| Author |
Message |
    
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 174
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 12:10 pm: | 
|
I am trying to fix the focus of a Konica Auto S2 RF. There are two lens parts in the shutter assembly. One can be screwed off from the back, the other from the front. The rear one has not been moved. The focus seemed to be off a bit.
This is my first attempt at custom collimation and focus correction. I did some background reading describing various methods on how to do this.
I used an SLR with known good focusing lens to verify the focus of the RF. I discovered that I needed to loosen the front lens part to get the infinity focus right. When an object at infinity is in focus, the lens distance scale has its infinity mark aligned with the distance index marker on the body. So far so good.
The rangefinder focusing works at infinity and at smaller distances now, i.e. the parallax correction corresponds to the focus in the film plane.
However, I notice that at shorter distances the distance scale does not correspond to the actual film plane to object distance. e.g. an object at 3m when in focus is shown to be at 2.85m or so on the lens distance scale. Is this expected? If not, what am I missing in the algorithm to fix the focusing of the camera?
Thanks. |
Dirbel
Tinkerer
Username: Dirbel
Post Number: 32
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 02:02 pm: |
|
If the rangefinder and the actual object distance match, the numbers on the focussing scale shouldn't matter.
I'm not sure how precise the focussing scale is supposed to be, but 5% (3m vs. 2,85m) doesn't sound too bad.
Ciao,
Dirk |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 175
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 02:16 pm: |
|
Yes, I was supposing that as well. My only concern was that the scale appears to be non-linear with respect to the true-vs-displayed distance. I was not sure if that is usual or acceptable. |
Scott
Tinkerer
Username: Scott
Post Number: 90
Registered: 07-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 03:13 pm: |
|
There's no way to shift that distance scale anyway, is there? And if the lens groups are screwed down to the bottom of their threads, then they're where they should be.
I agree-- if you're confident that the RF is showing you what's at the focal plane, then that's as precise as you can get. |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 176
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 03:23 pm: |
|
The front element is not screwed down all the way now. I had to unscrew it by a few turns to get infinity focus right.
With infinity at focus, the distance scale agrees with it. The distance scale is off by a bit only at shorter distances. But I can live with this. I wasn't sure if this was a symptom of a non-obvious problem in the optics' arrangement.
Now, with this front element in correct position but "loose", how do I secure it there? I wonder if I should insert some shims, between the shutter and the lens, made from tinfoil or if I should put a spot of adhesive to fix the lens in its position. Any suggestions? |
David_nebenzahl
Tinkerer
Username: David_nebenzahl
Post Number: 241
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 06:46 pm: |
|
It's not clear that you're doing things in the right order here, at least so far as I can tell. (Disclaimer: I'm not familiar with this camera at all.) My apologies if I'm missing something here.
The usual course of focus correction for a rangefinder camera is in two steps: first the actual focus is corrected, then the rangefinder is adjusted.
It sounds like you're saying that the actual focus of the lens is off, correct? In that case, the first step is to get the lens focusing correctly. This is done by adjusting the focus at infinity, and can be done with a collimator (in the form of a 2nd camera) as you described. So far, so good.
Except that you shouldn't be messing with either lens element, screwing them in or out or shimming them, assuming that they were correctly placed in the shutter (of course, that's an assumption you can't always make, but it's a starting point). Both elements should be completely screwed in, with whatever original shims were there, if any. You then adjust the focus by moving the focus ring/collar/sleeve/whatever means is used to focus the lens (this is a manual focus camera, correct?) where it is secured to the focus helical. Forget about the distance scale for the moment: you just want to get the lens focused at infinity (at the film plane) when the ring/collar/whatever is rotated all the way towards infinity.
After the focus is corrected, then you can adjust the rangefinder (also at infinity).
So far as the scale being off, that remains to be seen until you're sure the focus is correct. At that point, the scale should be fairly accurate; the discrepancy you reported (showing 2.85m when actually focused at 3m would not be correct, unless the scale was incorrectly printed or engraved, which would be surprising.) |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 177
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 07:00 pm: |
|
David: I seem to understand what you just described. However, I am not sure about your description of adjusting the focus (4th para).
