gg and film position, an old question revisited

[leonard_evens]

I've been using a Toho 4 x 5 view camera since July, and gradually

getting a better understanding of focusing. I have the impression

that most recently I typically have a modest systematic error favoring

the foreground over the background, and I would like to try to

understand it.

 

I've tried to check the depth of my Fidelity film holders with a dial

caliper which extends to function as a depth micrometer. The depth

seems to vary but I get something under .190 inches, which is the

nominal standard with film in place. But I don't place a whole lot

of credence in those measurements since the dial caliper is not really

designed to make such measurments accurately.

 

In my most recent test, using a 150 mm lens at f/5.6, I found I was

about 3 inches in front of where I intended to focus at about 8-9

feet. That corresponds to about 3 percent, which represents about a

0.2 percent error in the position of the film plane, or about 0.32 mm

or 0.013 inches. Since the focus in in front of where it should be,

that means the ground glass might be as much as 0.4 mm closer to the

lens than the film.

 

On the other hand, even focusing with a 7 X loupe, it is doubtful that

I can focus more accurately than about 0.3 mm on the rail. The depth

of focus at f/5.6 is probably that much when using a 7 X loupe.

 

I could of course make additional tests and get an statistical average

of the error. What I think I will do is shift my focus about 0.3 mm

forward on the rail (to favor the backround) of where the loupe says

it should be when taking pictures and see if I can detect any

difference. Fortunately, having rigged up a scale on the focusing

knob, I can manage to do that.

 

Also, I hope to receive a new, brighter gg soon, and maybe that will

help resolve the question.

 

Any illuminating comments? For example, have I managed to get it

backwards and the gg might actually be further away from the lens than

the film?

[david_goldfarb]

My first thought would have been focus shift from the lens, but since you've tested with the lens wide open, then it could indeed be a groundglass problem. Nonetheless, you should double check your focus with the lens at shooting aperture just to be sure that isn't contributing to the problem.

 

Are you sure everything on the camera is locked down before you insert the filmholder, and that you aren't moving anything in the process? The Toho looks like a great camera, but it's also very lightweight, and might just not be able to deliver the precision you need if you are often shooting with very short DOF.

 

Also try a 4x loupe. Unless you have a very fine groundglass or a Bosscreen, too much magnification can make it harder to see the image, because the loupe magnifies the focusing surface and the fresnel, if you have one.

 

If it's a consistent groundglass problem and the glass is closer to the lens than the film, then the easiest solution is to put some shims under the groundglass.

[ardea]

Leonard

 

I would tend to trust your dial indicater. Although it may not be that accurate,it should be quite repeatable to within a thousandth of an inch, (.001"). My suggestion would be to measure your film holders for depth with film inserted. I'm sure five measurements at each corner and averaged would give one a good base line for each film holder. I would then assertain where the image falls on the ground glass. For example the Bossscreen image falls on the paraffin surface that lies between the "ground Glass"; sometimes the ground on the ground glass is not facing the lens. That would be the point of measurement....but I am sure you know that. Basically measure your ground glass holder like you would measure your film holders.

Try measuring the distance on your GG holder to see if it matches the filmholder distance....

Merry Christmas

 

Richard

[robert_a._zeichner1]

Leonard, Measuring film holder depth is trickier than one would think. The compliance of the aluminum center plate makes depth measurement with all but the most sensitive dial depth calipers nearly impossible to do with accuracy. Additionally, the way the holder fits into the back and the spring tension of that back will likewise influence measurements. So, unless this is all carried out while the holder is retained by the back, the results will most likely be errant. You will also need a precision gound plate with holes drilled in the appropriate test positions to carry out such a test. The easiest way to make comparative tests, is to test with film at a wide aperture as you have done, but using a stepped test target that permits you to see what plane is coming into focus on the film as compared with the one that's in focus on the gg. I have adapted such a target for home construction and the materials list and assembly and test instructions are in an article I authored in the Nov./Dec.'96 issue of ViewCamera. In light of your intention to replace your gg, I would urge you to get a copy of this so you'll know what you are getting in to.

[leonard_evens]

In response to Robert:

 

I'm sure I would find your article interesting. I recently started subscribing to View Camera. It is a small operation which makes heroic efforts to serve its subscribers, but it doesn't seem to have a systematic method for downloading past articles---except for a few which have been put on its website. My local public library doesn't carry it. I presume View Camera could copy a previous article and send it to me, but with limited resources that would be bound to be expensive, and they don't say anything about it that I can find in the magazine or the website. Can you suggest a way I could get hold of the article?

