Scale focusing

[Chan Tran]

I have some cameras that the rangefinder patch faded so badly that I can't use it any more. I can estimate the distance quite accurately but the problem is that I can't set the distance on the lens very well. The scale in non linear and has too few markings. I want to add more markings on the lens distance scale. I think one can mathematically calculate the marking position but I haven't figure out how yet. Anyone already figured it out?

 

[xkaes]

What camera/lens is it? It would be important to know the closest focusing distance. This should be easy to find on-line if it is a common lens.

 

[OAPOli]

The formula is 1/f = 1/u + 1/v where f is the focal length, u the distance between the lens and image and v the distance between the lens and subject. u + v would be the distance between subject and image, which is what is marked on the lens. However, in order to make the scale, you would also need to know the diameter and thread pitch of the helicoid. The latter would require a disassembly, but maybe you can transfer the current scale onto paper to infer the pitch.

 

[Donald Qualls]

The spacing between the number is logarithmic -- that means 1 meter and 2 meter are further apart on the scale than 10 meters and 20 meters (on many 35 mm cameras, the next thing beyond 10 m is infinity). Look at a slide rule (you can find pictures and even a simulator online, if you're never handled one) to see logarithmic scales. If you can measure the lens extension with focus ring movement accurately, you can use the focus formula to fill in those gaps, but that only works reliably on lenses with the node at or close to the aperture stop.

 

[Dustin McAmera]

You can do it graphically; if you measure the distance along the lens scale, around the barrel, (actual millimetres along the scale) from the infinity mark to each of the marked distances, you can plot a graph (on squared paper or in Excel) of the marked distance on the scale in feet or metres (say d) against the physical distance along the scale in millimetres (say x). It should be a reciprocal curve (d proportional to 1/x); and you could work out its exact formula; but you might as well just draw the graph, and read off from it the scale distance for the extra marks you want to make. I think it'll be accurate and precise enough for distances you're going to estimate.

 

[Dustin McAmera]

That's the graph for my Yamato Pax M2, scaled in feet. For me, it already has enough marks (also, the focus is stiff, and the shutter doesn't work anyway)

In fact, you see I plotted that with d on the y-scale. If I'd plotted it with that on the horizontal scale, and the scale distance as the vertical, the fitted regression function would have given me the scale distance with absurd precision. The power should be exactly minus one really; there's probably some way I could force that in Excel, but I think the extra accuracy would be lost in the thickness of the lines marked.

In fact if I plot the graph as scale distance on the y scale, and 1/d on the x scale, it's a nice linear graph, and I can force the intercept to be zero, giving me scale distance y (mm) = 122.53/d with d in feet.

 

[Sirius Glass]

Using the depth of field scale can help if 1) the depth of field scale is on the lens and 2) one is willing to use smaller apertures to increase the depth of field.

 

[BAC1967]

Depending on the camera you can use a piece of ground glass to focus and mark the distance.

 

[BrianShaw]

Anyone already figured it out?

f/8 and estimate using the existing distance marks

 

[Dustin McAmera]

I have bookmarked a company named Vacuum Coatings who say they will recoat mirrors and prisms (I have no connection to them, and haven't sent them any work myself - I have a couple of cameras that could do with it). Regarding cameras, they say they only handle the mirror or prism element; you are expected to get it out, and put it back, and not send them a whole camera.

Vacuum Coatings Scientific Mirrors - Sextant, Camera mirrors and Prisms

www.scientificmirrors.co.uk

 

[neilt3]

Depending on the camera you can use a piece of ground glass to focus and mark the distance.

That would be my approach to put clear marking on the lens barrel and body .
Camera on a tripod set to a measured distance to be focussed on via the ground glass .
Mark it .
Nice onto next distance target .
Mark it .

No maths or getting the start point marked correctly and hoping all other marks are in the exact right spot , based on calculations.

Just make sure to use a good loupe for accurate focussing.

 

[Ian C]

If will be easier to answer if you specify the following:

1. What is the focal length of the lens (as accurately as possible).

2. What is the make and model of lens and of the camera. Is the scale on a slide? Or is it marked on the barrel of a helicoid-focusing lens? A photo would be useful as well.

 

[xkaes]

I asked the same question in Post #2 yesterday. Will we get a response?

 

[Ian C]

Assuming that you can accurately measure or estimate the subject distance s (subject plane to first nodal point of the lens), let

f = focal length of lens

s = subject distance

Δ = the distance the lens is displaced forward of its infinity position.

Δ = (f^2)/(s – f)

At infinity, Δ = 0


Examples using a 100 mm lens:

200 mm subject distance (1:1 magnification), Δ = 100 mm

500 mm subject distance, Δ = 25mm

1000 mm subject distance, Δ = 11.11 mm

10 m subject distance, Δ = 1.01 mm

20 m subject distance, Δ = 0.50 mm

50 m subject distance, Δ = 0.20 mm

100 m subject distance, Δ = 0.10 mm


If you’re using a helicoid-focusing lens, there is an extra step involved to calculate the required angle of rotation of the focusing ring. It’s reasonably simple.

 

[Chan Tran]

I wouldn't know where the front or rear nodal point of the lens and thus I wouldn't be able to measure the focal length either. The distance I know is from the subject to the film plane.

 

[BAC1967]

There is a seller on eBay that sells ground glass made to fit 35mm and 120 cameras. That's what I use to calibrate my rangefinders.

