A question about parallax

[SimonK]

I'm relatively new to the world of rangefinders and discovering the bane of this world - parallax - and the ways to combat the problem - distracting lines in the viewfinder, messing about with separate viewfinders...

Now, my question is - why not just put the viewfinder directly above the lens and the rangefinger lens out to the side?

It can't be a question of ergonomics; every SLR has the viewfinder in this position and is easy enough to put against your face.
It can't be a question of mechanics because nothing could be simpler than running a rod from the cam on the lens out to the edge of the camera.
I measured the distance from above the centre of the lens of a FED to where I would have put the rangefinder and it is 30 mm - twice the distance between vf and rf lenses on an XA, so the problem isn't one of separation distance.

There must be a good explanation for why rangefinders are so designed, but I can't for the life of me see what it is. Could some wise soul please enlighten me.

 

[Spyderman]

Take a look at Leica and Zeiss Ikon and Contax II and...

so the problem isn't one of separation distance

Every serious rangefinder camera has longer rangefinder baselength than XA.

Even if you could put the VF above the lens - would it remove prallax error ? no. You should first understand what "parallax error" is.

 

[oftheherd]

You have to understand that SLRs are a photographic aberration. If you read the press from the time they first started being foisted on the public, you will note there was a lot of well deserved bad comments. Terrible invention. 😀 😀 😀

Actually, no matter where you put the viewfinder, at some point you willl get close enough that parallax will be a problem. I think some cameras have trided to use a separate viewfinder, but even so, didn't get it right over the lens. You help the parallax problem but then run into the problem of seeing around lens hoods. Making lens hoods that allow you to see through them is possible. It is done, but it adds to the cost.

Everything is a tradeoff.

 

[ClaremontPhoto]

Parallax is a non-issue for normal working.

It's only relevant to close-up work.

For which a rangefinder camera is poorly suited anyway.

 

[SimonK]

Parallax is a non-issue for normal working.

It's only relevant to close-up work.

For which a rangefinder camera is poorly suited anyway.

Well, I don't know about that. I took a photo of a doorway from about 3 yards, nicely centred in the viewfinder (I'm an SLR user of 30 years) and of course on the print it is way off to the side.

 

[SimonK]

Take a look at Leica and Zeiss Ikon and Contax II and...



Every serious rangefinder camera has longer rangefinder baselength than XA.

Even if you could put the VF above the lens - would it remove prallax error ? no. You should first understand what "parallax error" is.

I know exactly what 'parallax error' is thanks, my point was that for most practical purposes a slight error caused by the viewfinder being 1 or 2 centimeters above the lens would cause must less problem than being 4 or 5 centimeters to the side.

 

[40oz]

are you asking why the finder isn't directly over the lens, or venting about parallax issues at minimum focus distance? Venting is allowed, BTW 🙂

Some rf cameras have framelines that adjust for parallax as you turn the focus ring. Some don't.

My Canonet has framelines that move with the focus ring. The finder is also too large to fit over the lens without making the body taller there. It's really compact given the f/1.7 lens, although the rf base is really short.

I don't know if Oscar Barnack is a member of this forum, but maybe someone has his email so we can all know the definitive answer 🙂 FWIW,

 

[ferider]

Well, I don't know about that. I took a photo of a doorway from about 3 yards, nicely centred in the viewfinder (I'm an SLR user of 30 years) and of course on the print it is way off to the side.

What you are observing is more than just the parallax problem. When you get close, the effective FOV decreases. A perfect RF would not only correct by shifting the framelines, but also by shrinking them as you move closer.

I believe the only cameras that do that have a motorized viewfinder (like the Contax G). The Hexar AF only shifts two of the four 35mm framelines, like this:

In Leicas, Bessas, ZI, Hexar RF, etc, the framline remain the same size. Which leads to a shift of the object along one diagonal. The same is true for the constant-size and static viewfinder of the more primitive Barnack (or copy thereof).

