# Jacques Henri Lartigue and his camera

Printable View

Show 50 post(s) from this thread on one page
Page 3 of 6 First 12345 ... Last
• 3-Jan-2008, 12:55
Scott Davis
Re: Jacques Henri Lartigue and his camera
I think it is also, just less obvious since we're only seeing 1/4 of it, if that.
• 3-Jan-2008, 13:14
Struan Gray
Re: Jacques Henri Lartigue and his camera
Quote:

Originally Posted by Lindolfi
there is no "residual distortion".

So why does your red line meander about so much within the white zone?

I once tried to fit an ellipse to the outer and inner black-white boundaries. I couldn't do it without invoking a change in speed for the car (or an accelleration of the shutter curtain, but that should show up in the background figures too).

I think Lartigue took the photo near a corner.
• 3-Jan-2008, 13:19
Ole Tjugen
Re: Jacques Henri Lartigue and his camera
Quote:

Originally Posted by Struan Gray
So why does your red line meander about so much within the white zone?

My first guess is that the young photographer didn't manage to keep an absolutely perfectly even panning speed. Does anyone think they could do better?
• 3-Jan-2008, 14:01
Lindolfi
Re: Jacques Henri Lartigue and his camera
Struan, fitting an ellips is something different from fitting a skewed ellipse. They are not the same mathematical shapes. A skewed ellipse has its minor and major axes at another angle than 90 degrees. A skewed ellips is the correct shape, since the wheel is deformed by the convolution of a moving line element (the slit) with a moving circle (the wheel). I think the fit of the skewed ellips is very good, given the noise in the photograph. Meandering has to do with the limited number of pixels present. Originals here look very clean indeed.

The fact that the wheel is not as sharp (not even the rim) as the rest of the car, suggests that the uneven road made the wheel vibrate. An acceleration of the car would have generated a systematic change in shape, not a random one.

If we assume that the shutter took about 1/5 seconds to move from bottom to top of the image, the shutter took 0.06 seconds to cover the hind wheel and so in order to produce a 10&#37; deformation of the rim from the perfect skewed circle, the car would have to break so fast that it would stop in about 0.6 seconds to standstill, which is quite impossible. Besides: the race driver would not have been able to hold the steering wheel with these bent elbows and still maintain this seated position at such an enormous deceleration. Remember, there were no safety belts. We can safely remove this item from the list of reasonable explanations for any residual deformation of the wheel in the image. Besides, according to me, there is no measurable deformation from a skewed ellipse.

Ole, the panning speed was very constant: The panning angular acceleration of the camera can be found by assuming that the three poles or trees at the side of the road were straight. If the camera would have accelerated, all three poles would have been curved to the left (from bottom to top). The left pole is a little bit curved, but the other two are not, so the camera angular acceleration can be safely set to zero.

Scott, you are right in your reply to cowanw, there is skewing, but less so of the passsenger wheel. From the dimensions of the car (a Delage type F), I reconstructed the distance of the photographer to the car and the rate of angular velocity of the camera and found in a 3D simulation I wrote that the furthest hind wheel should indeed by less skewed than the closest hind wheel. The angular change is less for that furthest wheel because of the rotation of the camera and the linear motion of the car.
• 3-Jan-2008, 14:45
Struan Gray
Re: Jacques Henri Lartigue and his camera
I used a skewed ellipse, of course. The outer rim of the white band gives an accurate enough marker to see a systematic error if you assume a linear motion of the wheel image across the focal plane.

A 10% distortion is fairly brutal. Much less is required to fine-tune the match. As you say, your linear fit is pretty good as it is.

If I remember rightly, a bigger problem is that any residual distortion can also be explained by perspective effects, and there is not enough information in the image to fully elminate those. The fact that the other rear wheel is less distorted, and that the top of the wheel is less thick than the bottom, suggests a wide-angle lens used from a low-viewpoint, but without specifics you end up well into angels on pins territory.
• 3-Jan-2008, 15:05
Lindolfi
Re: Jacques Henri Lartigue and his camera
From the difference in size of the two hindwheels and the width of the Delage Type F, I calculated a focal length of between 210 and 240 mm on 4x5, so that is longer than normal at that filmsize. Viewpoint height was at about 0.8 meter given the perspective and the dimensions of the car. Moreover, the car moved almost parallel to the plane of the film (horizon at half the photograph and fanning of lines on the road), so any wide angle deformation is not to be expected from a wide angle lens.
• 3-Jan-2008, 15:11
Ole Tjugen
Re: Jacques Henri Lartigue and his camera
Quote:

Originally Posted by Lindolfi
Ole, the panning speed was very constant: The panning angular acceleration of the camera can be found by assuming that the three poles or trees at the side of the road were straight. If the camera would have accelerated, all three poles would have been curved to the left (from bottom to top). The left pole is a little bit curved, but the other two are not, so the camera angular acceleration can be safely set to zero.

But since the poles (and people) only go down to the middle of the picture, we have no basis of comparison for the lower half of the wheels?

BTW, I believe most 1912 ICA's took 9x12cm glass plates and not 4x5".

While there are several different lenses that may have been fitted to it, 99% of all were 135mm with a few special order 120mm's and 150mm's.

An f:4.5 lens in those days would be either a Tessar-type, a triplet, or a dialyte.
• 3-Jan-2008, 15:17
Lindolfi
Re: Jacques Henri Lartigue and his camera
Thank you Ole. 9x12 is correct ofcourse, also according to McKeown's. The length is nearly the same as 4x5. However, if you are right about that 135 mm (still a normal length for 9x12 and not wide angle), the image we look at must be a crop. How can we know this is true and if so, why was the image cropped?
• 3-Jan-2008, 15:28
Ole Tjugen
Re: Jacques Henri Lartigue and his camera
The proportions of the image look right for 9x12cm, less 3mm for edges. I'll have to look closer when I get home and can compare to an original 9x12cm plate (exposed by myself last year).

Or could it be film? That's another few mm off the size?

A third possibility is tha the camera could have been fitted with a different focal length lens, but I somehow find it difficult to believe in anything longer than 180mm at most.
• 4-Jan-2008, 00:43
Lindolfi
Re: Jacques Henri Lartigue and his camera
Here is a correction concerning the situation in which Jacques Henri Lartigue made the photograph:

According to "The Camera" published by Time Life International in 1971, the photograph was made in 1912 during the Grand Prix in Dieppe and the car would be a Delage. However, the number 6 on the car and the type of car shows that this is not the case: The photograph was taken in 1913 in Amiens during the Grand Prix and the driver was Ren&#233; Croquet, driving a Th&#233;ophile Schneider car. In the race he finished 10th (reference).

According to the title Lartigue gave to the photograph "Courses de voitures, Papa &#224; 80 km/h, 1913", the passenger must have been his father. When the estimate of 80 km/h is correct, we can back calculate the travel time of the focal plane shutter, which would be about 0.07 sec at a vertical dimension of the plate of 90 mm and assuming an uncropped photograph.

This site
, shows a number of photographs with a similar deformation of the wheels and also another photograph of the same car during the same race (6th photograph from the top)
Show 50 post(s) from this thread on one page
Page 3 of 6 First 12345 ... Last