Ground Glass and film plane alignment - how is fresnel lens thickness taken into consideration?greenspun.com : LUSENET : Large format photography : One Thread
There is a very good article in May/June 99, Photo Techniques magazine about this subject. The author spent six years using a new 4x5 view camera not realizing the film and ground glass were mis aligned by the thickness equal to 3 sheets of film. He thought enlargements were suppose to get blurry becasue he had nothing to compare them to. I have been hearing this from many LF photographers, many of them using new equipment!
The method of testing this seems quite easy (with the proper measuring device of course) - with the 4x5 back removed, measure the distance from the back to the ground glass and then insert a film holder loaded with film, and measure from the same back position to the closest part of film. With any luck they should be close to equal. I am unsure what tolerance should be considered acceptable? Does anyone know? However, what none of these articles ever mention... where do you measure to if there is a factory installed fresnel lens in front of the ground glass. Should you still take your measurement from the back to the fresnel, or do you remove the fresnel lens and measure through to the ground glass? Any input would be appreciated. Thank you. .
-- Bill Glickman (Bglick@pclv.com), April 27, 1999
The tolerance is 0.007 inches, which happens to be the typical film thickness. Take the fresnel out before measuring.
Another good test is to put a newspaper flat on the ground, mark the middle, focus on the mark, and photograph it. The resulting photograph will show if the ggs is out of position.
-- Alan Gibson (Alan.Gibson@technologist.com), April 28, 1999.
Dear Bill and other contributors, Yes it is a big problem and indeed an underestimated one even for camera builders. The procedure which is described in the question would be the standard procedure accepted by most, even though, like some other threads have been saying earlyer, it is prone to give erratic results given the poor constuction qualities of most cameras in terms of micrometric precision. However, when I bought my Wista VX, I disliked immensely its ground glass. Like in many cameras, critic focussing would be difficult if not impossible because Wista integrates a groundglass-fresnel in one (plastic) item covered by a clear glass with a grid. When using a loupe with an eccessive magnification (7x) you just cannot focus.A 4x helps but not much. Let alone the fact that, due to the fresnel lens you can only view the image on the groung glass with one eye opened and in a certain position at a certain distance from the camera. Wista doesn't think that this is a problem because most of the customers do not complain, I did. Back to the original problem. To solve the problem, I ordered a Bosscreen 4"x5"( a kind of brightscreen wich is made from two pieces of glass with in between a very thin layer of bee-wax and paraffine, very , very, fine structure!) for my Wista. Well first I got a ground glass which was too large for the camera then one which was resting higher than the focal plane. When asking if they had ever made any focal plane measurements they answered that these had been made at some stage in the past from the firm's previous owner. I didn't get the impression that the guy knew what I was talking about, anyway, the second time the glass was resting too far from the focal plane and to prove it I set off to device an empiric demostation of this fact( he didn't get the theoretical aspects of the matter!). You take a ruler which has measures in clearly visible numbers. You want to use your longest focal length lens and the one with the largest aperture too.Place the camera on a tripod, fully extend the bellows and mount your lens, place the ruler in front of the camera at a suitable distance so that you can read on your ground glass a certain amount of numbers (obviously your camera, or rather its standards, front and rear, need to be at a 90 degree angle in relation to the position of the ruler). Try to make shure that you are focussing (precisely) say at the number 80 of the ruler (the real distance is irrelevant) by moving the ruler to and fro the camera. Once you are sure to see a measure clearly, shoot a Polaroid 55 or a film or even better both( so you can check if film and polaroid are on the same focal plane) at the largest aperture of your lens. When developing the film and or the Polaroid 55 you will see if the focussing of the film plane corrisponds to the image that you saw on your ground glass. If numbers which where closer to the camera appear sharper than the chosen measure, it means that the glass is too far at the back of the focal plane(!!!!) and the reverse for numbers which are further away. The solution in this last case might be spacers. Bosscreen acknowledged my test and modified the glass accordingly, however, being very critical about focussing amounts to a theoretical problem, in most cases because, and you can verify this yourself, the focussing capabilities of your eyes are erratic and are due to different circumstances. Try focussing several times the same object set nearby your camera, if you mark the distances of the standards on your camera, you rarely will come to the same distance (in tenths of millimiter) twice and your ground glass will never be much more off-set than this little! It makes sense to go to extremes for extreme photography (very close-up still life, microphotography, photogrammetric measurements) but than you would want to buy a specially deviced back with integrated ground glass and film holders like Sinar or Linhof produce for exeptionally extravagant prices! So, at times, it can be a problem but if not an evident cock-up, like mounting the frenel lens in a inverted position with the groung glass(!) this small differences shouldn't be a real problem. However, to answer you question, some builders put fresnel first (seing the camera from the back) and then ground glass and some do exactly the opposite! So ask the ones who built your camera or else do my empiric test. I am not an expert like some other contributors are, so, if I said any wrong things please forgive me! I do not intend engaging in one of those internet fights which all too often appear on this forum! I am prepared to accept, without further comments, any corrections or contributions. I'd like to remind all of you of the existence of the World Field Photographers Association (WFPA) and once again put its URL to your attention. http://www.johndesq.com have a look an join in it is fun and it is free! regards
-- andrea milano (firstname.lastname@example.org), April 28, 1999.
