500cm Body focus error?

<p>I bought a mid-1970s 500cm on ebay and have shot a few rolls through it. Replaced the light seals and all seems well. Then I decided to shoot a critical focus test. I used a Proxar 0,5m filter to focus on a newspaper, and underlined the part that I was focusing on. Shot at f2.8 with the lens focused at the closest setting with the proxar filter, with a tripod and used pre-release. The results I got today show the actual focus is slightly beyond what I saw in the finder. Not much beyond, about half a centimeter or so. So what I'd like to know, is this significant? Should I worry about this? I usually shoot at f8 or f11, and I think that shooting at these apertures will pretty much negate any minor focus error like this. I can scan the test frame if someone thinks that would help. <br>

I hope this is all within tolerance for any Hasselblad body.</p>

<p>Did you check at all four corners as well as the center?</p>

<p>There are two body problems that can cause a focus shift:<br>

1) The body has been damaged, resulting in the rear being out of square with the lens mount (or vice-versa). This is easily corrected if you have the correct jig, gauge, and hammer (yep... HASSY hammer). However, that setup cost $8,000 when I bought it many years ago. I don't know its current price.<br>

2) The mirror might be out of adjustment on one plane or another. You can check it by setting up a cross-hair target (suitably large), accurately positioned on the lens axis. With the camera on a tripod, confirm that the center of the target appears exactly centered in the finder screen. If not the mirror must be adjusted.<br>

This test is much easier with a grid screen or one with crossed lines. A micro-prism or split image at the center will complicate matters. You can put in a different screen for the test if need be. It won't introduce any error in the results.</p>

<p>And ditch the Proxar. Use a larger test target. All you care about is the size of the image on the negative. A picket fence works great.</p>

<p>- Leigh</p>

<p>Leigh,<br>

Without the proxar, I can't focus close enough to the newsprint to read it. I don't have an extension tube. Also, the only way to truly test precise focus is to do what I did. Any flat target would either be in focus or not in focus, and wouldn't show me the focus error, as I have found this way.<br>

Having the mirror adjusted means a body CLA, about $150 labor, and something I am trying to avoid here. <br>

I would hope someone could just tell me that 1/4 to 1/2 a centimeter focus error with the 80mm lens set at minimum focus is a normal error for these cameras.</p>

<p>Who said anything about a flat target? I certainly did not.</p>

<p>Since you already know all the answers, why did you ask the question?</p>

<p>- Leigh</p>

 

<p>OK, I didn't get your picket fence suggestion, sorry about that. I don't have a picket fence nearby, but I guess I could find one.<br>

I was hoping someone on this discussion area would know if the error I came across was something to be concerned about. The only solution I can think of is to get the body calibrated, or buy a much newer used Hasselblad (such as 501c or 501cm).</p>

<p>Dave,</p>

<p>I know that it was your hope that you will not have to spend money fixing something that was speaking when you formulated that question, but the answer to your question about the "normal error for these cameras" is "no error at all".</p>

<p>If you want a functioning camera, i'm afraid that the options you mentioned are the only ones: this one has to be fixed, or you need another one that is good (but how will you know when buying used? So test first.).<br>

May be that the mirror pads have gone and need replacing. May be that the mirror alignment is off spec for another reason. May be that the body length is no longer good.</p>

<blockquote>

<p>I would hope someone could just tell me that 1/4 to 1/2 a centimeter focus error with the 80mm lens set at minimum focus is a normal error for these cameras.</p>

</blockquote>

<p>I don't believe anyone can answer that question because the specs are not published anywhere that I know of.<br /> There are three adjustments in the body that would affect the accuracy of the viewed focus in the finder v. the actual focus on the film:<br /> 1) The flange focal length of the body, i.e. the distance from the rear mounting surface to the front of the lens bayonet. This length is set very accurately (±0.03mm) using a special jig and gauge.<br /> 2) The angle and plane of the mirror. This is set by adjusting two screws, using a special sighting device.<br /> 3) The support of the ground glass finder screen. This is set by four special screws which support the frame around the glass. This adjustment is set initially using a precision measurement gauge and jig, but the final adjustment is simply viewing three parallel lines from an illuminated source and centering the middle one between the two outer ones by eye. This is not precise by any means.</p>

