Scale-focus calibration

[JPD]

Don’t unnecessarily scratch your camera no matter how minuscule the scratch.
Just use tape or use a fine marker drawing arrows. More precise, and completely invisible once you’re done.
Even if they are pretty cheap, these folders are irreplaceable.
With that off my chest, I have a hard time seeing how this would work precisely.
Determining accurate focus at infinity is rather hard. A target 20 meters away is going to be only slightly less sharp than a target at infinity, even wide open. There is much room for interpretation between those two.
Much better to use close focusing, especially the distance the lens was designed to optimally focus at.

The scratch will not be seen when the focus scale is put back on, but a marker is a good idea. When you examine the negatives in an enlarger it's easy to see where the focus lies, so the room for interpretation is tiny. And when in use, the camera is not often used with fully open aperture anyway, and with scale focusing, even when using an accessory rangefinder, having some depth of field is important to "correct" for small focusing errors anyway.

 

[Helge]

I don’t understand why you guys insist on calibrating to infinity.
It’s not a very precise target.
It’s like using smoke as a test target when you have a razor edge.
If you adjust to a precise close range, the rest will follow automatically. And the longer away you get the less accurate focus matters.
DoF grows a lot and you usually stop down when shooting scenes like that to achieve maximum DoF.

 

[henryvk]

Determining accurate focus at infinity is rather hard. A target 20 meters away is going to be only slightly less sharp than a target at infinity, even wide open. There is much room for interpretation between those two.
Much better to use close focusing, especially the distance the lens was designed to optimally focus at.

It would seem to make much more sense for close focus, where you can actually determine what's happening with a grid or something like that for a target. I could see this working with ground glass and a loupe (or test film, if you really need to).

For infinity, Helge's method is better because you can actually determine what's going on with the naked eye. I use an SLR with a waist-level finder and split-prism, but instead of film with a hole in it I tape a long hair across the film gate. That's what I do for infinity focus and it's very quick and there's virtually no guesswork involved because the hair (with the right backlighting) shows up as a bright, very fine line in the SLR viewfinder and it's hard to mistake when it actually merges in the split-prism.

After that, I check focus at 1,5 meters with a piece of paper with a big black X on it. For this, I'm reasonably confident in a ground glass and 10x loupe.

Btw there should be a sticky with this collimation stuff.

 

[Helge]

It would seem to make much more sense for close focus, where you can actually determine what's happening with a grid or something like that for a target. I could see this working with ground glass and a loupe (or test film, if you really need to).

For infinity, Helge's method is better because you can actually determine what's going on with the naked eye. I use an SLR with a waist-level finder and split-prism, but instead of film with a hole in it I tape a long hair across the film gate. That's what I do for infinity focus and it's very quick and there's virtually no guesswork involved because the hair (with the right backlighting) shows up as a bright, very fine line in the SLR viewfinder and it's hard to mistake when it actually merges in the split-prism.

After that, I check focus at 1,5 meters with a piece of paper with a big black X on it. For this, I'm reasonably confident in a ground glass and 10x loupe.

Btw there should be a sticky with this collimation stuff.

It’s good for close range too. And since close range is better overall it’s Best™. ;-)

Problem is no one really knows where the film naturally falls in a folder. Especially in a 6x9.

Is it on the rollers, the rails, the pressure plate, somewhere in between?
It’s been discussed ad infinitum.
Pressure plates on folders is also something else than pressure plates on say a Rolleiflex.
Films natural curvature is used as a counter spring to keep the film flat to a higher degree.

To be complete sure, make several test holes in the sacrificial film. Average and take out outlier results.
A fraction of a millimeter means a lot.

And just use a digital camera. Almost every model from the last ten years has zoom in to pixel level.
Even use an external monitor to get a better look. Same setup you use for camera scanning.

 

[snusmumriken]

Make a clean hole in a roll of sacrificial film through backing paper and film, smaller than the opening in the red window and right on a printed number.
Make the hole a bit bigger in the paper, so you get a clean sharp edge of film to look through.
The precise shape doesn’t matter too much. A line is fine, a round sharply punched hole too.

Load the film and wind to the hole. Leave the red window open and backlight it with a powerful diffuse light.
Now you have an exact film plane gauge to use the SLR focus check method in.

Ground glass focusing will never ever be accurate on a folder. A. The film plane is not where the ground glass rests. B. The ground glass is too coarse and the contrast and image too small to accurately focus on. Also fiddly as hell.

