Film plane alignmentgreenspun.com : LUSENET : Large format photography : One Thread
I recently bought a used roll film holder and found that a large number of shots were slightly out of focus, more than my normal quantity. I began looking at the alignment of the film in relation to the groundglass and found some problems. I have an Omega45E camera, a Graphic roll film holder, a Toyo roll film holder, and a variety of cut sheet film holders.
I removed the back from the camera and using a straight edge layed across the flange that attaches to the camera I measured the distance to the ground glass, I'll call this the reference distance. I made this measurement at several points across the glass. I then took cut film and roll film holders,loaded scrap film into them, installed them in the back and measured the same way. The distance to the film was between +.023 inches and -.005 inches of the reference measurement to the ground glass. The Toyo roll holder was the worst at +.023.
This seems like a wide tolerance. The +.023 inches could explain the out of focus shots, the problem was worse with close subjects with the in focus plane being farther from the camera than appeared in the ground glass. How do I go about fixing this situation. I plan on doing more close up work over the winter months where this problem will be exagerated. Are there better measurement methods that I should use to determine the extent of the problem? How do I fix them?
-- Dave Schneider (firstname.lastname@example.org), November 09, 1999
The tolerance for a 4x5 system is + or - .007". What this means is that IF the ground glass is in the absolute perfect position, film that lies within that range should render acceptably sharp results. When you move up to larger formats, the tolerance gets larger. Conversely, if you move down to 6x7cm, the tolerance gets smaller. The key here is that the ground glass has to be as close to perfectly aligned as practical. How can you measure this? Well, unfortunately making a depth measurement by referencing two points along the opposite edges of a film holder or opposite edges of a film back can be very misleading. The real depth can only be known if the measurement is referenced to the entire contacting surface. This measurement integrates all of the bumps, burrs, knicks, warps and other real and present defects in such camera backs and film holders, which is what happens when you make an actual exposure!
I overhauled 16mm motion picture cameras for years and believe me, on this small format the tolerance becomes very difficult to measure (hundredths of a mm)! The only way we could prove our measurements and adjustments were on the money was with a film test. Even then, we would occasionally run into a camera that measured perfectly in every way we knew how to measure such a camera and it would fail the film test. Often, it was necessary to disassemble the camera several times to find the elusive burr or bit of dirt that prevented the film from lying in the correct position.
Likewise, I believe a film test is also the most reliable way to prove performance on large format cameras. Ideally, if you could surface grind a flat plate of chromed steel, drill with several holes allowing measurement of let's say every inch across the film area and, knowing the specification of depth on a film holder, made comparative measurements of all your film holders and roll backs once you're certain the gg is in the right place, you could quantify all of it and know which items lie outside the range of tolerance. Don't forget film thickness and be mindful of pressure plate/plenum compliance! That's a lot of work. You would still have to test with film to prove you were right. What I might suggest at this point is that you read my article in View Camera magazine (Jan '96, I think) and do an initial film test with your gg in its current position. You might even run the test with a number of film holders to get a feel for the overall performance of the system in the 4x5 world. The test is pretty revealing and should tell you if you're close. Then, try the test with the roll backs. I can't get my C2N to work very well and I know my gg is on the money. I'll be curious to hear how you make out with your testing and would be happy to try to answer any subsequent questions arising from your experiments. Best of luck.
-- Robert A. Zeichner (email@example.com), November 09, 1999.
Dave, this is a vexing problem. It begins with fresnel screens and effective location of the plane upon which you are focusing. Fresnels on the lens side of the ground surface shift that plane, exactly how much determined by specifics of the fresnel lens. I don't know about your Omega 45E, but Toyo told me it places the optical plane of G-series fresnel/screen systems at 0.1875" from the seating surface. This is a manufacturer's decision. The ANSI specification calls for a 4x5 film holder to provide 0.197" +/- 0.007" from the seating surface to the septum. One must then assume a film thickness. Many use 0.007" to represent a typical film base, with a resulting optimum focusing plane distance of 0.190" - - though Ron Wisner has decided on 0.192" to account for wear on the back's wood, and in deference to users of Tech Pan with its thinner base.
