Aligning BOSSCREEN to film plane [Archive] - Large Format Photography Forum

towolf

26-Oct-2014, 17:28

I’m looking to reconcile two contradictory findings in my pseudo-scientific endeavor to nail down focus on the BOSSCREEN ground glass that my camera came with.

I’m presenting my method and if someone can spot my mistake I would be very grateful.

I was shooting slides (E100G) for the first time and was just a little disappointed with the detail on the horizon of a landscape scene. The shot was made racked against the infinity stop of a Linhof Tech V. Now, I have checked the infinity stop against the GG painstakingly in my endeavor of projecting a laser through the rangefinder. My new paranoid hunch was that maybe the GG was not co-registered with the film plane. Now, the previous owner cannot be asked how he installed it, because he fell victim to cancer (which is why I have the camera).

Since I had a big load of old Xray film I commenced to photograph a table top with a newspaper on it at 45°

http://i.imgur.com/en57mTQl.jpg (http://i.imgur.com/en57mTQ.jpg)

I focussed on a particular line (thin line under "Asset Management" at 10cm on the paper ruler) but the focus on the developed sheet seemed to be 1 cm out.

Now, 3 columns of text were (roughly) 18cm wide on the paper and 7cm on the film giving a magnification of 1/2.6. The other calc 150/(545-150) found the same ratio of 2.6.

I calculated with 10mm offset "(10mm * (sqrt(2)/2)) / (2.6**2) = 1mm", and hence got 1mm film deviance from the ground glass.

http://i.imgur.com/1dVB3bw.png (http://i.imgur.com/1dVB3bw.png)

Also in the repeat.

http://i.imgur.com/QwVGFoS.png (http://i.imgur.com/QwVGFoS.png)

Then I lifted the BOSSCREEN from its frame and found no metal shims (which are supposed to be there on a Technika) but two 0.9mm thick rubber-like stacks of electrical tape instead of shims.

http://i.imgur.com/MaYHODWl.jpg (http://i.imgur.com/MaYHODW.jpg)

towolf

26-Oct-2014, 17:32

Since the 0.9 mm were agreeing with my estimate I peeled them off and fixed the screen directly onto the metal flange with the four set screws fully retracted. I took another test shot and focus seemed to match now.

http://i.imgur.com/zkPRva7l.jpg (http://i.imgur.com/zkPRva7.jpg)

However, and now comes the problem, today I took more test shots comparing the old infinity stops to my new ground glass setting but things at infinity didn’t get better, they got worse. For some reason what matched at close-up distance didn’t match at infinity.

So, today I wanted to try a different route to measure the distance of the film plane and Bosscreen wax layer each. Just breaking out the calipers would not have given the right accuracy obviously. And I have no advanced tools like micrometers. And in the BOSSCREEN the image layer is in the center of a glass sandwich preventing purely mechanical measurements.

But, I have also inherited an old 40x Zeiss stereo microscope:

http://i.imgur.com/rUhP33Ml.jpg (http://i.imgur.com/rUhP33M.jpg)

I dropped a Durst enlarger glass on top of the GG frame to measure the focus difference between a felt pen mark on the underside of the glass pane

http://i.imgur.com/YyO4W6sl.jpg (http://i.imgur.com/YyO4W6s.jpg)

and the wax layer inside the screen.

http://i.imgur.com/wWJAEDil.jpg (http://i.imgur.com/wWJAEDi.jpg)

towolf

26-Oct-2014, 17:33

Then I could use my callipers to measure from these two defined spots

http://i.imgur.com/gdZh4nMl.jpg (http://i.imgur.com/gdZh4nM.jpg)
http://i.imgur.com/fzUVDHHl.jpg (http://i.imgur.com/fzUVDHH.jpg)

to make a measurement with ±0.05mm accuracy (by reading the nonius at right angles)

http://i.imgur.com/SODUBmal.jpg (http://i.imgur.com/SODUBma.jpg)
http://i.imgur.com/4G6esHzl.jpg (http://i.imgur.com/4G6esHz.jpg)

towolf

26-Oct-2014, 17:34

The vertical focus offset between the lower and the upper level would give me the flange distance from the GG

http://i.imgur.com/UxctduIl.jpg (http://i.imgur.com/UxctduI.jpg)
http://i.imgur.com/wQY3FLIl.jpg (http://i.imgur.com/wQY3FLI.jpg)