[jerry2]

Leonard... it seems sooner or later we have all tried this. Initially I tested all my new view cameras with very price dial calipers also and the results showed all cameras to have mis aligned gg / film. However as the poster mentioned above, this method of measurement is extremely difficult to do accurately in most home set ups. (after implementing my focus test, described below, I found out all my cameras were perfectly aligned) Here is two suggestions....

 

If you really want to get this right, and you don't mind spending a few dollars, vs. all the time and film you waste testing, then find a shop that uses laser systems to test this. In USA, I know Mamiya USA has $40k laser that performs these tests...then the fix can be applied at the same time. This assures perfect alignment. Be careful with adding new screens after this procedure is implemented! The fee was pretty reasonable when I sent a camera in a few years ago.

 

If you want to pursue it yourself...your on the right track, however, you have to refine your approach. There is too much slop in the system focussing at 9 ft. If you have the bellows, use a 150 lens and focus at about 1 ft, about 300mm of bellows. Take a piece of paper with numbers printed accross each line, first line, 111111, etc. second line 222222222... etc.. do this all the way down the page. You will have to experiment with the font size, but 10 would seem about right. Now take this paper and place the middle of the paper about 1 ft from the lens. Tilt the paper on a sharp angle so the bottom of the paper closest to the lens is lower than the lens and the top of the paper is above the lens plane. Be sure this paper is placed on something very flat! View and shoot wide open!

 

Now focus the gg near the center of the paper so a given line number is in perfect focus, but the line above and below this are NOT in focus. You may have to experiment with the paper angle to acheive this. If possible only use back focus, so you can keep the lens set at a fixed distance from the paper. As you will notice, even the smallest increment you can adjust focus, you will notice immediate jumps of the line which is in focus. Once you have a full line in perfect focus lock everything down, take image. Then re check the gg to be sure nothing has changed. Therefore if your alignment is accurate, the film will show the same line in focus as your gg did. View the film, if the wrong line is in focus, then surely you have a gg / film alignment issue.

 

You can experiment by seeing how far off your focus was by re focussing the camera on the line that was in focus on your film. Regardless, unless you have excellent shop skills, the fix should be done by a shop that can shim, then re check with laser till its right along the entire film area. This assumes you want the ultimatte in gg/film alignment. If you shoot distant landscapes a small error will get burried in other issues, such as, film buckle, Depth of Field, Depth of focus, etc. However, if you do close up work, shoot at wide f stops, shoot brick walls, etc...then the problem will be very apparent.

 

As an example, focussing at 50 ft. vs. 55 ft, produces only a .1 mm change in the lens to film distance. As you know, most cameras and humans can not acheive focus down to .1mm. And even if we did, once we load a film holder, we will loose such precise accuracy. This may help you decide if you feel the problem is worth pursuing. Just play the with appropiate formulas and you can easily identify the usefullness of perfect accuracy and how much in accuracy you are willing to tolerate.

 

Hope this helps....

[john_randall3]

This to Bill L.

 

You have give one of the most complete and incisive

explanations of how to resolve the film plane/ground glass

alignment issue I have read yet. I hope the moderator of this

forum sees fit to archive this whole thread, cross-referencing if

possible, to assure easy retrieval.

Thanks again.

 

Best regards,

[jerry2]

John, thank you for the kind words. It's rare anyone takes the time to thank contributors. I too was baffled in my early days of View Cameras why there was no simple, home brew test for gg/film alignment. I was amazed at the mechanical methods people would attempt to determine if a problem existed. Not only can my method be used as a "rough cut" to see if a problem exist, it can also be used for very detailed analysis of how far the alignment is off and whether it warrants the $$ to have it repaired, or possibly with the right information, you can shim the camera yourself.

 

I would like to elaborate on one issue that may help some, as my above explanation was a bit brief. The benefits of this method is acheived by focussing at a distance that provides the best ratio between mis alignment distance vs. movement in the "plane of sharp focus" (PSF) distance. This distance is where the target should be located, it varies for each fl lens. As a general rule, the shorter the fl lens, the more accurate the test can be. I provided an example below to demonstrate just how accurate this test can be....

 

Using a 75 mm lens on 4x5, you place your target at 3500 mm from the lens. (Of course the accuracy of this number is irrelevant since all we are searching for is differences in gg focus vs. the same for the exposed film) At this distance....

 

.1 mm of gg / film mis alignment will equate to a change in the PSF of 20 mm. Within this 20mm field, if you have 5 lines of numbers to focus on, this will represent .1mm/5 = .02mm of mis alignment for each line the exposed film varies from the gg. Measuring down to .02mm by other mechanical means is near impossible, unless you are using lasers and have excellent jigs. This is the beauty of this test, for only a few bucks you can lay this issue to rest!

