Thoughts to ponder

[george013]

One more. In another thread with about the same subject, the portrait lens was mentioned. Why is a certain focal length yes or no suitable for portraits? Look at the next pictures from the Dutch Canon site.
http://www.canon.nl/youconnect/learn...-a-lens-makes/

It's in Dutch, sorry. But I'll link to the photo in it. They've all the important info.



What you see is a difference in magnification between the fore site and the back site of the head. Background is not important. When the working distance is short, the differences between them will be big in relation to that working distance and so the differences in M. With the long lens the working distance will be more, the differences in the head will be equal but not to the camera, The differences in M are less.

George

[Stagecoach]

You see the light but have some remarks about the switch?

That comment is totally confusing me George. What light have I seen and what switch are you referring to with regard to my post No 39?

In Richards example #3 it would mean that the camera has moved forward 193mm. Correcting that by moving the camera backwards would have enlarged this difference in M. With an amount you probably won't see.

Not sure what your reference to Richards example No 3 is referring to or your reference of 193mm?

Do you correct your dof-tables with that information.?

Not sure again what information you are referring to? Post No 39 makes no reference to DoF and is not intended to demonstrate anything with respect to DoF.

If there is something in post 39 that you consider incorrect can you explain exactly what it is so that I can change things so others are not misled.

[george013]

That comment is totally confusing me George. What light have I seen and what switch are you referring to with regard to my post No 39?

I'm sorry. I reread your post more carefully. You didn't see the light yet.

and I believe the FL/AOV of the lens used at any distance will affect the ratio in size between near and far subjects recorded on the sensor additional to that ratio variation which is due to sensor to subject distance.

It's about the difference in M of different subjects due to their differences in distance to the camera.

Not sure what your reference to Richards example No 3 is referring to or your reference of 193mm? .

Your second drawing. If the first picture was taken with 17mm and the third with 200 there's a difference of 183. Excuse for the typo. The lens is going forward as you know seeing your 2nd image. To keep the distance equal you must correct the camera by moving backwards with 183mm. I draw a simple diagram, even half the aov used by the camera, to show what happens.

.

Not sure again what information you are referring to? Post No 39 makes no reference to DoF and is not intended to demonstrate anything with respect to DoF.

If there is something in post 39 that you consider incorrect can you explain exactly what it is so that I can change things so others are not misled.

It's the acceptance that the distances are not corrected with the change of focal length. Except for makro.

George

[Stagecoach]

I'm sorry. I reread your post more carefully. You didn't see the light yet.

That is a matter of interpretation.

It's about the difference in M of different subjects due to their differences in distance to the camera.

And where does it say in my post 39 that the differences in ratio between the subjects projected onto the sensor, (or you can convert ratio to magnification if you wish), are not caused by the difference in distance to the camera. Surely Drg 1 demonstrates that fact very clearly

Your second drawing. If the first picture was taken with 17mm and the third with 200 there's a difference of 183. Excuse for the typo. The lens is going forward as you know seeing your 2nd image. To keep the distance equal you must correct the camera by moving backwards with 183mm. I draw a simple diagram, even half the aov used by the camera, to show what happens.

My Drg No 2 has absolutely nothing to do with Richards images. What Drg No 2 clearly shows as far as I can see is that if the camera sensor remains at a fixed distance to the subjects (camera on a tripod), changing its FL (example of zoom lens) will affect the size ratio between near and far subjects projected onto the sensor.

In other words, suggesting you can not take these images and crop them in post to be identical.

It's the acceptance that the distances are not corrected with the change of focal length. Except for makro.

So you are unable to say what is incorrect in post 39?

[Rob Ekins]

My brain hurts....

[george013]

Calculate the difference in M between fig.1 and fig.2. It might be easier to use 2 subjects of the same size. Just to keep it simple.
And if you draw the lens ax you can use simple geometric formulas. Even more simple.
Like in post 32. Reread post 32.

George

[Stagecoach]

Calculate the difference in M between fig.1 and fig.2. It might be easier to use 2 subjects of the same size. Just to keep it simple.
And if you draw the lens ax you can use simple geometric formulas. Even more simple.
Like in post 32. Reread post 32.

George

Geroge, Why?

The Drg No 1 Figs 1 and 2 you refer to show EXACTLY what happens when the camera is moved away from the subject. The ratio of size between S1 and S2 (subjects of different sizes and distances) CHANGES at the sensor !

There are no sizes involved, there is no need for the lens axis to be included or geometric formulas to be calculated to explain what is very simply shown in the diagram.