And the same story with the film holder

http://i.imgur.com/5qF5ddTl.jpg (http://i.imgur.com/5qF5ddT.jpg)
http://i.imgur.com/3UaNV3gl.jpg (http://i.imgur.com/3UaNV3g.jpg)

I measured everything 3 times and calculated both flange distances and got 4.8mm for the film holder (the specification standard seems to be 3/16 inch, so 4.76mm) and got 3.9mm for the new "naked" position of the BOSSCREEN wax layer. A difference of 0.9mm!!!

So it appears, after all, that the original rubber shims were correct when measuring (opto-)mechanically. But on my table top newspaper test there was a clear problem.

If anyone of the wonderfully knowledgeable people here can spot my error please help me find it.

Maris Rusis

26-Oct-2014, 22:39

towolf

27-Oct-2014, 00:11

It could be an optical path difference problem caused by the glass in front of the wax layer. Light passing through glass appears to travel a greater distance than if there was only air there. Optical pathlength is geometric pathlength multiplied by the refractive index of the medium the light is passing through.

That is an interesting point. However, didn’t I already compensate for this by measuring through the glass in question? Or does the refraction by the inner pane of glass depend on the incident angles? Does this explain the difference between 0.75 m and infinity?

That would be a flaw in the Bosscreen itself, no less.

Bob Salomon

27-Oct-2014, 03:18

First, contact a Linhof service center and get a set of the correct shims. They are the same as the current ones for the MT.
Second, Do you have the Boss Screen for the Linhof?

If not, find one, if you can or just get a new enhanced screen from anyone else.

towolf

27-Oct-2014, 04:13

First, contact a Linhof service center and get a set of the correct shims.

I can ask, but the last time I wrote email to linhof.de they quoted 25 EUR for an infinity stop. I’d rather figure this out myself. That is half of the fun in LF for me.

Second, Do you have the Boss Screen for the Linhof?

No, it says BOSSCREEN HOLLAND - PLAUBEL. Are there differences in the stand-off?

In the end I don’t want to buy anything or pay someone. I just want to know the best method to co-register the screen I have.

towolf

27-Oct-2014, 04:36

Something that just occured to me: Would the colour of light (let’s say 6500K vs 3200K) affect the results of my newspaper tabletop test? Would it contribute that much? It’s noteworthy that I was using blue-sensitive Xray film in both cases.

I got this idea because some lenses have a red dot to focus on (behind ∞) when using IR film.

Bob Salomon

27-Oct-2014, 06:54

I can ask, but the last time I wrote email to linhof.de they quoted 25 EUR for an infinity stop. I’d rather figure this out myself. That is half of the fun in LF for me.

Yes. Since the image forming layer is the wax between the two sheets of glass there are major difference. So your best course is to get the camera back set back to where it belongs, on the shims with the adjusting screws properly adjusted and install a proper screen.

No, it says BOSSCREEN HOLLAND - PLAUBEL. Are there differences in the stand-off?

In the end I don’t want to buy anything or pay someone. I just want to know the best method to co-register the screen I have.

Cor

27-Oct-2014, 07:32

towolf,

You used Xray film for your test, right ? Is it a flimsy/thin film ? I ask because you were pointing the camera downwards when shooting. So perhaps the film sagged a bit (bowed out) during exposure, and thus ruining your measurements..

best,

Cor

towolf

27-Oct-2014, 07:45

You used Xray film for your test, right ? Is it a flimsy/thin film ? I ask because you were pointing the camera downwards when shooting. So perhaps the film sagged a bit (bowed out) during exposure, and thus ruining your measurements.