If I may rephrase, this is what you are suggesting:
1. Screw in the lens elements securely.
2. Adjust the infinity focus using some form of a collimator but only by turning the lens focusing ring on the lens barrel.
This is exactly what I did initially to verify the focus. The only variables that I can see are the lens elements' position and the focusing knob. The former are fixed in your method. That leaves the focusing knob.
When I tested the camera's focus, infinity was being focused a bit before the knob reached its end. So infinity was being focused a bit prematurely. That is what prompted me to tweak the position of the front element.
Perhaps there is something else that I am missing? I did take out the lens mount from the body, cleaned a few things in the shutter and reseated it back as it was before.
Or have I missed something completely in your post? |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 178
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 07:21 pm: |
|
hmm ... I think I see what you meant. I will now try to see how to tweak or reposition the focusing ring itself on the lens barrel and what effect this will have on the distance scale.
Apologies for late realization.
Thanks. |
David_nebenzahl
Tinkerer
Username: David_nebenzahl
Post Number: 242
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 07:37 pm: |
|
You seem to have got it now. It's pretty simple, really. Basically, don't count on the focusing "knob" (ring/barrel/sleeve/whatever) being in the correct position. |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 179
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, March 31, 2010 - 08:44 pm: |
|
I have it sorted it out more or less. Now I just need to wait till I get some time to work on it in bright daylight and I can fix the one or two centimeters error that still remains. Thanks.
BTW, I searched again, apparently with better words this time, and got this thread that actually gave a pretty good outline as to how to go about doing this:
https://kyp.hauslendale.com/classics/forum/messages/2/14425.html?1233179312
Based on that: the two chrome screws on the focusing ring can be taken out and the focusing distance ring plate comes out. There are three grub screws holding the focusing ring. Loosening them allowed the ring to be moved without moving the lens. So I first focused at infinity and then loosened those screws, positioned the focusing ring at its infinity extremity and secured it back by tightening the screws. Remounting the distance scale ring plate was the last step. |
Scott
Tinkerer
Username: Scott
Post Number: 91
Registered: 07-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 01, 2010 - 12:58 pm: |
|
Of course, you have to find the solution that works best for you, but I would recommend starting the process by screwing all lens groups/elements snugly into their threads in the shutter housing.
If you want to see if shimming will help, I think the only apropriate place for shims would be between the back of the shutter and the helical ring (where the shutter is mounted to the helical with the slotted ring-nut. That is where the manufacturer would have shimmed it.
As far as discrepancies between actual focus (at the film plane) and apparent focus (in the rangefinder or viewing screen), I am much more concerned with discrepancies at close distances. That's just because, in my experience, focus errors at close distances are more apparent than focus errors at far distances.
The effect of DOF makes errors at infinity less apparent. That's why I prefer to check at close distances (like six feet) rather than at infinity. |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 180
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 01, 2010 - 02:39 pm: |
|
Scott: Good point about infinity focus being less critical than nearer distances, especially within 3m. In fact, I first noticed the focusing error in portrait shots when I ran the first test roll in the camera. From a little side view, the eyes of the subject were soft but her shoulder, nearer to the camera, was sharp.
As David suggested, I restarted the method by screwing in the lens elements snugly (well, only the front one, the rear one is already at its place as I received it and it is quite fixed there).
This time I adjust the focusing ring, as I previously described. Focus seems to much better than before. Only that at closer distances, say at 1m, the scale is off by a few cms. An object at 1.2m is shown to be at 1.30m or so on the scale. However, the rangefinder agrees with the focusing plane at all distances from infinity to the nearest, well, as far I can make out with naked eye.
I am not sure if that error is due my own error in determining exact focus in a SLR-based collimator I used. For now, I am going to leave it there. |
David_nebenzahl
Tinkerer
Username: David_nebenzahl
Post Number: 246
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 01, 2010 - 03:05 pm: |
|
You might re-check the focus using a focus screen (and I still say use a glass one instead of a piece of tape) to see if it agrees with the collimation. If it does, you're good to go.