<p>As I mentioned before, adding the Proxar makes it impossible to calculate performance parameters for the lens.</p>

<p>Using the data for the unadorned 80/2.8 CF we find that the depth of field is 1.2cm at its minimum focus distance of 90cm.</p>

<p>For critical work it's always best to measure the distance from the film plane to the desired focal point and set the lens focus ring to that distance. Obviously the ring is not calibrated in sub-cm increments.</p>

<p>- Leigh</p>

<p>The point of the preceding rather lengthy discourse is that the position of the focusing screen is <strong>NOT</strong> precisely controlled. In fact its positioning is quite subjective, dependent to a high degree on the skill and experience of the tech making the adjustment.</p>

<p>We would all like to think that the system is quite precise, but that's not reality.</p>

<p>- Leigh</p>

<p>There are several possible causes for this problem, as the previous posters mentioned. I suggest the following tests that could help you:</p>

<p>1. Take a picture of a distant scene with sharp edges (maybe a city skyline) with the widest lens available, wide open, focused at infinity. If you get a sharp image, then the body might not be the problem.<br>

2. Repeat your close focusing test, but this time do it twice. One with the camera in the normal position and the other with the camera upside down. If you get different results, it may be that the cushions where the mirror rest inside the mirror box need replacement. I think this is the most probable cause of the focusing problem. You need to know that this is a special repair and it is not performed in a regular CLA.<br>

3. If your get a sharp image at infinity and no difference in focus if you put your camera upside down, then either the focusing screen position or the mirror position might be the cause. (previous posters indicated how to deal with this)</p>

<p>The most common focusing error occurs when the screen is upside down in the finder. Hasselblad screens have a metal border. Only the upper edge of that border is rolled over the screen. The original 500cm screens were rather dark. It's possible that it was replaced with a third-party screen, which could affect focusing. In either case, shake out your piggy bank and get an Acute-Matte screen of some sort.</p>

<p>The second most common error is focusing by looking through, rather than at the ground surface of the screen. This happens because the diffusion is less than ideal for focusing in order to make the image appear brighter. Grid marks on genuine Hasselblad screens are embossed on the bottom surface. Make sure they are in focus to your eye when focusing the lens. Depending on your eyesight, this may require that the diopter of the magnifier be changed.</p>

<p>The magnification in a waist-level finder is about 4.5x, and is ideal for fine focusing (if not viewing in general). Alternately, you can remove the finder and place a loupe directly on the screen. Use a high quality loupe which can be focused.</p>

<p>The mirror is very low on the list of things which can cause focusing errors. If it is misadjusted, the most obvious effect would be a displacement of the image. The viewing screen rests on four screws, one at each corner. These are adjusted and sealed at the factory. If all else fails, this is the place to start, not by hammering on the mirror. Actually, the place to start is with Hasselblad service, or another qualified service agency - someone with a collimator.</p>

<p>There are enough paper weights to go around. Don't contribute to this excess by working on your camera without proper experience.</p>

<p><em>"1) The flange focal length of the body, i.e. the distance from the rear mounting surface to the front of the lens bayonet. This length is set very accurately (±0.03mm) using a special jig and gauge.</em><br /><em>2) The angle and plane of the mirror. This is set by adjusting two screws, using a special sighting device.</em><br /><em>3) The support of the ground glass finder screen. This is set by four special screws which support the frame around the glass. This adjustment is set initially using a precision measurement gauge and jig, but the final adjustment is simply viewing three parallel lines from an illuminated source and centering the middle one between the two outer ones by eye. This is not precise by any means."</em></p>

<p>It is, Leigh. Very precise, in fact.</p>

<p>The absolute focus distance error depends on the focus distance.<br />As you wrote, the maximum alignment error should not be larger than that 0.03 mm. That is 0.03 mm in 749.1 mm body length. That's 0.004%.<br />Calculated using the formulae that one uses to calculate such things, at the close focus distance of the lens, 90 cm, that would equate to an error in lens to subject distance ranging from -0.07 mm to +5.4 mm, depending on whether the lens to film distance is too long or too short.</p>

