Parallax Compensation - What's the Point?

[Vlad Soare]

Advertisement (gone after registration)

I know, the title sounds silly, almost like heresy. Of course a rangefinder camera will suffer from parallax error, so having something that compensates for that is surely at least desirable, if not necessary, isn't it? But bear with me for a moment.
When the lens is focused at infinity, the framelines occupy a certain position within the viewfinder. As you focus closer and closer, the framelines move towards the bottom right. Since the closer an object placed in front of the lens is to the lens the more obliquely it will be seen through the viewfinder, this adjustment seems to make perfect sense. But does it really?

Let's say I'm looking at a landscape with some point of interest in the foreground.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

When the lens is focused at infinity - say, on the trees - the viewfinder looks like this:

That's all nice and well, the framing is perfect. But I would like the object in the foreground - i.e. the little girl - to be in focus. And if I turn the focus ring from infinity to one meter, the viewfinder will now look like this:

Which is wrong, isn't it? If the position of the camera hasn't changed, then what's actually captured on film won't change either, will it? Why should I capture less of the house and more of the grass just because I refocused the lens? 

What am I missing?

 

[thomas_schertel]

This is the problem with rangefinders. But in realty the parallax matters only under 2 metres. You will stop down to make this photo. Maybe to 11 or so. With the 50 mm lens set on 5 metres all is sharp from 2,5 to infinity. The parallax compensation is now valid for 5 metres. If you place the girl, the house and the trees not to the margins of the viefinder frame, all should be well.

Yours sincerely
Thomas

Edited August 23, 2021 by thomas_schertel

[a.noctilux]

Vlad,

only short/or longer practices can 'explain' those weird things of rangefinder Leica M.

Don't forget that rangefinder Leica since start was created for 'dynamic photography' (1913-25), and continue with Leica M happily.

 

Not really related but as 'user decision' in one situation, here, parallax/framing/to render the most of subject of the picture 👇.

Just anecdote while we were talking about 50/75 lenses in the other thread M-A.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

Here, I had with me (and my wife !) M6, M5 with two lenses Noctilux 1/50 and Summilux-M 1.4/75 and slow film 25 or 64 ISO.

Start framing with 50mm with loose 'framing' and too much perturbating background, even at f/1 or 1.4, displacing myself a bit,

happily, I had the 75mm which as the 50 hides a large bit of framing, so here the parallax/tight framing can be hard to come by,

as the background was busy and a couple of cm at left or right and/or up/down can ruin the subject.

That was the rangefinder challenge, in turn with SLR this would be easier.

[pop]

vor 2 Stunden schrieb Vlad Soare:

If the position of the camera hasn't changed, then what's actually captured on film won't change either, will it

Indeed, it won't change. However, your viewfinder does not tell you where the camera (i.e. the lens which makes the picture) points to. It tells you where the viewfinder points to. There's no difference worth speaking of at a large distance. The closer you come to the object of interest, the more noticeable the difference will become.

[wattsy]

2 hours ago, Vlad Soare said:

I know, the title sounds silly, almost like heresy. Of course a rangefinder camera will suffer from parallax error, so having something that compensates for that is surely at least desirable, if not necessary, isn't it? But bear with me for a moment.
When the lens is focused at infinity, the framelines occupy a certain position within the viewfinder. As you focus closer and closer, the framelines move towards the bottom right. Since the closer an object placed in front of the lens is to the lens the more obliquely it will be seen through the viewfinder, this adjustment seems to make perfect sense. But does it really?

Let's say I'm looking at a landscape with some point of interest in the foreground.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

When the lens is focused at infinity - say, on the trees - the viewfinder looks like this:

That's all nice and well, the framing is perfect. But I would like the object in the foreground - i.e. the little girl - to be in focus. And if I turn the focus ring from infinity to one meter, the viewfinder will now look like this:

Which is wrong, isn't it? If the position of the camera hasn't changed, then what's actually captured on film won't change either, will it? Why should I capture less of the house and more of the grass just because I refocused the lens? 

What am I missing?

 

I think the mistake you are making is in thinking that refocussing on the little girl in your example would change the focus a great deal (and that the frame lines would move as much as you have indicated).You are also assuming that the scene as shown by the viewfinder generally is the one captured on film and that is not so. As Pop points out, the VF and the lens are not in line and therefore the VF is showing something different to what the lens is projecting on the film. This difference matters most for subjects in the closer focus range and so the frames move down and to the right to try and compensate for the difference in the VF position and lens axis.

Edited August 23, 2021 by wattsy

[harmen]

Put another way… if you ever forget to take off your lens cap, your entire shot will be black; not just the bottom right as your viewfinder would make you believe 😜

[250swb]

Advertisement (gone after registration)

Or put yet another way, you get used to it and learn to compensate for parallax at close distance.