This is also very good point. I was fretting about it already. Film flatness. The film is rigid and 0.2mm thick (on the thick end of the sheet film spectrum, and with double emulsion). But I have some doubts about it. I will repeat the film tests with carefully checked flatness. Maybe with a small patch of double-sided tape in the center.

So the general suspicion I have is that the 45° newspaper test was probably confounded and the measurement with the microscope was likely (more) accurate.

Bob Salomon

27-Oct-2014, 08:13

I can ask, but the last time I wrote email to linhof.de they quoted 25 EUR for an infinity stop. I’d rather figure this out myself. That is half of the fun in LF for me.

No, it says BOSSCREEN HOLLAND - PLAUBEL. Are there differences in the stand-off?

In the end I don’t want to buy anything or pay someone. I just want to know the best method to co-register the screen I have.

Yes, that is why yours says Plaubel. If it was the one that they made for Linhof it would say Linhof.

And, again, the best way is to put the camera back together properly. Then adjust the gg to the correct Linhof position.

As for the 25 EUR. Was that for one or for the pair with the set screws?

Maris Rusis

27-Oct-2014, 16:03

That is an interesting point. However, didn’t I already compensate for this by measuring through the glass in question? Or does the refraction by the inner pane of glass depend on the incident angles? Does this explain the difference between 0.75 m and infinity?

That would be a flaw in the Bosscreen itself, no less.

You can't measure where the wax layer in a Bosscreen is by focussing on it through the cover glass. All you get is where it appears to be because of the optical path length effect of the cover glass.

This is an easy test to do. Set up a view camera with the bellows off and exactly focus a subject. Now place a sheet of glass between lens and ground glass and you will notice the subject is now out of focus. To regain focus the length of the camera must be extended.

The wax layer of a Bosscreen cannot be placed by geometric measurement so as to coincide with the film plane in a film holder of correct T distance. It has to be placed optically so that when an image is sharp on the film it is also sharp on the wax. Because the exact position of the screen is influenced by the thickness and refractive index of the glass between the wax and the lens correct shim thickness is found by trial and error.

Bob Salomon

27-Oct-2014, 16:24

You can't measure where the wax layer in a Bosscreen is by focussing on it through the cover glass. All you get is where it appears to be because of the optical path length effect of the cover glass.

This is an easy test to do. Set up a view camera with the bellows off and exactly focus a subject. Now place a sheet of glass between lens and ground glass and you will notice the subject is now out of focus. To regain focus the length of the camera must be extended.

The wax layer of a Bosscreen cannot be placed by geometric measurement so as to coincide with the film plane in a film holder of correct T distance. It has to be placed optically so that when an image is sharp on the film it is also sharp on the wax. Because the exact position of the screen is influenced by the thickness and refractive index of the glass between the wax and the lens correct shim thickness is found by trial and error.

The Linhof shims are set to factory specs on the screws that they sit on. The correct Boss screen for the Linhof has the bottom layer of glass cut off where the glass would sit on the shims so the wax layer is then at the correct position. Like the MAster Technika all V, Master Technika, Master Technika 2000 and Master Technika 3000 cameras have the gg position zeroed at the factory to the same position. So then it is simply child's play to change the gg and have it at the correct film plane position.

However, you can not assume that the bottom glass layer on the Linhof version of the Boss screen is the same thickness as the ones for other model cameras.

So, you know the proper procedure. The first step is to get the gg shim position at the correct position. Then modify the Boss Screen or get the correct one.

towolf

27-Oct-2014, 17:40

You can't measure where the wax layer in a Bosscreen is by focussing on it through the cover glass. All you get is where it appears to be because of the optical path length effect of the cover glass.

This is an easy test to do. Set up a view camera with the bellows off and exactly focus a subject. Now place a sheet of glass between lens and ground glass and you will notice the subject is now out of focus. To regain focus the length of the camera must be extended.