Optimum focus distance is far more important than infinity focus with a front cell focuser. Unless you plan on only using the camera for horizon shots. Even at fifteen meters accurate focus is important with 6x9.
The optimum focus distance is often found by dividing the focal length by 20.
So a 105mm lens like the Vaskar should be optimized for about 5.25 meters (5 is fine too as long as you’re consistent).

I’m taking careful note of this. Just one question: if instead of a folder I was dealing with a 35mm camera with a 40mm front-cell lens, your rule of thumb suggests the optimal distance would be 2m, which would be close to the minimal focussing distance. Is that really correct?

 

[Helge]

I’m taking careful note of this. Just one question: if instead of a folder I was dealing with a 35mm camera with a 40mm front-cell lens, your rule of thumb suggests the optimal distance would be 2m, which would be close to the minimal focussing distance. Is that really correct?

To tell you the truth, it really don’t know if it’s about an optical fundamental or if it’s just used as a convenient number, that was used as a shorthand.

It’s a number I’ve seen mentioned many times on forums where old repair people and experts chipped in, and in books on vintage cameras.

However, 2 meters doesn’t seem too close.
It’s where many subjects will naturally be placed.
After all we are talking about a very very gradual and slight transition from optimal to a little worse, but correctable with stopping down.

It’s less gradual the closer you get to the camera from optimal, until you get within about 1 meter for medium format and 80 cm for 135.
That’s why diopters was popular instead of just cranking the lens all the way out.
But still you’ll have to look hard to see a difference from a unit focuser.

From optimal focus to infinity, the transition towards worse is even slower and hard to spot.

In fact it mirrors DoF transition in general for the lens.

 

[tom williams]

I've used this method a few times. It should work well if the DSLR/mirrorless lens is longer then the testing lens and you are certain the infinity mark on the DLSR lens is correct. It's akin to using an autocollimator. I usually use a clear piece of glass on the film rails with a sharpie mark on it, pressed on with a rubber band. Essentially what happens is that when the testing lens is at infinity, it "collimates" the image to infinity (i.e. parallel rays). The image is then properly refocused to the DLSR sensor as long as its lens is also at infinity.

OAPOLi, thanks for the explanation.

The image usually falls out of focus rapidly. How does the range you observe compare to the marks on the focus scale (from infinity to the next distance)?

The image does fall out of focus rapidly, from the point of least extension of the lens where the lines are in focus, toward lesser lens extension (smaller distances on the scale). The image will stay in-focus for a surprisingly large adjustment if I continue extension beyond the point where the lines first come into focus.

 

[tom williams]

blee1996, thanks for the detailed reply.

Just adding some details for your next experiment.

1. I used a mirrorless camera Nikon Z6, which can magnify the image which is handy. I read your D750 can do the same in live view. Then we are all good there.
2. I need a telephoto SLR lens that is wide open for critical focus. I used the Nikkor AiS 105mm f/2.8 macro manual focus lens. It has an added benefit of a focus lock knob, so infinity will stay infinity on the lens when I lock it down. And I keep the lens wide open. This way, the depth of field is kept the shallowest to assist in accurate calibration. I could try 200mm or 300mm lens in the future.

I used a Tokina Macro 100mm f2.8 D lens, wide open, on manual focus, and not checked for infinity mark accuracy.

1. Make sure the digital camera and SLR lens are actually focus on infinity, and do not always trust the infinity marking on the SLR lens. Once the SLR lens is set on infinity, I point my digital camera to as far an object as I can find, e.g. a TV antenna on a remote mountain top. I use the magnifying function on the digital camera to double-check.
2. In terms of what to put on the film plane of the folder, I think ground glass, blank film, tracing paper are all fine as long as you can keep them exactly where the film will be and keep it as flat and tight as possible. Draw a cross hair on a matte surface, so you can focus both on the line as well as on the grain.
3. On the folder: there should only be a very narrow position of the outer lens housing in relation to the fixed part of the lens, where the infinity stand. If you have take care of all the above steps and there is still a "wide range" of infinity position, can you check if the lens assembly is put back together properly after your cleaning?

The calibration procedure became necessary after cleaning the lenses. I think it would be hard to screw up the reassembly, but I'm qualified. I've just shot a few images with the camera in question, of a resolution target at various distances corresponding to scale positions, so there will be convincing information shortly.

 

[Dwight Anderson]

For those using scotch tape across the rails, I've found that marking with a pen isn't good enough. The ink leaves a rough edge in the grain of the tape that makes it harder to collimate using an SLR. I started using an exacto knife to to lightly score a line in the tape (not cut it). This is much easier to focus on than an ink line.