Alan, .007 is what I was afraid to hear. On a brand new 8x10 metal field camera with all new film holders many of my measurements are .025" apart from ground glass to film! This is equal to the thickness of 4 sheets of film! It appears the film holders and ground glass are not consistently off in any one direction, therefore shims would not help!
Milandro, your input was very helpful, and now I don't feel alone like everyone makes you feel. The point is here, if the equpment is off, it just adds to the errors that us humans are prone to, eyesight, camera shake, camera movement when loading film holders, etc. So if I am off the equivalent of 4 pieces of film to start and then add my field errors, its amazing even the chromes looked focussed. Of course I first noticed this problem when small enlargements were made. I figured if I bought all new equipment and paid the extra dollars, I would not have to deal with issues like this, but how wrong I was! I imagine for older used cameras things could be much worse! I was amazed that Wista lacked knowledge or quality controls in this area, the whole purpose of a camera is to block light and maintain ground glass and film alignment. Whats even worse is, I bought the same brand film holders as the camera maker to assure all components are speced by the same manufacturer, and it did not help at all! Very frustrating, but better off spending the time to fix the problem now vs. the tremendous traveling expenses to shoot on location and film / processing cost to notice it after the fact!
-- Bill Glickman (Bglick@pclv.com), April 28, 1999.
I bought a used Horseman 980 with a 105 and 180 Topcon lens. Testing them with Newspaper target, I was horrified to see the 8x10 enlargements of negs taken wide open- FUZZY. I measured the distance from the G Glass to the bosses that meet the camera body and then the film distance to the same bosses on the film back. The measure was off the thickness of the glass!! The PO had the Glass in upside down.
-- George Nedleman (email@example.com), April 28, 1999.
The ANSI standard for the depth of a standard 4 x 5 inch film holder is 0.197" plus minus 0.007" . Most film has a base of 0.007" . When film is loaded in the film holder, the depth is 0.190" . This is the measurement used by Sinar cameras. Wisner cameras use a compromise of 0.192" to allow for wear on the wood and because Tech Pan film, used by some photographers to achieve ultra-sharp images, has a base of 0.004".
The ANSI standard for 5 x 7 and 8 x 10 inch film holders are as follows (film thickness has not been deducted):
5 x 7" 0.228" + - 0.010
8 x 10" 0.260" + - 0.016
See my View Camera Construction FAQ at: http://home.sol.no/~gjon/lffaq.htm
-- Jon Grepstad (firstname.lastname@example.org), April 28, 1999.
To clarify the various issues presented, and introduce a new one, the problem -- of getting accurate ground glass and film plane alignment -- has three components. First, is the ground glass and/or fresnel correctly installed (i.e. is the fresnel behind the ground glass, as it almost always should be)? That is the easiest to solve. Second, did the manufacturer, in designing and constructing the camera, ensure that the ground glass and film plane would be within acceptable tolerances. Third (here's the new problem), film, whether sheet film or roll film is not flat. Short of using vacuum backs (there are probably other problems there), even if the other two problem issues are resolved, there will always be problems with critical focus. Therefore, conducting empirical tests (i.e. shooting newsprint) by shooting your lens wide open will most likely lead to disconcerting results.