<p>The Proxar changes thing, and i'm too lazy to do the math.</p>

<p>Thanks for your replies. I have purchased a second hand Acute Matte D screen, and have it in my hands today. I will install it and try it out. The old screen, the original as far as I can tell, has very course circles, and they are a distraction.<br>

I should have mentioned at the outset that I have an A12 back with a non-matching insert. The frame spacing on the roll is fine, there is no problem there. But could this non-matching insert be the cause of this focus error? Maybe I need to spend a few hundred on a late A12 back with matching insert?<br>

I'll ditch the proxar filter on my next test, will shoot this weekend, when the rain stops.</p>

<p>Not being a matched set should not be a problem.<br>

But it being a used back might.<br>

Not as such, not just because it is a used back.<br>

But who knows what it has been through? The back too can be out of alignment.</p>

<blockquote>

<p>... the maximum alignment error should not be larger than that 0.03 mm. That is 0.03 mm in 749.1 mm body length. That's 0.004%.</p>

</blockquote>

<p>Hi Q.G.</p>

<p>The flange focal length of the lens is 74.9 mm (not 749). This is the distance from the face of the lens mounting bayonet to the film.</p>

<p>The body length (lens bayonet to rear body face) is 71.40 mm with an adjustment tolerance of ±0.03 mm, which is ±0.04%. Yes, that is quite precise.</p>

<p>But that dimension is from the lens bayonet to the back of the camera, not to the focusing screen. The dimension to the screen is not specified at all.</p>

<p>The adjustment procedure for the screen specifies a gauged dimension as a starting point, but from there the screen is further adjusted by eye.</p>

<p>I've never understood why they do it that way. It seems like an obvious source of error of the type described in the OP's first post.</p>

<p>Since the gauged preset position is to the top of the screen, but the focused image is on the bottom surface, it must have something to do with variations in screen thickness. Remember, these procedures were written 50 years ago.</p>

<p>Oh well. If it was easy, anybody could do it. ;-)</p>

<p>Thanks.</p>

<p>- Leigh</p>

<p> </p>

<p>1/2 meter focus error sounds like alot. Is this error at a large distance or a short one?</p>

<p>Hello Leigh,</p>

<p>Yes, i put the decimal in the wrong place (in the percentage too).</p>

<p>The calculations however are correct.</p>

<p>The distance from lens to focussing screen and distance from lens to film (!) are (or should be), of course, the same. So yes, it is specified.</p>

<p>The method used, 'eyeballing' something, is not problematic. When you read, say, a vernier calliper, you are eyeballing thngs too, and the result is easily as precise as we want it to be.<br />What is important is what it is you are looking at, and how that translates minute differences into something that can easily be seen. And the method they use is quite good. No worries.</p>

<p>They do not adjust, or look at, the top of the screen. They adjust the points the screen is resting on. So as long as the distance from resting point to the bottom of the screen is always the same. It doesn't matter that one type screen might be thicker than another.</p>

<p> </p>

<p>Hi Kelly,</p>

<p>It's 1/2 centimeter. <br>

But we don't know the distance to the subject. We do know it was short, perhaps 75 cm.</p>

<p>- Leigh</p>

 

<p>Hi Q.G.,<br>

Not meaning to be picky, but...</p>

 

<blockquote>

<p>The distance from lens to focussing screen and distance from lens to film are (or should be), of course, the same. So yes, it is specified.</p>

</blockquote>

<p>What something "should be", and what it's specified to be, are two very different things. The distance to the screen is not specified in any Hasselblad literature. The tool used to make the preliminary adjustment is the same one used to set the body length, but with a different adapter.</p>

 

<blockquote>

<p>The method used, 'eyeballing' something, is not problematic. When you read, say, a vernier calliper, you are eyeballing thngs too, and the result is easily as precise as we want it to be.<br>

What is important is what it is you are looking at, and how that translates minute differences into something that can easily be seen. And the method they use is quite good. No worries.</p>

</blockquote>

<p>We're not talking about vernier calibrations. We're talking about three bright lines, of significant width and separation, projected onto the focusing screen. Final adjustment of the screen position is based on observing these lines and centering the middle one between the other two.</p>

<p>Have you ever used this tool? I have, and it's quite a difficult procedure. You have to adjust the screws by a small fraction of a turn, and each the same amount. I've attached the adjustment procedure from the service manual.</p>

