Curtain shutters and egg shaped wheels

[Jim Galli]

In Jacques Lartigue's (http://www.cracow-life.com/poland/jacques-henri-lartigue) famous early racing car pictures, I understand the people leaning because I think that J. H. L. was panning the camera hand held, so anything stationary would be at a different place as the shutter moves down, and if he is in fact 'panning' I can see that the car would be reasonably sharp at an effective 1/1000th second, but why are the wheels egg shaped? Why does the camera see them any differently than the car body?

I want to re-create this effect with a Model T speedster at speed using my 5X7 Speed Graphic. I know Lartigue likely used a chimney graflex type, but if I prefocus for a known distance that the car will be at speed, that shouldn't make any difference. I'll hand hold and 'pan'. Could even build a wire 'sports frame' like the later press 4X5's had.

[Paul Ewins]

From what I understand of it (and it made sense *at the time that I read it*) the effect happens not because you are panning but because the actual speed that the shutter runs at is quite slow, even although a narrow slit will give an effective 1/1000 speed.

So, with the camera held stationary and an object moving past quickly, when the shutter slit starts moving the object is in position X, by the time it has moved down 4" the object is in position x+1 so the top and bottom don't align. Because the curtain moves from top to bottom and the image is upside down the bottom of the image, i.e the wheels, is recorded first and as the car move across more of it is recorded untl it finishes with the top of the image. This is why the car always leans in the direction it is travelling.

Most older 35mm SLRs had shutters that moved horizontally so this wasn't an issue. The newer ones with vertical shutters should display the same effect, but with a travel time of 1/150 - 1/250 sec compared to (I think) 1/5 sec for the Graflex it is going to be so slight that you won't pick it up.

[Jim Galli]

The Graflex had an effective 1/1000th speed. It had 6 different possible speeds of travel, and 4 different widths of slit opening. 1 1/2", 3/4", 3/8" and 1/8" slits all gave different speeds. So 24 possible different speeds all the way from 1/30th to 1/1000th.

So perhaps to get the egg shape wheel, the camera should stay still, or pan slightly slower than the moving vehicle. If the camera were held static, the "people" wouldn't be angled in the photo I linked to.

[Stephane]

This is as explained by Leslie's Stroebel "view camera techniques" on page 76: first he used a focal plane shutter, then he panned.
The car is not distorted because Lartigue panned and the wheel is elliptical because of its rotation. The panning caused tilted of the people and vertical poles in the opposite way of the elliptical wheel.

[Steven Barall]

I think the car body is more distorted than you think it is. Also, because you can freeze the forward motion of the car doesn't mean that you can freeze the rotation of the wheels especially with a spoked wheel where the rotation is obvious. Good luck. I look forward to seeing the photos you make.

[Gordon Moat]

First is focal plane shutter, but not the kind that move from one side of the frame to the other. The idea is that the top of the image frame gets exposed prior to the bottom of the image frame; while it would work the other way, your egg shapes would go the other direction. Next step is panning, and this is where shutter speed becomes important. To get a regular pan blur effect, with little distortion, I often use 1 over the speed of the vehicle as a guide. Drop the shutter speed even slower, and more blur happens.

Here is an example of a soapbox derby racer, shot using a hand-held Nikon. It is tougher with slower vehicles, because the shutter speed is really slow. This makes it tougher to hand hold and get enough detail. Ideally you would have a tripod that allows panning with the action, but keeps the camera level.

Ciao!

Gordon Moat Photography (http://www.gordonmoat.com)

[jnantz]

jim

you might need to have your even slower
hold the camera upside down and gravity
will help slow it down below 1/30 ...

good luck!
john

[Jim Galli]

Here's (http://www.kettering.edu/archives/wilkerson.jsp) another one. This time the photog obviously wasn't panning as the grandstand is straight.

[Chauncey Walden]

Jim, I played with a few numbers and I don't think your T will have to over exert itself to do this. If you fix your camera position so that the wheel will subtend about 1 inch on your 5x7, my guess is that the T will only have to be doing 12 to 15 mph. No panning required. He was panning in his first pic, but not quite as fast as the car. If you actually measure the travel time of the curtain on the Speed we can figure it out pretty close. I was using 1/5 sec and 6 to 8 inches displacement on the wheel. It varies with your tension setting, of course, and you would want a small slit for shortest exposure. I hate to use shutter speed here as that (travel time) only varies by the tension while the exposure varies by the slit.

[Tracy Storer]

Knee-jerk = pan a bit too slow...remember, the image is upside down in the camera, so the bottom of the image is exposed first in a top-down shutter.

[Gordon Moat]

Here's (http://www.kettering.edu/archives/wilkerson.jsp) another one. This time the photog obviously wasn't panning as the grandstand is straight.

The slit of the shutter travels from the top of the camera to the bottom. The bottom of the picture was exposed first, and as the car traveled forward, the rest of the car was exposed. Notice that the radiator is also tilting forward.

