new backfocus thread-solution

[wlaidlaw]

Advertisement (gone after registration)

David.

 

I think the logic runs as follows. If your lens is correct at infinity but images have still not converged, then the camera thinks it is focussed on something not quite at infinity. Therefore if it could make images converge, the lens would be focussed at beyond infinity. Now roll that back to a nearer object, when the camera body thinks it is in focus, the lens will actually be in focus further away than the focus point i.e back focus.

 

Wilson

[dalippe]

David.

 

I think the logic runs as follows. If your lens is correct at infinity but images have still not converged, then the camera thinks it is focussed on something not quite at infinity. Therefore if it could make images converge, the lens would be focussed at beyond infinity. Now roll that back to a nearer object, when the camera body thinks it is in focus, the lens will actually be in focus further away than the focus point i.e back focus.

 

Wilson

Thanks. That is more or less what I was thinking. Unless anyone corrects us authoritatively before tonight, I'll give it a try tonight and share the results.

 

David

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[dalippe]

Holy cow! It actually seems to have worked.

 

I was originally planning to be very principled about this, not tweaking until my lenses focused but tweaking until I believed I had the best possible coincidence of the two images on the moon. Alas, there was no moon tonight (stormy in NY...). So I just went ahead and tweaked until my 75 Summilux was correct at the longest distance I could conveniently shoot indoors, about 16.5 feet.

 

Then, much to my surprise, my 90 'cron and 50 'lux were also right at this distance (both had previously back-focused). Furthermore, the 90 'cron was also right at ~5 feet, where it had previously back-focused. I didn't check the 75 'lux at 5 feet, and the 50 'lux front-focused by a very small amount at this closer distance. I'll test more lenses and more distances tomorrow, but I'm very hopeful.

 

The adjustment was easier than I expected because it seems to take a reasonable amount of turning of the screw to have a noticeable effect on focus. I had figured I would need to go back and forth, honing in on the correct setting with smaller and smaller adjustments. But no such precision was necessary to achieve precise results with respect to focus.

 

In case anyone is paying attention, I'll report the results of further lens tests over the next couple of days...

 

David

[amansjeanphilippe]

hello

I have this problem with my M8, I send it to slom and they do that job. But I think it is not the right answer to this problem because!

the problem is not set with visoflex

the lenses can't be focused on the infinity stop for i.e. night pictures

lenses that have floating element are not well set or in correct position.

 

May be, the best home solution will be putting thin paper under the flange.

 

J.Ph.

[gareth_c]

Holy cow! It actually seems to have worked.

 

Same!

 

I did my 75mm summilux the other night. Again, just indoors. And within 10 minutes all my lenses seem great. Not been able to do much testing yet but they have certainly improved hugely. Might still need to tweak a tiny bit but no more than that.

[dalippe]

A small bit of testing this morning before leaving for work of 75 'lux at 25 feet looks good also. On the way to work, I pulled over to try focusing on a vertical on the horizon a few miles away. I came to believe that my the images were going just barely past coincident at infinity and that the best image in the RF was obtained by pulling back from infinity by just a hair.

 

To anyone who knows: is this as it should be? Is a few miles close enough to infinity that I should not be able to achieve a perceptible improvement in the RF by pulling back a hair from the infinity on the lens barrel? I really am talking about something that is barely perceptible...it took me a minute with a magnifier to convince myself it was true. Perhaps I overshot on my adjustment?

 

Thanks.

 

David

[ken_nyus]

Advertisement (gone after registration)

Yet another success story here (so far).

 

I am a new M8 user, and got the 28mm summicron as a first lens. The 28mm seemed to focus fine, but then I got my hands on a used Noctilux, and backfocus was a problem. I got out the test charts and ran my tests and the Noctilux was back focusing about 3cm at 1m. I could also not fully get the rangefinder to go out to infinity with this lens, it was just a little short. Reading about all the problem with focusing the Noctilux here I just assumed the Noctilux was the problem, spoke to DAG on the phone, and prepared myself for sending my only other lens away for a while.

 

In the mean time, I succumbed to equipment desires, and purchased a new 50 summilux asph. I was impressed with the summilux, but this lens also backfocused. Now I was not sure what to do, 1 lens ok (new), 2 out (1 used, 1 new).

 

Bought the allen wrench yesterday, and looking out my office window today I realized I had some good far away targets to look at, so I got up the courage to try the adjustment. I have the 50 summilux with me today, and looking at the far targets I realized this lens also did not really get all the way out to infinity. I made two small adjustments before I realized I was turning the wrong way, but with 3 small adjustments back the other way this lens is working great now.

 

I will do more tests over the weekend with the other two lens but so far so good!

[Guest jimmy pro]

I just tell my subject to quit moving forward every time I hit the shutter button. If they say there not moving I just tell them look, this is a Leica and it cost five grand, so it's gotta be your fault.

[BobDyl]

Thank you thank you thank you thank you. Perfect focus across all lenses...

[rafael_macia]

The close focus accuracy can be checked by putting the camera on a tripod, angle it down 45 degrees to a yardstick.

Mark a bold mark next to one of the inches and use that same mark (inch), as your point of focus..

I had a Summilux which showed focus drift with this test. Wide open it was on the mark, but stop it down to F4 ? and it was two or three inches past the mark.