-- Howard Slavitt (email@example.com), April 28, 1999.
Bill says "It appears the film holders and ground glass are not consistently off in any one direction..." It could be that the holder or GGS is twisted (in which case it should be trashed). Or do you mean that the difference varies over time, possibly due to temperature effects?
As Jon points out, the tolerance for 10x8 is 0.016", so 0.025" isn't too bad. If the ggs and film are not in the same plane, they might still be parallel, or they may be non-parallel (e.g. correct in the centre, but the ggs may be too far forward at the top, and too far back at the bottom). Non-parallel is a harder condition to identify.
If the manufacturer isn't aware of these issues, I can find no excuse.
However, my view is that we needn't be too paranoid, because we have depth of focus, i.e. a tolerance at the film plane due to the small aperture of the lens.
Depth of Focus = 2.C.N'
where C = permissable circle of confusion, and N' = effective aperture. For C = 0.1mm and N'=16, Depth of Focus = 3.2mm or 0.126", and this is eighteeen times the official 5x4 manufacturing tolerance. (Of couse, it is much less wide-open.) If your subject is essentially a plane, such as a landscape (or newspaper), you will be using movements to attempt to keep it all in focus, and the Depth of Focus will easily take care of the manufacturing tolerance. The film may not be flat, but this will also be easily taken care of by the Depth of Focus.
How accurately can we focus? It will depend on the camera, GGS, and loupe, but I think my focusing tolerance is slightly worse than the manufacturing tolerance.
If your subject is not a plane (e.g. railway tracks leading to a building, and you want everything in focus) then you are relying on some Depth of Field, and this will reduce the Depth of Focus available to you.
-- Alan Gibson (Alan.Gibson@technologist.com), April 29, 1999.
Alan - Your response really puts things in perspective! You have applied the simple depth of focus formula to this issue, something I admittedly overlooked to see what the potential effects mis alignment would be. Sometimes we can not see the forest through the trees, and this was a case in point. It's so applicable, I suggest the main page have a link to explaining this, because at one time or another someone with focussing problems will always need this information, and there seems to be not one all encompassing explanation of this. This responses in this forum is about as good as it gets!
Also I would like to point out, that measuring these problems with dial depth guages and micormeters is not an easy task. The problem is getting a completely level reference source to measure from that has no flex in it, specially for 8x10. The the reference source can not be moved between the different measurements, i.e. glass measurement vs. film holder measurement. I was told by a manufacture that even their testing methods proved to be inaccurate, although much more accurate than our home methods, until they purchased a $40,000 laser device that does all their checking. So my point is here, get a good reference piece of material such as a milled flat piece of metal with holes drilled in appropriate places, than a method to assure the depth gauge is hitting the glass or film holder almost perfectly perpendicular, then still treat your results cautiously. If you are off say 50%, as Alan points out, it seems depth of focus will correct any error that small. How far off until one is concerned would need to be evaluated taking all varialbes into consideration. If the test shots prove their is an alignment problem, then maybe you should send the equpment to be aligned with a laser machine and good repair man. I thank everyone for there input, it has been a great learning experience for sure!
-- Bill Glickman (Bglick@pclv.com), April 29, 1999.
Bill made some very good points. these tests are not easy and require statistical verification for repeatability and accuracy. Additionally, other potential pitfalls when using aluminum plates as reference planes is thermo-dimensional changes. With some care and practice, though, i think it is possible to come up with reliable figures, at least for 4X5. Going up in size will make things much more difficult. Julio
-- Julio Fernandez (firstname.lastname@example.org), February 08, 2000.