The drawings and what they show are correct, they demonstrate what has caused the 'compression' effect in a practical way.

[pnodrog]

An online calculator for a 2m subject at 100m compared to a 20m subject at 250m taken from same view point with a 50mm lens and a 250mm lens shows no difference in the ratio between the resultant image sizes. The ratio between the two resultant heights in the images did not appear to be dependent on focal length. Any difference was less than 0.2% which was either rounding errors or minuscule result of focal length but certainly smaller than any observable effect.

Note: Subject ratio is actually the image height ratio. Smallest over largest.

[Stagecoach]

An online calculator for a 2m subject at 100m compared to a 20m subject at 250m taken from same view point with a 50mm lens and a 250mm lens shows no difference in the ratio between the resultant image sizes. The ratio between the two resultant heights in the images did not appear to be dependent on focal length. Any difference was less than 0.1% which was either rounding errors or minuscule result of focal length but certainly smaller than any observable effect.

The difference will be very small Paul and also decrease as focal length extends I suspect. But in theory the diagram suggests there is one, if the focal point changes in relation to the sensor.

I'll feed the figures in for shorter distances and see if anything shows.

Edit : I have fed the figures in for shorter distances and the results show no difference in the ratios.

So what's wrong with Drg No 2 ? Or is it that the calculators use the focal point for subject distance OR the difference is so small it is ignored?

[george013]

The difference will be very small Paul and also decrease as focal length extends I suspect. But in theory the diagram suggests there is one, if the focal point changes in relation to the sensor.

I'll feed the figures in for shorter distances and see if anything shows.

Edit : I have fed the figures in for shorter distances and the results show no difference in the ratios.

So what's wrong with Drg No 2 ? Or is it that the calculators use the focal point for subject distance OR the difference is so small it is ignored?

Once more. I wrote this formula several times already.
M=di/ds a common formula.
Given 2 subjects with a distance x between them and the camera focused on the first subject will give:
M1=di/ds
M2=di/(ds+x)
When you shoot at a closer distance ds will be less and the x-factor will have a bigger influence. A big difference in magnification between them. A typical wa issue as said. But not just for the wa, but for the distance being used.
When you shoot at a further distance, ds will increase. The influence of the x-factor will get smaller. Meaning the difference in magnification of both subjects will getting less.


What's wrong with fig.2. Nothing, but I it doesn't show the differences in M for the 2 subjects, compared with fig.1. Or fig1(M1/M2) compared to fig2(M1/M2).

If you've problems with numbers, tell us so we can find another way.

George

[pnodrog]

George your post #32 is a diagram basically illustrating perspective and the change in ratio of the subject in relation to it's background. As you illustrate the background appears bigger in relation to the subject as we move further away from the subject. However in a normal photograph that subject in the distance (and it's relationship with the background) is usually no longer of importance as it has been replaced with a closer subject/point of interest which has a more natural looking relationship with it's background. However once we crop or zoom extensively to restore the original subject to again become the main point of interest we become far more aware of it's relationship to the background and some people use the term compression to describe it.

The odd relationship that distance items have with the background exists in most photographs but we seldom zoom in (enlarge and view closely) enough for it to look unusual. In most cases they lurk out of interest to our unaided vision.

[dem]

What's wrong with fig.2. Nothing, but I it doesn't show the differences in M for the 2 subjects, compared with fig.1. Or fig1(M1/M2) compared to fig2(M1/M2).

It does, George. Where it says "S2 = 1.6 times S1" for FL/AOV1. Here 1.6 = 40.0 / 25.0 is the ratio of the two image sizes produced by lens 1 + same for lens 2. One of the images is going to be blurred but this only becomes an issue when you try combining this diagram with the definition of magnification M = FL / (lens-object distance - FL) which is defined only for images in focus.

Note that in this diagram both subjects are very close to the camera. If we say that the lenses are 5 cm and 10 cm away from the sensor (note these are not their focal lengths, their focal lengths will have to be 40 mm and 60 mm for the first subject to be in focus), then the subjects are about 25 cm and 30 cm away from the sensor.

Assuming that we can focus both lenses at these distances and that DoF is wide enough to accommodate both subjects 5 cm apart, we still get only 6.7% difference in magnification ratios (1.60 vs 1.50). It is more than 0.2% we got for images in the opening post, but it is still small.

For all practical purposes this effect can be ignored because it is going to be masked and overpowered by other effects that influence our perception of depth - perspective distortion controlled by the camera-subject distance, increased blur due to a tighter framing, inclusion of more surrounding objects due to a wider AoV.