 

As I became more sophisticated in this test, I got some thin wood about 1/4" thick and glued them together in a stair step arrangement. Then I taped the line numbers to the wood facing the lens. I used about 20 pieces of wood to be safe. Each step moved 2 mm further away than the previous step. This method is much easier to see on film than tilting the paper as I described in my last post. With this arrangement I could measure gg / film mis alignment down to .01 mm. (.1/10 = .01 mm mis alignment per line, i.e. each line the film varied from the gg represented a gg / film mis alignment of .01mm) This was a bit overkill, since at these exacting standards other issues can create problems, mainly film buckle. I would reccomend 4mm stair steps, that is accurate enough for view cameras. But to my amazement, and I did the test very carefully, and was sure to always re check the gg after each exposure.... all my cameras were dead-on focus, other than a tad of film buckle. But yet my best attempts using all sorts of jigs and expensive measuring calipers produce results convincing me I had a problem.

 

One last caveat.... after your assured the middle of your film has passed the gg / film alignment test, you may want to confirm the entire 4x5 plane is also aligned properly. To accomplish this, run the test again, putting the focus lines at the top and bottom of the film by changing the focus position. As you can see, a thorough test can chew through some time and film...

[leonard_evens]

With respect to Bill L's last remarks:

 

I don't understand your arithmetic. The formula for the (approximate) relation between percentage error in image position and percentage error in subject position is

 

pct im err = - magnification X pct subj err or

 

pct subj err = -(pct subj err)/magnification

 

This is reasonably accurate for small errors, and the smaller the errors the more accurate it gets.

 

With a 75 mm lens and subject at 3500 mm, the magnification is

the reciprocal of 3500/75 - 1 or about 0.0218978. If the subject is at 3500 mm, the image is at 76.64233577 mm. Taking the image error to be 0.1 mm, as you seem to be doing, the percent error is about 0.1305 percent. Dividing this by .0218978 yields about 5.9595 percent. For 3500 mm, that amounts to about 209 mm.

 

I also did the arithmetic exactly using the lens equation to find

the difference in the subject positions for image position 75.5 and image position 75.6. It comes out to just about 234 mm.

 

What did I do wrong?

[leonard_evens]

I see now!

 

He meant 20 cm. That is the only way the apparatus he describes makes any sense.

[jerry2]

Leonard,

 

Sorry, I wish the 20 cm was it....but I think we are doing two different things here...I have seen your formula once before, but never applied it, so let me explain in more detail what I am trying to convey....

 

With a 75mm lens, if you focus at 3500 mm this will create a lens to film distance of 76.5mm. (Do we agree on that?) Moving the lens .1mm further away from the film will advance the PSF 210 mm further away from the lens. Hence why I wrote that .1 mm of lens to film distance equlates to 210 mm advance of the PSF. This is what we are trying to detect on film, since we can visualize changes of even smaller increments.

 

This is the same formula that tells us with 75mm lens, focussed at 150mm it will create 150mm of lens to film distance. So as you can see, I approached this not by looking for film error, the formula is only designed to advise you of the amount of lens to film distances for different focus points...I just exploited it to meet our needs in this application. Try working on this formula and see if it makes better sense. Please advise....

[leonard_evens]

Bill:

 

I think we substantially agree. (See below for the exact figures.)

 

But go back and look at your original response. In MY browser, you say ".1 mm of gg / film mis alignment will equate to a change in the PSF of 20 mm". 20 mm is not 210 mm. Did I miss something?

 

The exact calculations using the Gaussian lens equation 1/u + 1/v = 1/f yield the following.

 

1/75 - 1/3500 = 1/76.64233577

 

1/75 - 1/76.74233577 = 1/3303.424801

 

1/75 - 1/76.54233577 = 1/3722.065783

 

so the two displacements of the subject plane are about

 

3500 - 3303= 207 mm

 

3722 - 3500 = 222 mm

 

76.5 mm seems slightly off.

[jerry2]

Leonard, I am so sorry... I can't beleive I overlooked this, even after you notified me of such.... This is what happens when you post stuff after 14 hr workdays... To be clear to other readers...

 

I wrote... .1 mm of gg / film mis alignment will equate to a change in the PSF of 20 mm. Within this 20mm field, if you have 5 lines of numbers to focus on, this will represent .1mm/5 = .02mm of mis alignment for each line the exposed film varies from the gg.

 

Yes, this 20 mm is a typo, it should have been 200 mm, or as Leonard points out, 20 cm. In the next post I wrote the exact number of 210 mm. Geeezzz, sorry once again.... anyway, does this help you at all in your quest? Let us know how you make out.