Parallel lines and circles

[Stew]

I have limited experience with a large format camera and I have not had to worry about things like converging lines. Tomorrow I have to make a photograph in which the camera will be at a 50 to 60 degree angle to the top of a table. The top is made of ceramic tile . The only object on the table is a soup bowl about 8 inches in diameter. The image will consist of most, but not all, of the bowl plus part of the table top.

I did a quick test tonight for composition. I'm using a 4x5 camera and a 150mm lens. The shot covers an area on the table of about 12 inches by 15 inches. The grout lines between the tiles on the table top converge. Also, to my eyes, the bowl appears slightly oval rather than round.

How can I maintain the 50 to 60 degree angle of view and at the same time avoid convergence of the parallel lines?

How can I make sure that the bowl appears as a circle, or am I just imagining that it is slightly oval?

Comments most appreciated.

[Wilbur Wong]

Why do you not want the grout lines to converge?

With most 150 mm lenses, 30 to 40 degrees from normal (30 to 40 degrees from perpendicular, or 50 to 60 degrees from the plane of the table, if I understand you right) I don't think you would have enough lens coverage to maintain parallel lines, by using direct or indirect shift.

If you want the bowl to be perfectly round instead of oval you must have the film plane (the camera back) parallel to the table and bowl. That is simply the only geometry of the set up which will produce round bowls and show no convergence of the grout lines.

If this is what you want to achieve, simply set up the camera to shoot straight down on the bowl and counter.

If you want the appearance of a dinner looking at a bowl of soup, the most natural appearance would indeed have perspective of converging lines and an oval bowl rim.

[Stew]

Wilbur,

Thanks for helping me preserve my sanity. I was being told that I should be able to do this, and for the life of me I couldn't figure out how to maintain the angle of view and get rid of the convergence at the same time.

[Conrad Hoffman]

You can do this to some limited extent, but I can't see the reason for it. As said above, if the back of the camera is parallel to the table, the lines won't converge. You can then shift the lens and the back as far away from each other as the lens coverage will allow, then reposition the camera off center until you have the shot you want. The back, lens, and subject all remain parallel. That will be the limit of offset due to lens coverage or possibly camera movements, but I wouldn't expect the results to look natural.

[james mickelson]

Prop the table up so that it is paralell to the camera back when the back is at it's maximum tilt. Put double stick tape on the bottom of the bowl so it doesn't slide off. A couple of books should be able to lift the far side of the table enough to let you use your back tilt to keep the grout lines paralell

[Mark Sampson]

...james' idea will work fine until you try to pour the soup into the bowl. ;-)

[Stan. Laurenson-Batten]

It is worth bearing in mind that perspective is altered differently depending on whether the front or rear controls are used. To keep proper perspection as far as is possible, it is best to use the
front controls of tilts and/or swings, that is at the lens plane. Set tilts first then swings.

If you are using a camera with asymmetrical movements as opposed to symmetrical, then your task is easier. Bear in mind; when transfering the movement from the rear standard to the lens standard, that they are reversed. The rear standard is then reset to zero and then, if necessary refocus.

When the front lens standard is swung or tilted the image circle becomes an oval, this can effect the coverage of the lens on the film plane/area. Should you be using a roll film back, or format other than that for the camera/lens was designed then other the complications will arise in the control of the image. There are other ways as pointed out above, your choice will depend on whether perspective control is met to your requirements.

[Stew]

"james' idea will work fine until you try to pour the soup into the bowl. ;-)"

Gravity is such a nuisance.

[neil poulsen]

You might be able to do this with the 150mm Schneider XL. It has a total angle of coverage of 105 degrees at f22. That translates to a half-angle of 53 degrees at f22. After doing some calculations, this translates to about 43 degrees rise. (There's more to it than just subtracting 10 degrees.) Stop the lens down a lot, and you might be able to get 50 degrees or a little more with the back in horizontal orientation.

In this situation, the lines in the table are converging towards a vanishing point. It would appear aunnatural to have them parallel. I don't see how you can keep the bowl round and try to make the lines parallel. Even with the converging lines, there could be a little distortion causing the bowl to look oval.

Another possibility is to do what you can with the lens you have available, and then to also use the transformation function on a scanned image in Photoshop.