Next we move on to actual film in holders. Sheet film is poorly positioned in standard holders. First, septums have that +/- 0.007" tolerance. Bad enough, but my measurements indicate that many don't meet the requirement and, among those which do, compliance typically occurs in only parts of the image area. Next, thickness of the channel allows film to move forward toward the lens. Finally, if you like T-Max film, note that it curls opposite of most emulsions, i.e. it will bulge out in the center. Modern Lisco/Fidelity/Riteway 4x5 samples all have an extra cutaway in the restraining channels near the loading flap to ease loading. This only makes matters worse, permitting even more departure from ideal position. To find out just how bad the visual effect of incorrect positioning is, check out Joe Englander's article in the April 1995 issue of Camera & Darkroom. The way out here is to upsize to 8x10. Even though it has a larger ANSI tolerance of +/-0.016", magnification in the final print is less, and results are visually better. I've gilded the lily by depth-checking numerous holders in stores and only purchasing those which are well inside the tolerance. This strategy has not been as successful for me in 4x5, where tighter tolerances can cause degraded result in severe situations, even with carefully selected holders. It seems one's only way out might be Schneider's Hi End Back system, but I've not yet concluded it's worth the significant investment required.
Finally to your roll film question. Above I mentioned severe situations. A factor which increases severity is lens focal length. With your roll film images you are likely using short lenses, and those demand much greater precision in film positioning. Unfortunately, roll film holders are usually much worse than their sheet film relatives in this regard. The culprit is reverse curl feed paths. When you thread film into a holder, the paper leader follows a circuitous route to the take-up spool. You then close the back and wind on to frame 1. Now make an exposure. If your holder is well matched to your focus screen, such as mine is in the Horseman VH and Horseman roll holders I have for it (Horseman selected 0.185" for both, and maintains good manufacturing uniformity), you will get incredibly sharp results. I have even confirmed this with a 65mm lens at f/5.6 shooting aperture. The problem is, after you advance to frame 2, the film currently in the gate had been sitting on the reverse-curling feed roller while you were working on frame 1. Its acetate base took a "set" and now bulges toward the lens. I have measured frame 1 at 0.185", +0 and -0.001" at center, -0.002" edges. However, subsequent bulging frames were as bad as -0.03", with terrible visual results. The way I've "fixed" this problem is to shoot only five 6x7 frames per 120 roll. I expose #1, then advance quickly and continuously past #2 to #3. That way the film now ready to expose has spent no time sitting on the reverse-curling feed roller. It's a film-consuming way to go, but the only solution I know of, and the one I use when an 8x10 is more than I want to carry.
Your Toyo results make sense, i.e. focused subject plane farther away than expected when the film is closer to lens than is the focus screen plane. I appreciate your post, since I have been curious about the Toyo holder. Short of actually buying one, there was no way to determine the film path employed; Mamiya America Corporation's telephone response didn't make things clear. Apparently, your back is about as bad as my Calumet slip-in roll holder, which measures similar film position departures. Your measurement methods seem fine to me. I can offer no suggestion on how to fix your problem, short of the Schneider back or using five frames per roll on a Horseman holder. Hope all this offers some insight why.
-- Sal Santamaura (firstname.lastname@example.org), November 14, 1999.
I repeated my measurements again this past weekend with a more systematic approach. I removed the ground glass from the back and measured the distance from the mating surface to the glass. This was done at several points around the perimeter of the glass. I then measured the same distance on several sheet film holders and both roll holders. Again the measurement was made at several points around the perimeter. With this method I found a much smaller variation between holders. I am now suspecting that I have some sheet film holders that are warped and may even have some twist in the roll film backs. The other possibility is that the back of my camera has a bit of twist. This would explain the large differences in my measurments. Anyway, I don't think the Toyo roll film holder is any worse or better in this regard than any other roll or sheet holders I have. I wanted to set the record straight on that. The Toyo holder is easier to load than many roll holders and there is definitely little possibility of curl setting in. I will post again if I get this sorted out. It's mostly an intellectual exercise. My poorly focused shots are equally distributed between all holders when I look at a larger sample. I need to work on my technique.
-- Dave Schneider (email@example.com), November 15, 1999.
Dave, I think your system of measurement is still flawed. Here's the problem: If you remove the back from your camera and measure the depth of your gg in relationship to the part of the back that mates with the camera, you do so under the assumption that that surrounding surface is parallel to the gg. You also assume it's perfectly flat, which it won't be. It may be close, but it really doesn't have to perfect. Only the film plane of the holder, when inserted in the back, must be coincident with the gg. If you were to make a perfectly flat plate that would rest on the mating surface of the camera back, drill a dozen holes or so to allow a dial depth gauge to be used to measure first the gg depth, then after inserting a holder, the holder depth, you will get results that mean something. Let's say your holders a a bit warped. When you insert them behind the springy thing that holds them against the camera, are they still warped? Hard to know, actually. But if you sample a dozen spots on the gg and then, the same dozen spots after inserting the holder, you will know the answer. You still need to test with film to be certain! Forget making measurements from points along the edge. If you catch a low or high spot, you can be off by several thousandths of an inch.
-- Robert A. Zeichner (firstname.lastname@example.org), November 15, 1999.