Bill, Let me add my (long) 2 cents worth for you and for the benefit of others following this thread that might have the same problem. I own a Wista with the frenel/plastic one piece unit that you describe and am aware of the focussing problems you describe. Depending on the focal length of the lens, the light leaving the ground glass is more or less perpendicular to it. Less when the lens focal length is further from the length that the frenel is optimized for (I assume a normal focal length is optimal). Therefore, extreme wide-angle and long lenses will result in light leaving the gg at an angle, making it difficult to use a high-power loupe resting flat on the glass surface. The solution to this problem is to use your loupe perpendicular to the rays and not parallel to the ground glass. This is done by (in my case) turning the loupe over (or removing the barrell) so that there is nothing between the actual lens in the loupe and the gg. Then one can easily find the proper angle and distance from the gg for proper focussing, even in the corners. I use an 8x loupe and have no problems with this method. If the concentric circles of the frenel lens or the texture of the ground glass are/is in focus then you are at the right distance. The proper angle is found by searching off-axis until you find the brightest image. Really a lot easier to do than to describe. As for the proper alignment and relationship of frenel and gg I quote Ron Wisner from an article I snatched from the Veiw Camera Magazine site:
"One subject which comes up often is the correct position of the fresnel lens. In years past, several arrangements have been used, including placing the fresnel behind the ground glass, in front of the ground glass, and incorporated into a ground glass made of plastic. However, there is only one correct arrangement. The ground glass surface should face the camera lens, and the fresnel is placed behind it, on the outside of the camera, facing the photographer. The textured surface of the fresnel should be placed down, against the ground glass.
There is one particular reason for this arrangement. In manufacturing cameras and film holders, one overriding concern is the correct position of the focal surfaces of the respective parts. In the film holder this would be the position of the septum, against which the film rests, and in the camera this would be the position of the diffusion surface, or ground side of the ground glass. Nothing is more important than the proper registration of these two elements.
If the fresnel were placed in front of the ground glass, interposed between the lens and the ground glass surface, a lack of registration can occur. This can be explained by considering the effect of a parallel sided glass plate such as a glass filter on a beam of light. Rays passing through such a plate are displaced by about one third the thickness of the plate, and depending upon the angle when passing through the plate, will be displaced laterally as well.
Furthermore, the greater the angle, such as in a wide angle lens, the greater the effect, resulting in an apparent curvature of the image, when no such real curvature exists. The result will be erroneous focusing of the edges of the image."
Testing to make sure your gg and film are in approximately the same plane and properly aligned, coupled with the correct frenel/gg arrangement should get get one close enough for anything but the most critical work. Also, measure ALL of your film holder, especially if you have a motley collection of old and new, wooden and plastic such as I do and discard those that are off. Hope this all helps. Regards ;^D>
-- Doremus Scudder (ScudderLandreth@compuserve.com), April 30, 1999.
After applying the Depth of Focus formula, I came to the conclusion that my film holders / ground glass alignment was not my problem. To test this, I devised the following test that worked out great. Using this test method, it can determine how much out of alignment your ground glass / film holder is. Here is how I tested it.
I typed up on a 8./ x 11" piece of paper on landscape orientation the numbers 1 accross the top line, 2 across the second line, etc. all the way down to the bottom, line 63. The size font you use will need to be experimented with based on the lens focal length and the format size. The goal is to get the paper to take up the full ground glass left to right. Put the test paper on a tilted typing stand or lean against a wall using a straight supporting piece like glass. Change the tilt angle so that only one line is in focus. (you can use 2 lines, but for the purpose of this discussion I will use 1) You note that line 32 is in focus, and lines 31 & 33 are partially in focus, and lines 30 & 34 are not in focus. Of course you need to use at least a 4x loupe on the ground glass to determine this. Then, take your picture and look and examine the results. If the lines of exact focus are off, determine exactly how many lines off they are. Say, one line off. By knowing the tilt angle of the test paper you can determine with a little bit of geometry the exact distance apart each line is from the lens. This will give you a good indication of how far off, and in which direction your alignment is.
-- Bill Glickman (Bglick@pclv.com), May 03, 1999.
Dear Bill, Maybe you should go over to Phoenix and measure an Arca Swiss or Sinar P camera and see how those numbers compare to your Toyo cameras.
-- Ellis Vener (email@example.com), February 08, 2000.
Dear Bill, Maybe you should go over to Photomark in Phoenix and measure an Arca Swiss or Sinar P camera and see how those numbers compare to your Toyo cameras.
-- Ellis Vener (firstname.lastname@example.org), February 08, 2000.