 

[bernard_L]

However, in order to make the scale, you would also need to know the diameter and thread pitch of the helicoid. The latter would require a disassembly, but maybe you can transfer the current scale onto paper to infer the pitch.

No need to involve the pitch; see below. Ditto for focal length.

The spacing between the number is logarithmic -- that means 1 meter and 2 meter are further apart on the scale than 10 meters and 20 meters

Beg to disagree, but in order to not leave false information on record:
a) The scale is not logarithmic.
b) If the scale were logarithmic, 10m-20m would be just as far apart as 1m-2m;, they are much closer.
c) The scale is such that (to a good enough approximation) the inverses of the marked distances are in a linear progression; i.e. for example, 1m (1), 1.5m (2/3), 3m (1/3) and infinity (0) are equally spaced.

You can do it graphically; if you measure the distance along the lens scale, around the barrel, (actual millimetres along the scale) from the infinity mark to each of the marked distances, you can plot a graph (on squared paper or in Excel) of the marked distance on the scale in feet or metres (say d) against the physical distance along the scale in millimetres (say x). It should be a reciprocal curve (d proportional to 1/x); and you could work out its exact formula; but you might as well just draw the graph, and read off from it the scale distance for the extra marks you want to make. I think it'll be accurate and precise enough for distances you're going to estimate.

+1
Plus, doing it graphically, you bypass/absorb/ignore the knowledge of the helicoid pitch, the nodal point, etc...

Assuming that you can accurately measure or estimate the subject distance s (subject plane to first nodal point of the lens), let

f = focal length of lens
s = subject distance
Δ = the distance the lens is displaced forward of its infinity position.
Δ = (f^2)/(s – f)
At infinity, Δ = 0

These are based on Newton's equations, with distances measured in object and image spaces from heir respective foci. A typical MF folder with front element focusing has the distance measured from the lens; a typical "modern" 35mm camera has the distance measured from the film plane, with significant differences between the two cases at short distances; for instance, for 1:1 reproduction, the subject distance is 2f from the lens, and 4f from the film plane. And then there are real-world thick optical systems.

Save yourself headaches and use the graphical method.

 

[Chan Tran]

Now I have a piece of 8x10 ground glass, how do I cut it so I can use it on the 35mm?

 

[Donald Qualls]

Beg to disagree, but in order to not leave false information on record:
a) The scale is not logarithmic.
b) If the scale were logarithmic, 10m-20m would be just as far apart as 1m-2m;, they are much closer.
c) The scale is such that (to a good enough approximation) the inverses of the marked distances are in a linear progression; i.e. for example, 1m (1), 1.5m (2/3), 3m (1/3) and infinity (0) are equally spaced.

Thanks for the correction.

Now I have a piece of 8x10 ground glass, how do I cut it so I can use it on the 35mm?

I'd suggest using a piece of stiff plastic (I used to use a piece cut from a CD case, but those are almost collector's items now) with frosted tape on one side; put it frosted side toward the lens, where the film would go.

 

[ic-racer]

An easy way to make a focus scale is to use an Autocollimator. But, as others have pointed out, there are many other methods.

Anyway, for the sake of completeness, this is the math to construct a focus scale with an Autocollimator. In real life, my Autocollimator came with a chart with the values already calculated. Some other Autocollimators (like the c6800) actually have the test lens focus scale numbers engraved on the Autocollimators focus scale for easy reference.

Yes, the test lens focus scale numbers are actually engraved on the c6800 collimator because, for example, with a 6.5mm displacement of the collimator focus (the 20ft engraving), ALL test lenses will be focused at 20 feet. All lenses, all focal lengths etc. etc.

 

[reddesert]

The equation for a thin lens is that which Ian C posted: extension past infinity = f^2 / (d_subject - f).

The cameras in question are rangefinders. They probably were never set up to focus closer than d_subject = 10 to 20 times the focal length anyway. That means you can approximate:
d_subject - f ~= d_subject
extension ~= f^2 / d_subject

So you can simply use that helicoid travel goes as 1/subject distance, as bernard_L said.

The questions of nodal points of the lens, whether you're measuring subject distance from lens or from film, etc, these are all theoretical when you're trying to draw a scale focus on a rangefinder camera that never focused closer than about 1 meter to begin with. In the end, you're probably going to guess the focus distance, which will be as much a source of inaccuracy as the scale labels. Don't overthink it.

 

[MFstooges]

I read somewhere about sharpie fix for the faded RF patch since my thread about fix lens RF. I'm not sure if it works.

 

[Donald Qualls]

I read somewhere about sharpie fix for the faded RF patch since my thread about fix lens RF. I'm not sure if it works.

I works very well on my Petri 7s. My other fixed-lens RF cameras have good-enough patches, and my Kiev 2 is very good.

 

[polka]

With my Minox35EL, scale focusing is very easy !

As the focal length (35mm) is approximately the same as the width of the 24x36 frame (36mm), you just have to estimate the edge to edge size of some element in your view finder at the desired focus plane (anything like for instance a sidewalk or a pole or a fence or a tree etc.) and set it as the focusing distance.

This will work directly with any 35mm focal. For other focals use some approximate proportion reductions (like 1.5 for 50mm focal, 3 for 100mm focal, etc.).

No use for a rangefinder !

POLKa