[Helge]

vor 8 Stunden schrieb Vlad Soare:

I know, the title sounds silly, almost like heresy. Of course a rangefinder camera will suffer from parallax error, so having something that compensates for that is surely at least desirable, if not necessary, isn't it? But bear with me for a moment.
When the lens is focused at infinity, the framelines occupy a certain position within the viewfinder. As you focus closer and closer, the framelines move towards the bottom right. Since the closer an object placed in front of the lens is to the lens the more obliquely it will be seen through the viewfinder, this adjustment seems to make perfect sense. But does it really?

Let's say I'm looking at a landscape with some point of interest in the foreground.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

When the lens is focused at infinity - say, on the trees - the viewfinder looks like this:

That's all nice and well, the framing is perfect. But I would like the object in the foreground - i.e. the little girl - to be in focus. And if I turn the focus ring from infinity to one meter, the viewfinder will now look like this:

Which is wrong, isn't it? If the position of the camera hasn't changed, then what's actually captured on film won't change either, will it? Why should I capture less of the house and more of the grass just because I refocused the lens? 

What am I missing?

 

Did you move that picture during that test or didn‘t you?

Center the picture to your camera and move it from a wider distance to 1 m. (no movement left or right).

Now fo two tests:

1) leave the distance setting on you lens at infinity and see how the part you see will move it’s center 

2) now set the distance setting according to the real distance of the picture, the center shouldn’t move

Understood the flaw in your derivation?

It doesn‘t work if you don‘t look through the viewfinder and if you do not move the object 

[Michael Geschlecht]

Hello Vlad,

Welcome to the Forum.

To summarize a number of points brought up in a number of Posts above in this Thread:

A range/viewfinder camera uses a viewing & focusing system that is separate from the lens/film/sensor system that is capturing the image.

These 2 systems are physically in 2 different places.

The frames & the rangefinder patches that move in the viewfinder system are designed to adjust for the differences in what the lens/film/sensor captures & what the frames/rangefinder patch show in the viewfinder when the lens is set at different distances. This is why they travel from the top left when the lens is focused at Infinity to the bottom right when the lens is focused at its closest setting.

At optical Infinity (For example: The Moon in the sky.): The image is so far away that the viewing & taking systems are optically & mechanically parallel.

When the lens is focused at Infinity: The picture with the frames/rangefinder patch is at the top left.

As the lens is focused closer the viewfinder image of what the lens captures needs to be adjusted for the AREA OF COVERAGE seen in the viewfinder window: Which changes as the lens is focused closer than Infinity.

Because of the mechanical placement of the 2 systems:

 Looking from the front: The viewfinder window is above & to the right of the lens.

A distance of about 4.5 cm/1.75 inches.

 BECAUSE of this small physical distance between the center of the viewfinder image & the center of the image created by the lens.

There is a degree of movement from the top left to the bottom right, as seen in the viewfinder, needed for the range/viewfinder system, so that the image captured is what is being framed in the range/viewfinder system as the lens is being focused closer than Infinity.

Additionally: As the lens is focused closer the image captured is also increasing in size. This is NOT adjusted for in Leitz/Leica "M" cameras. Altho it has been adjusted for in some range/viewfinder systems in other makes of other cameras at various times.

This is a different discussion which can also be discussed.    

Best Regards,

Michael

 

 

Edited August 24, 2021 by Michael Geschlecht

[Vlad Soare]

Guys, I know how a rangefinder works. I know the viewfinder is just an approximation of what the lens sees.
What I don't get is why this approximation should depend on how the lens is focused.
I mean, if I put the camera on a tripod and leave it there, completely still, and all I do is rotate the focusing ring, the framelines will move. What's that supposed to tell me? As long as the camera is perfectly still, both the viewfinder coverage and the actual image projected by the lens will remain unchanged, won't they? So, how's the movement of the framelines supposed to help me? What's it supposed to tell me?

In my example, let's say I want to get the framing shown in the second picture, but I want to shoot wide open, with the girl in focus and everything else out of focus. If the camera were focused at infinity and positioned so that the viewfinder looked like that, then everything would be fine (except for the girl, who would be out of focus). Now, I move the camera so that the girl is in the center, then use the rangefinder to focus on her. Then I reframe. But when I reframe, the camera will end up in a different position than it was before, because the framelines I'm relying upon are in a different position in the viewfinder than they were before. How does this help me? 

I'm sure the answer is simple and is right under my nose, but I just can't seem to grasp it. 

 

Edited August 24, 2021 by Vlad Soare

[harmen]

It tells you if at least those objects that you have focused on will be in the frame.  The image will never look exactly how you see it in the viewfinder.  The framelines can show you what falls within the frame at the selected distance.  For example of everything that is at 1m away, where does it show in the frame.   Or for everything far away, say beyond 5m, where does it fall in the frame.  

