Depth of field question

[Randy]

I am not sure where to post this, so if it needs to be moved, take it away mod's.

I was perusing Shorpy (http://www.shorpy.com/) recently and came across a photograph that intrigued and confused me:

http://dl.dropbox.com/u/52893762/parlorcar.jpg
link (http://www.shorpy.com/node/9954)

Shot in 1905 on 8X10 glass plate. If you open the link (http://www.shorpy.com/node/9954) and click on "View Full Size" it will be evident that it is very sharp from the closest chairs to the distant wall. I have never shot much at minimum aperture, so I don't have a lot of experience as to the kind of depth of field you can get out of large format. In this photo it seems obvious that no tilts or swings were used, nor could have been beneficial to the extreme depth of field (am I using the correct term, or should it be "depth of focus"?)

So is this photograph strictly the result of using a very small aperture? I am also guessing that the photographer knew where the hyper-focal distance was for his lens?

Thoughts? And if anyone has some examples they have shot and achieved great depth without using camera movements, just hyper-focal focusing and small apertures, I would like to see some.

[Vaughn]

Actually the closest chairs are not sharp, but they are a minor part of the image as not to matter. Does not look like too short of a lens for 8x10 -- the near chairs are not significantly bigger than the far chairs.

But probably closed down to the smallest aperture. The length of the exposure is not an issue since there is nothing moving. My 'trick' in this case would be to pick a place about 1/3 of the way into the scene (compensating for the more distance view in the mirrors) and then start closing the lens down. If the near and far come into focus at about the same rate, then I know I placed the plane of focus properly into the scene. If, for example, the near chairs came into focus before the far chairs did, then I would know that I have focused too close, etc.


In this photo it seems obvious that no tilts or swings were used, nor could have been beneficial to the extreme depth of field (am I using the correct term, or should it be "depth of focus"?)

Camera movements do not increase the Depth of Field (that is what I call it), instead, places it to the image's best advantage, so you are right that movements in this image would not be of much help.

Vaughn

Two examples where front tilt would not have helped.

Wawona Tunnel -- 8x10 platinum/palladium print
Tolaga Bay Wharf -- 16x20 Silver Gelatin print from 4x5

[Andy Eads]

I just killed another hour looking at the marvels on Shorpy. There are perils associated with this forum!

[Brian Ellis]

For optimum depth of field with minimal aberrations focus on the nearest object you want in focus, note where the lens is on the camera bed or rail at that point, then focus on the farthest object you want in focus, make the same note, measure the distance between the two points on the bed or rail, then consult a depth of field table. This is all explained in much better and greater detail in the article by Tuan on the forum home page about focusing a view camera.

FWIW, I'd suggest avoiding hyper-focal distance tables, unless perhaps you plan to only make a print of the same size as the tables were designed for (which I've read is usually an 8x10 print). And even then I never found one that really worked as it was supposed to so I've always avoided the whole concept.

[cowanw]

I have never done this, but theory suggests that, if the railway car is say ten feet high, set up your camera at five feet and tilt such that the plane of focus is as near to horizontal as possible. if you can then set an fstop such that you have five feet of depth of field on each side of the plane of focus, you should be good to go. Real world comments?

[C. D. Keth]

I have never done this, but theory suggests that, if the railway car is say ten feet high, set up your camera at five feet and tilt such that the plane of focus is as near to horizontal as possible. if you can then set an fstop such that you have five feet of depth of field on each side of the plane of focus, you should be good to go. Real world comments?

Good plan if you're a distance from the railway car or above it so the focus can skim the top. If you're closer, you'd see the problem. The top and the bottom of the car would likely be soft. When you use tilt or swing, the depth of field does not fall away evenly on each side of the plane of focus any more. It still follows the usual rules and becomes less the closer you get to to the camera. That means the shape encompassed by the tilted depth of field becomes like that of a wedge with the sharp end intersecting the planes of the lens and of the film.

[Vaughn]

I have never done this, but theory suggests that, if the railway car is say ten feet high, set up your camera at five feet and tilt such that the plane of focus is as near to horizontal as possible. if you can then set an fstop such that you have five feet of depth of field on each side of the plane of focus, you should be good to go. Real world comments?

It may not be possible to get the plane of focus anywhere near horizontal and still keep the back of the camera vertical.

