all that off my chest... I often focus on my main subject. I often shoot with shallow DoF. I often shoot with the intent that my image will look more or less natural and as the eye sees it with things in the distance being less defined than things in the foreground. I also tend to ignore diffraction, a boost to contrast and fiddling with levels and saturation post capture seems to hide the nasties well enough if and when I decide that small apertures are required and no one has ever said to me anything like "That shots ruined by diffraction" or "What's your CoC here?"
Sorry about all that :D
I have two issues with hyperfocal shooting and neither of those is absolute.
1) If you calculate hyperfocal distance the calculation will assume some CoC. What I have seen, critically sharp images would require a small CoC. In my experience I would say that setting the CoC equal to pixel pitch is needed for optimal sharpness.
2) If we use a to large CoC we get into a situation where almost nothing is really sharp. Infinity will not be in correct focus and neither will foreground be perfectly sharp.
BTW, setting the CoC to pixel pitch simply is defining the depth of the plane of focus. That is not DoF.
Hi,
But of course it is DoF, although personally I use 1.5x sensel pitch because when the CoC reaches that diameter I can see microcontrast starting to suffer. This is the perfect parameter to calculate Focus Stacking intervals, IOW DoF.
Cheers,
Bart
Only for pixel peeping. That is not a real world viewing condition. Naturally, the problem comes up that given the same shooting conditions and optics, a given format will have different DoF just because the sensor is divided into more or less units. Naturally, that is false.
BTW, a CoC value does not change micro contrast. It actually does nothing to the image. CoC is simply modeling perception. You have a model for 100% monitor view and your personal taste.
BTW, setting the CoC to pixel pitch simply is defining the depth of the plane of focus. That is not DoF.
CoC means circle of confusion. DoF tables are calculated based on maximal permitted CoC. That means that close and far DoF limits will be rendered with the CoC used as parameter, but somewhere in between there will be zone with much smaller CoC.
On a very high performance system a point will be rendered as a point in the focal plane. If the image is defocused the point will not any longer be rendered as a point but as disk. The diameter of the disk is CoC.
Is 1/width of the PSF more or less a Gaussian function of distance from the focal plane? Such that focus setting and aperture size are its two parameters?
When doing focus stacking (typically merging several shots of larger aperture, with smaller PSF in each focal plane), do you typically end up with a more rectangular function? (within some range, everything is really sharp, outside that range sharpness falls of more rapidly).
At the same time, I don't understand why anyone would routinely shoot smaller than f/11 with the likes of a D800.