So, under the bottom-line, moving the viewfinder as close to the lens as possible only solves one of the two fundamental problems. Both create in-accuracies of similar magnitude.

Roland.

 

[mh2000]

I think it is thinking like this that let camera designers get away with poor placement of VF's. But I disagree, RF's should be expected to be shot in the 3-6ft range... which seems like a "normal" working distance to me... but maybe that is why I have shot SLRs more than RF's...

Parallax is a non-issue for normal working.

It's only relevant to close-up work.

For which a rangefinder camera is poorly suited anyway.

 

[ferider]

If I remember right, Barnack viewfinders work for 10m onwards.

M's start at 1 or .7m, depending on the model.

If you are further away you have to crop .... Note that also with many SLRs cropping is necessary to get exactly what you see.

Not with my OMs or Nikon F though 🙂

Roland.

 

[Benjamin]

So, does one really need to worry about parallax error, with regards to framing at least, with a lens of a minimum focus distance of say .7m anyway?

Does Bruce Gilden have to worry about parallax error at a close distance?? And let's say perhaps that he used a 50mm instead of the wides that he obviously tends to favour? Could you end up photographing only half a face for example?

 

[ferider]

So, does one really need to worry about parallax error, with regards to framing at least, with a lens of a minimum focus distance of say .7m anyway?

Does Bruce Gilden have to worry about parallax error at a close distance?? And let's say perhaps that he used a 50mm instead of the wides that he obviously tends to favour? Could you end up photographing only half a face for example?

Not half, but 80 percent or so, with a Barnack yes, with an M6 no. The issue gets worse when the lenses get longer.

Roland.

 

[Benjamin]

Oh okay, thanks.

 

[ruben]

What you are observing is more than just the parallax problem. When you get close, the effective FOV decreases. A perfect RF would not only correct by shifting the framelines, but also by shrinking them as you move closer.

I believe the only cameras that do that have a motorized viewfinder (like the Contax G). The Hexar AF only shifts two of the four 35mm framelines, like this:

In Leicas, Bessas, ZI, Hexar RF, etc, the framline remain the same size. Which leads to a shift of the object along one diagonal. The same is true for the constant-size and static viewfinder of the more primitive Barnack (or copy thereof).

So, under the bottom-line, moving the viewfinder as close to the lens as possible only solves one of the two fundamental problems. Both create in-accuracies of similar magnitude.

Roland.

Actually there is a cheapo of high quality glass that not only moves the bright lines but also shrinks them according to distance, and therefore merites your world championship cup as "Perfect Rangefinder": the Konica Auto S, and Konica Auto S2.

Teasing,
Ruben

 

[ferider]

I need one of those !!! ............. teasing, too. 🙂

It gets more complicated with interchangable lenses, Ruben.

All the best,

Roland.

 

[JoeV]

I think the main design problem to the OP's suggestion is that the coupling betwen the lens and the rangefinder mechanism would therefore be extemely lengthy and perhaps would change the entire shape of the camera. You'd need a linkage connecting from the lens mount area all the way over to the left side where the rangefinder window would need to be located.

~Joe

 

[Al Kaplan]

The medium format 6x7cm Koni-Omega also had expanding/contacting framelines just like the Konica Auto S an S2, along with interchangeable lenses of 58, 90, 135, and 180mm.

Getting back to 35mm cameras, mounting a finder directly over the lens would eliminate side parallax, and then only in horizontal pictures, but shoot a vertical and you'd have lots of side parallax. There ain't no free ride!

 

[Chris101]

Hey Reuben, what happened to the wizened old man? Some kid has taken over your avatar!

 

[payasam]

Finder above lens can lead to a lot more than "slight error". Besides, we must remember that even when there is compensation for parallax, the different absolute positions of lens and finder can cause trouble: for example, distant objects behind near ones.

 

[russianRF]

Isn't it generally true that the farther apart the viewfinder and the rangefinder windows are, the more accurate the focusing?