Printable View
"You don't get the horizontal lines parallel by leveling the camera horizontally."
What? Leveling horizontally works in the same manner as leveling vertically. As Doremus said (and I said, perhaps unsuccessfully) is that.."If you nail the camera position, then all you should have to do is simply camera center the building on the ground glass. If you camera position is correct, the horizontals will be parallel. If the aren't, you need to reposition the camera."
That's simply not true, Thom. Left/Right-Leveling only ensures that a horizontal line running through the middle of the GG is level. It does not lead to non-converging horizontals automatically. See for example "View Camera Technique" by Leslie Strobel for detailes and example pictures. I can make a drawing if you like to explain better.
What Doremus said is not a contradiction to this.
Forward/Backward Leveling ensures that the back is parallel to a vertical plane only in vertical direction. Thus, no matter if the camera is tilted to the left or to the right, all vertical lines on the subject will be parallel to each others (they do not converge in the image).
Left/Right Leveling simply rotates the image on the GG. But for keeping horizontal subject lines parallel to each others on the GG, you can not make use of any spirit level. You need to make the back parallel to the subject plane horizontally. This can not be achieved by using spirit levels.
To get non-converging (parallel) horizontals AND verticals, the camera back needs to be a parallel plane to the subject plane (which means, the planes may not intersect horizontally and vertically).
To get these horizontal and vertical lines to be horizontal and vertical lines on the GG (without affecting their convergence!), you need to use left/right leveling.
Let me weigh in here just to clarify:Quote:
Originally Posted by Rollinhofuji
Thom, sorry to disagree with you, but I believe Rollinghofuji has got it right. Leveling the camera left-to-right is my first step and sets both the vertical and horizontal lines that intersect in the center of the ground glass. Subsequent leveling front-to-back then brings the back parallel to a building vertically, ensuring that the vertical lines at the edges of the frame are parallel and do not converge.
The next step, if one wants the camera back exactly parallel to the face of a building, would then be to pan left/right (around the correct vertical axis, of course) to bring the back parallel horizontally to the face of the building and ensure that horizontal lines do not converge. This points the camera down a line perpendicular to the face of the building in question (but not necessarily at the center of the building).
My advice to Rollinghofuji uses this principle in reverse; i.e., if you have the camera positioned on a line perpendicular to the exact center of the building, and if you point the camera back down that line at the exact center of the building, then the back will automatically be parallel horizontally to the face of the building. You can use this fact to check camera position.
Hope this clears things up,
Best,
Doremus
Ah, I see now where we are misunderstanding each other. I am assuming the OP is trying to be exactly centered and perpendicular to the subject. If those criteria aren't met, and the camera position is at an angle to the subject and not perpindicular, then the top and bottom horizon lines will not show as horizontal on the ground glass and he is panning to achieve that. If I am correct in this assumption, then I must ask; why not use the rear swing to bring those lines into true? In other words, if you are perpendicular to the subject, yet not centered on the subject, can you not swing the rear to bring those horizontals true? Then swing the front to the same degree and shift as necessary to compose. Anyway, as others have stated, what you see on the gg is gospel. That's the great thing about using a view camera with movements.
Thom, in fact I never thought about that - thanks for this hint! Maybe I took it all too complicated. This seems to be quite an effective approach.
Thom,Quote:
Originally Posted by Thom Bennett
Sure, your method works fine, but is in essence exactly the same as panning the camera. Sometimes it is easiest to swing back and lens parallel and then "pan" using the shift. As long as the film plane the lens plane and the plane of the front of the building are parallel, there will be no converging lines on the ground glass (back parallel) and everything will be in focus (lens plane parallel).
But, if Rollinhofjuji wants parallelism and the center of the building to be the optical center of the image, he really needs to be on that line perpendicular to the center of the building. There's no other way to position the optical center and keep parallel lines from converging.
Best,
Doremus
Doremus, again you got the point...it IS the same. Might be more convenient anyway, will have to check. The good thing about Thom's rear swing method is the fact that the tripod doesn't need to be level (it's sufficient if the camera is level at a given pan setting). Because the camera will stay level that way.
For future pics, I will check which methods works best for me (especially if it is really required to have everything centered, or if good old view camera movements may give even better results).
Thanks all for your helpful input!
If you are able to take two lengths of string or rope of equal length from each corner of the building - you would be given the exact center - this can probably be done with a laser measuring device as well - although I would just use string and geometry.
Thanks Robert! ic-racer also mentioned that method which is suprisingly easy (and extremely precise).
So true...Quote:
Few things in the architectural world line up perfectly believe it or not.