Typically though, I use tape or ground glass and a 50mm lens as a loupe. I focus for infinity on a lighted tower about two miles away first. Then I set up in my kitchen and focus on the digital clock on my stove. The blue led numbers are very easy to see using the 50mm. I use the closest marked distances on the lens, usually 3, 4 and 5 feet, measuring from the film plane to the clock. In most cases, I don't have to adjust from the previous setting at infinity, but I like to have the close distances correct.

After all, scale focusing in practice is only as good as your distance estimating skills and I'd rather know the close focal points are correct.

 

[Helge]

After all, scale focusing in practice is only as good as your distance estimating skills and I'd rather know the close focal points are correct.

You just measure the distance outright, with pre-measured string or a metal tape, use an external tele rangefinder (easy to calibrate and more precise than a viewfinder RF) or you use a laser measuring thingy.

 

[Bill Burk]

When I recalibrate the scale focus on folders, I first use a groundglass on the film plane and focus on something at infinity. Then I print a numbered scale with lines about 1,5 mm apart, that I tape on the lens mount, and make a scratch on the front lens mount, like on this picture, so it points to a line in the middle of the printed scale. Then I load the camera with a test roll of film and take pictures, with wide open aperture, of something at "infinity", turning the front element from 1 to 8 (or 12), develop the film and check which one nailed the focus, let's say it's #5. Then I turn the front element so the scratched mark points to the corresponding number on the taped scale. Remove the scale, and put back the focus ring and tighten the lock screws.

https://live.staticflickr.com/65535/51184059283_611c2f99ef_w.jpg

This has worked very well for me.

Very clever!

 

[henryvk]

You just measure the distance outright, with pre-measured string or a metal tape, use an external tele rangefinder (easy to calibrate and more precise than a viewfinder RF) or you use a laser measuring thingy.

Or try the average lenght of two paces, which should be around 1.5 meters

 

[JPD]

I don’t understand why you guys insist on calibrating to infinity.
It’s not a very precise target.
It’s like using smoke as a test target when you have a razor edge.

I don't think, for example, street lights 100-200 meters away are that hard to focus on and comparable to smoke. When they are as "point-like" as possible, they are in focus. Precise enough for a 6x6 or 6x9 folder with a standard lens.

I want the focus to be spot on at the infinity stop when I unfold the camera, and it's not certain that the focus scale is all that precise at closer distances. The focal length can vary a little between lens samples. Rollei TLR viewing and taking lenses were hand picked to match, but I doubt the lenses and focus scales on folders were matched.

 

[Helge]

I don't think, for example, street lights 100-200 meters away are that hard to focus on and comparable to smoke. When they are as "point-like" as possible, they are in focus. Precise enough for a 6x6 or 6x9 folder with a standard lens.

I want the focus to be spot on at the infinity stop when I unfold the camera, and it's not certain that the focus scale is all that precise at closer distances. The focal length can vary a little between lens samples. Rollei TLR viewing and taking lenses were hand picked to match, but I doubt the lenses and focus scales on folders were matched.

But “It’s not a very precise target”.
What might be fine or even appear spot on at anything beyond say a hundred meters distance, cascades down, and multiplies into a grossly missed focus at closer ranges.
If you use a target at a 100 meters and it’s seems fine and you look at another target a kilometer away and it also seems fine and you used either target to calibrate, then that translates into very off close ranges.
The demands on precision is simply not as taxing at longer distances.
It’s like using a rubber band to measure out a kilometer and then subdividing that distance to make a meter stick.

So unless you only shoot stuff at infinity, which seems unlikely unless you live in the Swiss Alps or the Rockies, calibrating to infinity is a very bad idea, and I can guarantee you not what they did at the factory originally.

You might get lucky once in a while, and get something close to good enough at close range. But don’t kid yourself. It could be better.

Also, a further complication if you use a real target (not the SLR method) aerial perspective and other psycho optical phenomena might affect your perception of good enough.

If you want perfect infinity with a front cell focuser use a tripod and stop down.

 

[takilmaboxer]

I'm sorry you're getting so much bad advice. Post #20 nailed it, that's the way to go. And remember, your final calibration must be done by exposing actual film.

 

[Helge]

I'm sorry you're getting so much bad advice. Post #20 nailed it, that's the way to go. And remember, your final calibration must be done by exposing actual film.

Unless you can tell mere exactly where and how I’m wrong, you are only contributing bad advice.