Carl Koch "Photo Know How" - Errors

[acrobatic_citron]

Doing a first pass through Carl Koch's well regarded Photo Know How book where you do the exercises and assignments. Chapter 3.5 in the single later edition he states

The in other words part does not match the equation... Then in his worked example

If you plug his worked examples back into original formula

25/37.5 =0.667 0.6666 which does not equal 30/20 (1.5). He's flipped one side of the equation.

Viewing distance :25cm
Subject Size :30cm
Image size :20 cm( inferred from equation)
Subject distance : 37.5cm (from equation)

Am I going bonkers or is it wrong? Is his calculation wrong or the equation or me?

 

[Don_ih]

From this:

subject distance = (viewing distance x image size) / subject size

So, I guess it's wrong?

 

[acrobatic_citron]

I think so too, but which is wrong? The stated equation or his workings ha, I could take the pictures with both subject distances and probably work it out that way but wonder if anyone will save me all that effort

The co author J Jost is still alive and, active on instagram so have messaged to see if he can provide any clarification

 

[reddesert]

Don't worry about the formulas. Use your brain and think about the problem. The screenshot of the example is missing a bit of information because the image size is only implied, but I can assume it from the context.

Take the first example of the (magazine-size) advertisement. The viewing distance of the ad is 25 cm. The image size of the bottle in the ad is 20 cm (what 20 cm means is not specified in the screenshot, but clearly that's meant to be the image size of the bottle in the ad). So you are viewing the ad from a slightly larger distance than the image size.

The actual size of the bottle is 30 cm. Thus, to keep the perspective the same between the photo setup and the viewed image reproduction, you need to put the camera (actually, the lens) slightly farther away than the actual size. This is 30 * 25 / 20 = 37.5 cm.

For the poster example, the viewing distance is 400 cm and clearly he intends the image size to be 80 cm. So you need to put the lens 30 * 400 / 80 = 150 cm away from the bottle.

Whether you have to follow this distance recommendation literally is up to you. There is not a literal requirement to have the exact same perspective in viewing, but it can be useful.

 

[acrobatic_citron]

He doesn't provide much more context tbh haha, he never states Image size so we must assume it's the 20 and 80 as logically thats what fits.

Thanks for taking the time to go through it

 

[MattKing]

Thread Title tweaked.

 

[Ian C]

I too thought that the author’s intention was to maintain the same perspective at two different viewing distances and image sizes.

To maintain the same perspective, the ratio of image height/viewing distance must be constant. This also means that we get the same angle of view in both situations.

Case 1.

Image height = h1 = 20 cm

Viewing distance = v1 = 37.5 cm

h1/v1 = 20 cm/37.5 cm = 0.53

Angle of view = 2*arctan(h/2v) = 2*arctan(20 cm/2*37.5 cm) = 29.9º


Case 2.

Image height = h2 = 80 cm

Viewing distance = v2 = 150 cm

h2/v2 = 80 cm/150 cm = 0.53

Angle of view = 2*arctan(80 cm/2*150 cm) = 29.9º


The aim of the calculation is to determine the second viewing distance v2, given h1, v1, and h2.

I don’t understand the use of the original size of the object in the calculation. It seems an unnecessary complication.

All we need is v1/h1 = v2/h2

Then, knowing v1, h1, and h2

v2 = v1*h2/h1

v2 = 37.5 cm*80 cm/20 cm = 150 cm

This doesn’t match the original setup in which the IMAGE of a 30 cm tall bottle is viewed from 25 cm—unless the image of the bottle is about 13.35 cm tall in the printed advertisement.

If we view a 20 cm tall image on a printed advertisement from a 25 cm viewing distance, the angle subtended is 43.6º. This is not addressed in the information shown from the book in post #1.

 

[acrobatic_citron]

I too thought that the author’s intention was to maintain the same perspective at two different viewing distances and image sizes.

This sounds correct, on the previous page he states a rule of:

'A print appears in correct perspective only when the viewing perspective - in terms of angles subtended by various parts of the subject - matches the taking perspective'

Yeah he did not define image size very well imo, some have assumed is the size of the subject on the print, myself the size of the print medium along the long edge

 

[reddesert]

I think he means the size of the photographed object, such as the bottle, in the reproduced image.

His goal, which is stated but perhaps not clearly, is that the perspective (relative dimensions) of the object should be reproduced accurately for the viewer, from the viewing distance that is likely given the image medium. So in the example of a poster, he assumes that you're going to be looking at a poster from 4 meters away, and that the size of the bottle image in the poster is 80cm high. The physical bottle is 30cm high. To maintain the perspective, you should put the lens at (30/80) * 400 = 150 cm from the bottle when photographing it.

If for example you put the lens/camera a lot closer than 150 cm, the photo would be looking down into the bottle bottom and up into the bottle top, more than the viewer of the poster would normally see. This is kind of a subtle point and obviously only applies to 3-d objects - if you were photographing a 2-d subject onto flat film, it wouldn't matter.

These kinds of perspective distortions are useful to think about, though. This is why the usual advice for photographing people (especially headshots) is to use a short telephoto such as 85-105mm on 35mm film - it makes you stand further away, compressing the subject's perspective. If you use a wide angle lens and stand close, it exaggerates the 3-D structure of the face, which of course we've all actually seen when standing close to a person, but our brains compensate for it in a way that is hard to do when looking at the 2-D photo.