It's a little hard to check focus on a little bitty screen, especially if you're up in years. I use a loupe (8X) to see clearly and try to focus on something sharp and distinct, like the telephone poles in the distance down the street.
We've had this discussion before about what distance to check focus at. Even if one never takes a picture of a subject at infinity, the focus should still be checked at this distance. Infinity is the only place on the focusing scale with a positive stop (the focus ring/barrel/etc. can't go any farther than that), and it's the easiest to determine correct focus for (no measuring necessary). |
Aphototaker
Tinkerer
Username: Aphototaker
Post Number: 181
Registered: 12-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 01, 2010 - 04:26 pm: |
|
Yes, I will try with a better screen next time. I do not have a focusing screen at hand for this purpose. I will try ground glass and see if that helps with accuracy.
I am using a 50mm f/1.8 lens as a loupe. I might try with a longer one next time. |
Mareklew
Tinkerer
Username: Mareklew
Post Number: 34
Registered: 03-2010
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, April 04, 2010 - 03:18 pm: |
|
David: you can always get a split-image focusing screen and put it on the film rails. The only thing is, you have to look at it EXACTLY straight from the back. There's a trick to do it: adjust aperture till the wedge starts to darken. If you look straight, both halves will start darkening at the same moment. If one darkens first, you aren't looking straight square. Then check focus.
Marek |
Mareklew
Tinkerer
Username: Mareklew
Post Number: 35
Registered: 03-2010
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, April 04, 2010 - 03:20 pm: |
|
Forgot to mention: for the trick to work, your split-wedge has to sit exactly in the optical axis of the lens, but that's relatively easy to adjust, at least close enough.
Marek |
Fatbear
Tinkerer
Username: Fatbear
Post Number: 1
Registered: 02-2012
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, February 22, 2012 - 07:22 pm: |
|
I suspect the focusing distance on the scale is probably to the optical center of the lens, not to the film plane. Setting infinity focus by pointing at something infinitely far away (or into a collimator) and making sure the image is in focus on the film plane is very convenient and will remain pretty accurate at even moderate distances. But when you adjust the focus of a lens you are adjusting its optical center to and from the film plane. So it's really the distance from the subject to the optical center of the lens that should matter. |
Glenn
Tinkerer
Username: Glenn
Post Number: 972
Registered: 07-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, February 23, 2012 - 06:55 am: |
|
I do not know why this old thread has come to life again, but the niceties of optical theory really have no place in this problem. The accurate position of the optical center of a lens is an unknown factor to most photographers, even manufacturers only consider it during the design stages, they never mark it by a datum line on the lens barrel. However; all the better cameras have alway had that most important factor, the film plane clearly marked. Yes, we all know how the traditional optical diagram is laid out with the traditional 'ray' through the intersection of the optical axis / optical center, but in the real world it is the subject to film plane distance that is measurable with simple 100% accuracy. |
Donnie_strickland
Tinkerer
Username: Donnie_strickland
Post Number: 144
Registered: 09-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, February 23, 2012 - 10:03 am: |
|
I don't know either Glenn. People keep trying to make this more difficult than it should be. I think Rick Oleson says it best:
"The lens only has one absolute position in terms of a specific distance, that's infinity - at all other distances you'd have to compare the engraved distance scale against a tape measure reading or something similar and I wouldn't trust that.
Equally or possibly more significant is that the infinity setting can be produced by a collimator so that you have a standard instrument with which to set the infinity point to ensure that it's optically correct and not just a "that light pole looks sharp" approximation. True, a collimator can be adjusted to set a different distance than infinity, but this is not generally what they are made for: they are made to deliver a collimated beam, which is the optical equivalent of infinity. If you have such an instrument, then it's much easier to correct at infinity than to pace off distances and worry over whether the engraving on the lens barrel is correct...
Also, the lens generally has a stop at infinity. If it is correct there, and incorrect at closer distances, it's still perfectly usable as long as it matches the viewing system (RF or groundglass). But if you correct it at a close distance and it's off at infinity, you may never be able to attain correct infinity focus because you may hit the stop before you get there." |