As a start, I would suggest 1/speed (mph) of the vehicle. So if a vehicle is traveling 30mph, then your shutter speed would be 1/30 second. After that you can decide if going faster or slower shutter speed would improve the effect.

The only reason to pan would be to keep the object in the frame of view. As a vehicle goes faster, there will be less of a chance of capturing it with a slow shutter speed. Also, the wheels on those old cars were much taller than modern automobiles, so the effect is very exaggerated. You can notice the oval wheels on pictures of modern vehicles in motion, but the result is not as pronounced.

[Jim Galli]

Here's my math and an example:


http://tonopahpictures.0catch.com/H.A.M.B./EggShapedWheels.jpg

The car is moving through the frame at 50 mph which is 880 inch / second.

The camera is a 5X7 Speed Graphic.

The tire on the vehicle in the photo (this is a photo-shop rendering) is approximately 5/8 inch tall on the film plane.

Effective shutter speed on the SG is set at 1/500th. The slit that travels in the curtain shutter is 1/8th inch. There are 40 1/8th's in the 5 inch travel, or 40X1/500th. So the speed of the curtain from top to bottom is 1/12.5th of a second.

Since the tire is 5/8ths inch tall in the plane, it gets 5/500ths of a second of movement during the shutter's travel, or 1/100th sec. total time that some part of the tire will be painted on the film.

So the top of the tire will have traveled 9 inches approx. further than the bottom of the tire during the 1/100th sec it is getting exposed. Straight surfaces like the car's radiator and front windshield will appear to be leaning forward whatever that angle is.

Does that make sense class?

[Chauncey Walden]

Yes, but is it safe to get that Phaeton up to 50mph?!?!?! How about 25mph and 1/250? Safer for the driver;-) and easier synching with the photographer.

[Jim Galli]

After getting 'safe' rammed into every orifice all day long at work, it gets to be a 4 letter word. Still, there ain't much to run into out on the dry lakes bed. Except maybe the photographer, but I'm fast for an old guy. ;)

[Glenn Thoreson]

Jim, if you pan with the camera, the wheels won't go egg shaped. That shape depends on forward movement to make it work. I have seen pictures of a bicycle rider done with a Speed Graphic at 1/60 that gave really wild elongation. If you use a too fast shutter speed it will stop all the motion and spoil the result you're after. I would try 15 MPH in the car and start with 1/30 for shutter speed for a start. The tricky part is to not get the car so motion blurred that you can't tell what it is. The old info I have says the effect happened at up to 1/60 +/-. What it doesn't say is how the speed affects it.

[Chauncey Walden]

Hey Jim, I know you are fast. I used to be fast when I was young but now my wife says I am just half fast! At least, I think that is what she meant. I'm thinking there may be one glitch in your computation. In order to expose 1/8 inch of film for 1/500, the slit has to travel 1/4 inch (1/8 to fully open and 1/8 to fully close) so the total travel time would be 1/25. Does that make sense? Thinking about it again at this late hour, no. The curtain only has to move 1/4 inch to expose the first 1/8 inch. At the end of that time, it is fully open for the second 1/8 inch and so on. Man, it is too late and I am only half fast....

[Jack Dahlgren]

To make it easier to think about imagine the curtain slot is the same size as the film. If you want a 1 second exposure the curtain has to travel the entire distance in 1 second, then the back edge will stop the exposure. The curtain has to travel for a total time of two seconds because the back edge is a second behind, but it is traveling a distance of twice the film width. With a 5x7 this would be 5 inches per second. Now if you made the opening in the curtain half the size, the exposure would be 1/2 a second, but the speed would still be 5 inches per second. Cut the opening to 1/8th and the exposure is 1/40th of a second and the curtain is still moving at 5 inches per second. Want that exposure to be 1/400 and the curtain needs to move 10 times as fast. 1/500 would be 12.5 times as fast and the curtain moves at 62.5 inches per second.

[Mark Sawyer]

In Jacques Lartigue's (http://www.cracow-life.com/poland/jacques-henri-lartigue) famous early racing car pictures, I understand the people leaning because I think that J. H. L. was panning the camera hand held, so anything stationary would be at a different place as the shutter moves down, and if he is in fact 'panning' I can see that the car would be reasonably sharp at an effective 1/1000th second, but why are the wheels egg shaped? Why does the camera see them any differently than the car body?


I think everybody pretty much explained the oval shaped wheels. Sometimes they pick up an egg-shape when the car is passing at an angle and the wheel is a bit farther away (smaller" during the latter part of the exposure. In Lartigeu's photo, I'd guess he was panning about half the speed of the car, so the background leaned one way and the car leaned the other way.

Jim, I'd experiment with holding the Speed Graphic upside down for a couple of exposures too, to see how it looks with the car leaning the other way. Or you could drive the Model T in reverse at 50mph... :D

The timing will be the hard part, since to get the effect of the car leaning and the background not, the camera won't be panning as the car goes past...