[C. D. Keth]

It may not be possible to get the plane of focus anywhere near horizontal and still keep the back of the camera vertical.

I know I couldn't do it with the camera and lenses I own. Given that challenge, I'd probably light it to draw the eye toward the sharp stuff and away from the stuff I have to leave sharp. In that railroad car, that's roughly the lighting scheme that feels natural anyway.

[rdenney]

Even when the subjects were suitable, we should remember that few lenses of a century ago and more provided much coverage. Extreme movements would have required a long lens with respect to the format in those days, assuming the back had to remain vertical. Movements at the front of the camera would not have been used to angle the focus plane to the extreme that we can do today.

And even on Shorpy's largest image size, I'm looking at a 2x enlargement that is displayed at about 100 pixels/inch. That can display at best about 2 line-pairs/mm, which is maybe only a fourth of what people with good vision can see on close inspection. At that (very low) potential sharpness, the circle of confusion is somewhere around half a millimeter at best. Without the 2x enlargement, the film would only have to achieve a quarter millimeter circle of confusion to attain all that the Shorpy site can display. DOFMaster suggests that given a 12" lens on 8x10, focused at 9 feet, f/64 would achieve that standard of sharpness from 6 to 15 feet. f/128 would achieve what we see here.

Stated another way, a real 16x20 print might reveal variation in sharpness on close inspection that you can't see on the screen display on Shorpy's site. And the photographer very likely never intended the photo to be printed larger than a contact print, where the inherent micro-contrast would make the image seem sharper in any case.

Rick "noting that the photographer, if intending contact prints, would not care a bit about diffraction effects" Denney

[amac212]

Uh oh. I never knew about shorpy until now. Time slipping right through my hands again... :)

[Leigh]

if the railway car is say ten feet high, set up your camera at five feet and tilt such that the plane of focus is as near to horizontal as possible.
That's just the Scheimpflug principle. It shifts the plane of focus from vertical to some other angle.

It does not apply in this case because both the floor and the ceiling are equally in focus, which would not happen if
the original photographer used Scheimpflug for the shot under discussion.

- Leigh

[Vaughn]

Not exactly what was being asked, Leigh. He wanted to know if in theory, one could use the Scheimpflug principle in such a way to place the plane of focus (PoF) parallel (or almost) to the floor, and by stopping down, have enough DoF to get both the floor and ceiling in focus.

1) One would set the camera, not at 5' (in his example), but at 3.3 feet...because the 1/3 of the DoF would be below the PoF and 2/3 above the PoF.

and 2) I think one would not be able to keep the back of the camera vertical to achieve a horizontal PoF, thus creating a perspective (if that is the correct word) problem.

I liked his question...made me think.

Vaughn

[cowanw]

Thanks, Vaughn. I have seen landscape shots with pebbles on the ground in front of the camera to infinity, in focus. Actually Corran in his sport football photo on this page
http://www.largeformatphotography.info/forum/showthread.php?95462-October-portraits/page12
Illustrates my thinking, with a plane of focus about 6 feet wide at chest height or so.
I am aware of the spreading nature of the depth of field in this situation; the farther away, the wider the view angle sees anyway.

[Vaughn]

Thanks, Vaughn. I have seen landscape shots with pebbles on the ground in front of the camera to infinity, in focus. Actually Corran in his sport football photo on this page...

Actually it looks more like he used tilt to purposefully throw the PoF, and thus the DoF way out of whack -- to give a very narrow and precisely placed DoF -- in a similar way to the recently popular images that make normal scenes look like photographs of minatures. Looks like he tilted the from standard backwards, rather than the usual forward tilt (easy to do on a Graflex press camera).

[Drew Bedo]

What about curvature of field? Might the 100+ year old lens have a less-than-flat plane of focus? If so, would that help in keeping the composition sharp?

[Kirk Fry]

If you look carefully at the large picture in the link you can see the camera in the mirror. The front standard is pretty much straight up and down. I can't read the name on the lens. Given that you can't see the photographer, I bet it was a very long time exposure.

[C. D. Keth]

What about curvature of field? Might the 100+ year old lens have a less-than-flat plane of focus? If so, would that help in keeping the composition sharp?

Curvature of field pretty much goes away when you stop down so I think trying to take advantage of that would just be a step backward since, no matter how to hack it, a fairly deep stop is needed for that photo.