[Helge]

@Vlad SoareYou left the object out of the equation, it moves together with the change of the focus setting.

Read my post before 😉

Edited August 24, 2021 by Helge

[harmen]

It might be easier to see when you look from above, like this.  While the lens always sees the same, you will notice that the frame lines in the viewfinder have to move left to indicate what the lens sees all along when you consider objects further away.  The nearby part is close to the right edge of the view through the finder, while objects further away show more to the middle.  At no point have they actually moved.  It’s just that the viewfinder can only properly provide a frame for a single selected distance at a time.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Also note how the perspective is truly different.  Through the viewfinder, the tree is right behind the person.  Through the lens the person is between the house and the tree. Edited August 24, 2021 by harmen

[Vlad Soare]

I see. Or at least I think I do. The framelines try to show what part of the focusing plane will (almost) certainly end up in the picture. The camera assumes that if I focus on something, that means I am interested exclusively in the plane that particular object lies in.
So, I guess my example (focusing on one plane while framing according to another plane) is an improper use case and isn't what Leica had in mind when they designed the system.
Right?

Edited August 24, 2021 by Vlad Soare

[pop]

vor 1 Stunde schrieb Vlad Soare:

What I don't get is why this approximation should depend on how the lens is focused.

It doesn't, actually. It adjusts to what the camera thinks is the distance between you and the subject. Stands to reason: if you adjust your lens to focus at four feet, the object you want to take a picture of is conceivably four feet away from your camera.

This whole discussions makes little sense to me, anyway, if you have a camera with a viewfinder with parallax compensation and a tripod.

Set your camera to infinity. Take a picture of something four feet away from you, e.g. your friend. Arrange the picture so that the top of the head nearly coincides with the top of of the frame. Without moving either camera or friend, turn the focus ring so that the camera focuses at the distance of four feet. Take the second picture. Now look through the viewfinder, adjust the camera so that the top of the friend's head nearly coincides with the top of the frame. Take the third picture.

Look at the pictures and observe what's in the picture and what not.

You now should have the answer to your question. If you don't see any difference, parallax does not work for you and you can cheerfully ignore the function of your gadget.

[Vlad Soare]

2 minutes ago, pop said:

Set your camera to infinity. Take a picture of something four feet away from you, e.g. your friend. Arrange the picture so that the top of the head nearly coincides with the top of of the frame. Without moving either camera or friend, turn the focus ring so that the camera focuses at the distance of four feet. Take the second picture. Now look through the viewfinder, adjust the camera so that the top of the friend's head nearly coincides with the top of the frame. Take the third picture.

Look at the pictures and observe what's in the picture and what not.

That's a great suggestion. I'll try it. 👍

[Helge]

vor einer Stunde schrieb Vlad Soare:

I see. Or at least I think I do. The framelines try to show what part of the focusing plane will (almost) certainly end up in the picture. The camera assumes that if I focus on something, that means I am interested exclusively in the plane that particular object lies in.
So, I guess my example (focusing on one plane while framing according to another plane) is an improper use case and isn't what Leica had in mind when they designed the system.
Right?

I think, Leica thinks you set the focus to the item you would like to see sharp and complete on your picture.

That items at another distance (and potentially not showing up sharp) might not be exactly framed is a systematic issue all viewfinder camera have.

[Vlad Soare]

I see. If you put it like this, it makes sense. 👍

Thank you all. It's all clear now.

 

 

[Michael Geschlecht]

12 hours ago, Vlad Soare said:

Guys, I know how a rangefinder works. I know the viewfinder is just an approximation of what the lens sees.
What I don't get is why this approximation should depend on how the lens is focused.
I mean, if I put the camera on a tripod and leave it there, completely still, and all I do is rotate the focusing ring, the framelines will move. What's that supposed to tell me? As long as the camera is perfectly still, both the viewfinder coverage and the actual image projected by the lens will remain unchanged, won't they? So, how's the movement of the framelines supposed to help me? What's it supposed to tell me?

In my example, let's say I want to get the framing shown in the second picture, but I want to shoot wide open, with the girl in focus and everything else out of focus. If the camera were focused at infinity and positioned so that the viewfinder looked like that, then everything would be fine (except for the girl, who would be out of focus). Now, I move the camera so that the girl is in the center, then use the rangefinder to focus on her. Then I reframe. But when I reframe, the camera will end up in a different position than it was before, because the framelines I'm relying upon are in a different position in the viewfinder than they were before. How does this help me? 

I'm sure the answer is simple and is right under my nose, but I just can't seem to grasp it. 

 

Hello Vlad,

For further clarification: Please re-read Post #9, just above.

Best Regards,

Michael

[Advertisement]

Hallo,

Look here: Leica M

[Stephen.s1]

..all will be well with the EV (if it's mounted in the center).. and of course there's always the monitor.. Eat your hearts out D lovers.