Photographing"the lean"

[photoscientia]

The shutter has to travel vertically while the camera is panned horizontally to get the effect. It's pretty obvious that the bottom of the image was exposed first, and the top last. It doesn't matter what shutter speed you set within reason, the amount of leaning is governed by the panning speed and the speed of travel of the blinds, which is constant, not by the slit width. You could probably get the same effect with an old rangefinder Contax, or a russian copy, especially if the shutter was in the state that they usually turn up in.

Lartigue was given his first camera at the age of six, so he already had 4 years experience under his belt by the time this shot was taken.

[M.]

The interesting question here is why doesn't the gas tank (with the number 6) appear as out-of-round as the wheel. I think I was wrong, and he must have actually panned at car speed. Neat-o!

[TAN K H]

Ah, I am wrong after all! Yes, a vertically travelling shutter it is!

[David Richhart]

Thanks to all for the thoughts. Uncle Ansel discusses this picture and the use of the focal plane shutter in his book "The Camera". It is just a general discussion with no specifics.

According to AA, when the shutter slides the same direction as the moving object it "stretches" the image. When the shutter slides in the opposite direction it "compresses" the image. When the shutter is travelling from top to bottom the bottom of the tire is recorded first because the image is projected upside down.

But, as some of you have pointed out, that extremely distorted wheel next to such a sharp image of the driver is a real puzzle.

Guess I'll have to load some film and go play in traffic... -Dave

[David A. Goldfarb]

It's hard to tell without seeing the negative (which would reveal uneven exposure), but the shutter could accelerate a bit as it comes down. More likely, the panning speed and/or the speed of the car could be uneven, so that the camera and the car might have been more in synch at the top of the picture (the end of the exposure) than they were at the beginning of the exposure. Since the poles or trees in the background are fairly straight, I would guess the panning and shutter speeds were even, and the car slowed down slightly to the panning speed.

[Struan Gray]

The distorted wheel is not oval: the bottom half is much more distorted than the top half; and for that matter, the bottom of the wheel on the far side of the car is also less distorted than the one closest to us.

The panning speed therefore more closely matched the apparent motion of the car when the shutter slit was in the middle of the image than when exposing the road at the start of the curtain travel. If you draw some circles on the image you will see that in fact the very top of the distorted wheel is a pretty good circular arc, which agrees with the undistorted image of the body of the car and its occupants.

Assuming that no darkroom tricks have been employed, the car must have slowed down or the panning rate must have sped up. Since the trees and people are so uniformly distorted, I'd guess it's the former, but it's common for people trying to track a moving object to pan too slowly at first and then catch up, so it's possible that's what happened.

Panning at exactly the car's speed will give circular, blurred wheels, not ovals. Pan a little too slowly and the wheels lean forward if your shutter travels from top to bottom of the focal plane (bottom to top of the image). Pan a little too fast and they lean the other way. In both cases the car body will also distort, so you need an uneven pan speed (or car speed) to get the effect shown here.

[M.]

The car is photographed at an angle (right rear quarter view) which, even in a still picture, would produce oval appearing wheels and gas tank. Since the car was actually moving they would be less oval (more circular) when the angle is smaller and the FP slit was near the bottom, and more oval progressivly as the car got at a bigger angle to the viewer. This partially explains the "lumpy" shape of the wheel. I don't think it answers the whole question. Go to it Dave!

[John H. Henderson]

I just got a Contax IIA with a vertical shutter. I need to go try this.

Some modern 35mm cameras have vertically-travelling shutters - particularly the Copal square metal shutter. The Minolta XD-11, I believe, has this shutter, and this weekend I was looking at a friend's Nikon EM which also had this shutter. The new Voightlaender Bessa-R and Bessa-L also use this shutter.

[Ellis Vener]

I suspect that while Lartigue was panning with the car the shutter was moving in the opposite direction and also it wasn't running smoothly at a constant speed, hence some parts of the image are more distorted than others. It is very unlikely that any fancy tricks were being played and that the image is at least partially,a "happy accident".