[george013]

It does, George. Where it says "S2 = 1.6 times S1" for FL/AOV1. Here 1.6 = 40.0 / 25.0 is the ratio of the two image sizes produced by lens 1 + same for lens 2. One of the images is going to be blurred but this only becomes an issue when you try combining this diagram with the definition of magnification M = FL / (lens-object distance - FL) which is defined only for images in focus.

Note that in this diagram both subjects are very close to the camera. If we say that the lenses are 5 cm and 10 cm away from the sensor (note these are not their focal lengths, their focal lengths will have to be 40 mm and 60 mm for the first subject to be in focus), then the subjects are about 25 cm and 30 cm away from the sensor.

Assuming that we can focus both lenses at these distances and that DoF is wide enough to accommodate both subjects 5 cm apart, we still get only 6.7% difference in magnification ratios (1.60 vs 1.50). It is more than 0.2% we got for images in the opening post, but it is still small.

For all practical purposes this effect can be ignored because it is going to be masked and overpowered by other effects that influence our perception of depth - perspective distortion controlled by the camera-subject distance, increased blur due to a tighter framing, inclusion of more surrounding objects due to a wider AoV.

I'm referring to fig.2 in dr.1
Dr.2 is showing the effect how the focus distance can change by only changing the focal length and not the camera. That's in Richards example 200-17=183mm. Graham is mentioning height of the image. I'm totally not interested in that. What I want to investigate is the change in magnification of 2 subjects with a distance x between them. The shown drawings are useless for that. I need an explanation.

Paul,
I just want to state that what is concerned as a wide angel effect or a long focal length effect is essential an effect caused by distance. Angel of view or focal length is no issue. How the photographer wants to deal with it is up to him/her. Both extremes can be used for an artistic effect. But it would be wise if he/she knows what causes that effect.

George

[johngunkler]

I thought"compression" meant the comparison of apparent distance between foreground and background objects between shots taken with different focal length lenses. No?


Sent from my iPhone using Tapatalk

[JohnRostron]

I thought"compression" meant the comparison of apparent distance between foreground and background objects between shots taken with different focal length lenses. No?

This assumption was the basis of my previous mathematical analysis. I also assumed that the 'perception of distance' would be approximated by the magnification (which is easier to quantify).

John

[george013]

This assumption was the basis of my previous mathematical analysis. I also assumed that the 'perception of distance' would be approximated by the magnification (which is easier to quantify).

John

John and John,

Read post 32 and 41.Everything has been said.

A spreadsheet


Examine the range 10 till 100. They have a constant increase of 10m. M(d20) is 50% of M(d10), but M(d100) is 90% of M(d90). All with a difference of 10m.

Be aware it's not about magnification but about the difference in magnification between 2 objects with a fixed distance at different camera distances.

George

[johngunkler]

This assumption was the basis of my previous mathematical analysis. I also assumed that the 'perception of distance' would be approximated by the magnification (which is easier to quantify).

John

Meaning the image size of a background object on two photos where the foreground object stays the same size, in the simplest case?


Sent from my iPhone using Tapatalk

[Stagecoach]

I thought"compression" meant the comparison of apparent distance between foreground and background objects between shots taken with different focal length lenses. No?

Yes

With the proviso that it is understood that the "compression" (or one of the myriad of terms used to name this phenomena) is caused by the distance viewed/captured by the camera from the subject/subjects.

The "different" focal length lenses used do not 'cause' the phenomena but allow you to control your subject size. E.g. we are not going to want to take a picture of a model from 50m away with a short FL lens because the model will be tiny in the image so we use a much longer FL lens.

Meaning the image size of a background object on two photos where the foreground object stays the same size, in the simplest case?

Yes

And here's a link to one of the many examples on the web https://www.slrlounge.com/lens-compression/.

The technical reason for this phenomena is very simple and what basically consists of a couple of lines taken from any subject within the framing through a focal point within the lens/camera onto the sensor can be scrutinised and formulas concocted to provide numerical figures that define any unknown or relationship between any points in a number of different ways

[pnodrog]

The technical reason for this phenomena is very simple and what basically consists of a couple of lines taken from any subject within the framing through a focal point within the lens/camera onto the sensor can be scrutinised and formulas concocted to provide numerical figures that define any unknown or relationship between any points in a number of different ways

Well put. But I do hope my life never depends on my understanding this statement....

[george013]

Watch this picture.
http://www.irishtimes.com/news/world...tion-1.3082936

George