<p> </p>

<blockquote>

<p>They do not adjust, or look at, the top of the screen. They adjust the points the screen is resting on. So as long as the distance from resting point to the bottom of the screen is always the same. It doesn't matter that one type screen might be thicker than another.</p>

</blockquote>

<p>The tool that they use is a focusing screen, but one specially designed for making this adjustment. It's not a regular screen. The initial adjustment of the four screws is gauged from the top of this screen.</p>

<p>- Leigh</p>

<div></div>

<p>I was injecting that 500cm in title was focus error :) :)</p>

<p>That would make 5 meters, anyway ;-)</p>

<p>Leigh,</p>

<p>You have to understand that that "has to be" is an absolute must. So if you specify one dimension, you also specify the other.</p>

<p>The vernier callipers were an example. The principle is that if you can make something small big enough to measure using whatever you like, the imprecision that other method (looking at something, like, for instance (!), how lines line up on a Vernier callipers) is prone to will be decreased by the same factor as the thing it is supposed to meassure is increased, magnified, to be able to use that method.<br />If, say, we can judge whether two lines 1 mm thick line up or not just by looking at them, with a precision of 0.5 mm, all we need is a microscope providing a 100x magnification to be able to judge with equal ease whether two lines 0.01 mm thick line up with a precision of 0.005 mm.<br />Etcetera. You get the drift: you don't judge a metering method by the way you read the scale.</p>

<p>That is: you do, but you shouldn't. ;-)</p>

<p> </p>

<blockquote>

<p>I was injecting that 500cm in title was focus error :) :)</p>

</blockquote>

<p>I suppose, were that the case, the error would be significant.</p>

<p>But it's the camera model number (properly 500C/M).</p>

<p>- Leigh</p>

 

<blockquote>

<p>... you don't judge a metering method by the way you read the scale.</p>

</blockquote>

<p>Quite the contrary.</p>

<p>Not only can you judge the method by its implementation, but you MUST do so if you expect repeatable results.</p>

<p>That's why precision gauging systems incorporate features to eliminate parallax error, among others.</p>

<p>Human error is the greatest source of inaccuracy in any metrology system.</p>

<p>- Leigh</p>

<p>You think so?</p>

<p>Anyway, the thing to be measured is translated into something we can quite easily judge using our error prone human vision while still producing results with the desired precision.<br>

The mark of <strong>successful</strong> "metrology".</p>

<p>Stop fretting over what is not a problem! ;-)</p>

<p><em>Have you ever used this tool? I have, and it's quite a difficult procedure. You have to adjust the screws by a small fraction of a turn, and each the same amount. I've attached the adjustment procedure from the service manual.</em></p>

<p>I don't quite grasp what this tool is, nor its principle of operation. But from a practical aspect, you have answered your own question. Since you have used this tool, what change would you observe if you move a leveling screw by 1/10th a turn? Let me explain...</p>

<p>The adjusting screws have a certain pitch, say 0.5 mm (probably finer). From the context of your reply, I surmise that changes in the line spacing of the [collimator] due to small adjustments of the leveling screws is easily visible. 1/10th of a turn would represent 0.05mm, which is comparable to the focusing tolerance assigned by Hasselblad. If the pitch were 0.25mm, the tolerance would be spanned with 1/5th of a turn.</p>

<p>Gauging the rough setup from the top of the special screen is appropriate if you know the thickness of the screen. That's another parameter with its own tolerances, but this is a "rough" setting by your own words. The actual adjustment is based on an image of the [collimator] on the bottom of the screen, the same as when the camera is in use.</p>

<p>The <strong>optical </strong>distance of the flange to the bottom of the screen should be identical to the distance from the flange to the film plane. Unlike the film plane, there is no way to measure this distance directly, rather using some sort of optical tool. The term "should be" reflects the intended results, subject to adjustments, described with typical scientific caution - adjustments have tolerances, so "precisely" is not a term to be used lightly.</p>

<p>"Precision" is a term used by metrologists when errors are expected, but you have some idea of what the maximum error will be. We incorporate the concept of precision when we say the temperature is 20 deg C +/- 0.5 deg, putting a number on the expected (maximum) error in this measurement.</p>