|
Distortion in Henri Lartigue’s race car
photograph.
It is all relative Andrew Davidhazy thoughts on the matter  To begin with: In Henri Lartigue’s camera the slit, focal plane shutter, moves from top to bottom but due to optical inversion the car’s image was upside down within the camera and so the bottom of the vehicle was exposed by the moving slit before the top of the car. The car and its occupants are
reproduced slightly leaning forward in
the direction of motion because Lartigue did not exactly match the
panning speed of the camera to keep thier image still on the film. So
as
the shutter slit passed over their image their recorded position on the
film keeps shifting in the direction the camera was moving.
As for the distorted wheel. If the car had been standing still even if the wheel were spinning the wheel would have reproduced as a circle. Assuming the camera was not panned. However, in this case since it took time for the slit to travel over the rotating wheel as the car moved forward, the hub of the wheel (and consequently the car) moved slightly away (forward) from the point where the slit first made contact with the rim of the wheel. By the time the slit reached the top of the wheel the point of contact of the slit with the rim of the wheel was now forward from where it had been when the slit reached the bottom of the wheel. Since the hub, by virtue of it remaining essentially fixed on the film and the points of contact being behind the hub and later in front of the hub result is an oblong shaped figure leaning in the direction of motion of the vehicle. This leaning effect results in ever greater leaning as the speed of the vehicle increases.
Car and hub (!) moving left to right
 Wheel and hub stationary Wheel rotating and hub moving Wheel rotating and hub moving faster As for the wheel, during the time it takes the moving slit of the shutter to go from the bottom to the top of the image of the wheel, the contact points (red stars in drawing) are in different locations on the film relative to the hub (and body of the vehicle) and the circular shape becomes oblong leaning in the direction of motion of the vehicle. What about the poles and people in the background. Well, that is more easily explained. Again realize that the scanning action of the shutter was from top to bottom (at the image plane the shutter slit exposed the feet of the people first and later the heads. So while the camera was panned with the car the full lengths of images of the subjects get recorded later and later as the slit recorded those subjects closer and closer to the edges of the frame. Note that these illustrations were made using the moving sensor of a flat bed scanner to stand in for the focal plane shutter of a camera equipped with one. As the scanner wand moved I moved a circular subject perpendiculat to its scanning direction. As for Lartigue's camera references are made to an ICA 4x5 glass-plate camera.
Note that the uphill look to the Lartigue photograph is misleading. Most likely the surface was level. After leveling the road surface the top of the photograph was pulled left while the bottom to the right. This caused the number 6 on the tank to be perpendicular to the ground. That is how it was on the car. That also brought the top of the wheel, the contact point of the pavement and the middle of the hub to lie perpendicular to the ground. Because the car had passed by Lartigue a bit one can see the farther tire offset from the closer one. Essentially looking at the car at an angle. This is the reason that the line from top to bottom of the wheel does not exactly line up with the center of the axle. It lines up with the center of the hub. If Lartigue had panned at the right speed the result might look like this. (without the vertical lines!) In this "corrected" version the wheel is pretty much round but the people and poles are inclined at a steeper angle than in his photograph. The axle seems off-center but that is because the photograph was taken at an angle to the wheel.  And if he had made the photograph with, let's say, a Brownie (leaf shutter camera) the result might look like this. Poles and people upright and wheel circular but there would probably be some horizontal blur unless the Brownie had an extremely short exposure time which is ulikely. As you can see there are many items in motion relative to each other and this analysis is not precise but more or less something to consider. More to come later … maybe.
Andrew Davidhazy. Prof. (ret. RIT) write to me at: andpph@gmail.com my website is here: www.andpph.com |