Yes, I’m wondering about that. But consider this:

You made two marks for focus positions between a close and a far point. Let’s say 5 mm without the glass. Now you insert your sheet of glass, but you also move the GG farther away until the subject is *again* in focus. The scale will still read 5 mm.

So with my setup, if I manage to keep the focus of my microscope at 4.8 mm and shim the Bossscreen appropriately so that it is in focus at 4.8 mm (even though the wax is actually farther away) I will have correctly shimmed the GG.

I made a quick visualization

http://i.imgur.com/kTaN4uv.png (http://imgur.com/kTaN4uv)

The wax layer of a Bosscreen cannot be placed by geometric measurement so as to coincide with the film plane in a film holder of correct T distance. It has to be placed optically so that when an image is sharp on the film it is also sharp on the wax. Because the exact position of the screen is influenced by the thickness and refractive index of the glass between the wax and the lens correct shim thickness is found by trial and error.

So much is clear, which is why I didn’t measure purely mechanically but with the help of finely adjustable optics.

VictoriaPerelet

27-Oct-2014, 21:06

Maris Rusis

27-Oct-2014, 22:48

VictoriaPerelet has the story right on Bosscreens and other configurations. Just for fun here are some numbers:

If the wax layer is covered by soda lime float glass (my guess) with a refractive index of 1.523 (my guess) and a thickness of 2mm (my guess) then the Bosscreen has to be shimmed back 0.688 mm so that the lens "sees" the wax layer at the same optical distance as the film. Different guesses calculate out to different shim thicknesses. All this is ok for paraxial rays only. For rays arriving obliquely, say at the edges of the sreeen from a wide angle lens, those rays have a greater pathlength through the coverglass. This results in image shift that gets worse the more the rays are off-axis. I've heard this phenomenon called "spherical aberration in the image space". I doubt that ordinary view camera work with real world cameras and subject matter will encounter these curious difficulties.

Bob Salomon

28-Oct-2014, 03:10

That is common problem for focusing screens that that have extra glass surface in front of focusing surface. This was discussed here many times, short summary here:

http://www.largeformatphotography.info/gg.html

Mainly extra glass needs to be compensated by "about" 1/3 of its thickness. But "about" depends on lens focal lenght, glass properties etc. that was not problem with film and moderately long 4x5 lenses. Arca used similar design with fresnel in front of GG.

Once high res digital backs become standard (well not on this site:)) and very short focal lenght become standard problem turned out real pain to deal with. Arca introduced N type backs etc, Linhof has gazzilions of focusing screen revisions, but problem was never solved completely. Sinar always used GG and had no problems.

Modern digital is focused on computer screen via live view and that is not problem anymore.

Overall with 4x5 film it is not real problem - film is art and not pixel pocking. Focusing gear pinion on your camera has bigger positioning error than 0.3mm. If this is is really annoying for you get plain GG.

Linhof revisions are simply screens with different rulings on them or screens of much much brighter output and both types have exactly the same placement in exactly the same holder. They do not change with the type of back used unless the new back has a different sized sensor.

A better example would be Rodenstock using the glass cover plate as the last element of the lens with some of their digital lens series and selling an optional glass plate that is used when those lenses are used with film.

towolf

28-Oct-2014, 05:35

All this is ok for paraxial rays only. For rays arriving obliquely, say at the edges of the sreeen from a wide angle lens, those rays have a greater pathlength through the coverglass. This results in image shift that gets worse the more the rays are off-axis. I've heard this phenomenon called "spherical aberration in the image space". I doubt that ordinary view camera work with real world cameras and subject matter will encounter these curious difficulties.

So, would you say my opcitcal method of establishing the shimming position is going to lead to more-or-less correct results?

I already noticed that when I use my microscope focussing method through the front or through the back I get a difference of 0.36mm. The rear glass (facing the user) is quite a bit thicker, so that would explain it.