The yardstick, with the marked off inches is good for this test. I would also put the camera in a vertical position, looking down.

When I was adjusting mine, I would forget which way I slightly turned the 2mm, when I thought "left or right". Because you are looking at the camera to make the adjustment, and looking from behind the camera to focus.

I had to remember "towards me or away from me" (2mm allen) each time I would remount the lens, and check my adjustment.

 

***sensor dust alert!

[dalippe]

I've been on vacation for the last two weeks so didn't post any updates. In the end, my adjustment required a bit of final tweaking to get coincident images at infinity although even at medium distances (20 ft) there was no discernible problem with the focus after my first adjustment. My experience was that two distant verticals with some details on them (a color pattern) were easier to judge than a planet. In the end, I now seem to have perfect focus across all tested lenses (50 f/1.4 does require more testing though) and perfectly coincident images at infinity.

 

I highly recommend trying this if you are in a situation where you have some lenses you can be confident are calibrated. I was in this situation because DAG had adjusted a few lenses for me. Since I was still having small back-focus problems, I was considering sending him the whole kit (including bodies), but now it looks like I can avoid that and simply have him adjust individual lenses in isolation as needed. That, of course, is how it should be in an interchangeable lens system!

 

All of this was done on one body. I will give my other body a try when I get a chance in the next few weeks.

[alex7075]

It's two days I'm trying to make my brain cells work to get to a solution, but apparently they refuse or they simply can't. So I make my problem public.

 

Equipment: M8, Summilux 35 asph., very old Summarit 1,5.

Both lenses wide open at infinity (end of the run): sharp pictures.

But:

rangefinder with the Summilux: images converge;

rangefinder with the Summarit: images converge before the focusing ring reaches the end of the run (>frontfocus).

The same situation is identical at any focusing distance.

If I adjust the cam for the Summarit, I guess the Summilux would exhibit backfocus; so where is the problem coming from? It's not the mount of the lenses, because they are ok at infinity.

Is there maybe a relationship between the focusing ring, the optics and the ring that moves the cam that can be fixed?

Thanks for your help.

[JHAG]

My experience, cher Alex : you'll never get both of your lenses adjusted.

It's a three terms equation, plus two variables at leas (focus shift on both

lenses).

I've been through it last year. My solution : DAG.

[alex7075]

My experience, cher Alex : you'll never get both of your lenses adjusted.

It's a three terms equation, plus two variables at leas (focus shift on both

lenses).

I've been through it last year. My solution : DAG.

 

Merci beaucoup.

I reconsider the skills of my cells: it is not possible to solve such an equation .

If the DAG solution worked for you, I will go to the Leica Service here in Paris.

Best Regards,

Alex

[JHAG]

Alex,

What is exactly the "Leica Service in Paris".

As far as I know, adjustment requires machinery

that only Solms, Leica NJ, DAG and maybe Luton

in the UK have access to.

[jklotz]

I just performed some tweaks. It appears as though my lenses are better across the board. More testing tomorrow, but according to my brick wall, things are looking better....

[alex7075]

Alex,

What is exactly the "Leica Service in Paris".

As far as I know, adjustment requires machinery

that only Solms, Leica NJ, DAG and maybe Luton

in the UK have access to.

 

Thanks for the hint.

I am thinking of "Photo Suffren". You probably know the place. I don't know if they can fix this kind of problem, but they have some equipment that fixes perfectly the rangefinder in a couple of minutes.

As for the lenses, I have no idea; if some special machinery is needed for this specific issue, I just hope they have it.

Edited February 23, 2009 by alex7075

[jaapv]

You're thinking far too complicated. These adjustments can be done by any competent camera repair shop with knowledge of Leica M.

[carstenw]

It's two days I'm trying to make my brain cells work to get to a solution, but apparently they refuse or they simply can't. So I make my problem public.

 

Equipment: M8, Summilux 35 asph., very old Summarit 1,5.

Both lenses wide open at infinity (end of the run): sharp pictures.

But:

rangefinder with the Summilux: images converge;

rangefinder with the Summarit: images converge before the focusing ring reaches the end of the run (>frontfocus).

The same situation is identical at any focusing distance.

If I adjust the cam for the Summarit, I guess the Summilux would exhibit backfocus; so where is the problem coming from? It's not the mount of the lenses, because they are ok at infinity.

Is there maybe a relationship between the focusing ring, the optics and the ring that moves the cam that can be fixed?

Thanks for your help.

Is it really a problem? I mean, it sounds like you are getting focus with both lenses, but the lens doesn't show infinity on the scale. As long as you can focus, I don't personally see the need for expensive solutions.

 

Or perhaps I am reading it wrong, and your M8 will give correct focus only with one lens at a time?

[Advertisement]

Hallo,

Look here: Leica M

[alex7075]

Is it really a problem? I mean, it sounds like you are getting focus with both lenses, but the lens doesn't show infinity on the scale. As long as you can focus, I don't personally see the need for expensive solutions.

 

Or perhaps I am reading it wrong, and your M8 will give correct focus only with one lens at a time?

 

In fact, at any distance, when the images converge in the rangefinder with the summilux, focus is ok;

when they converge with the summarit, there's frontfocus (so I have to guess the focus).

Right now the rangefinder is calibrated to the summilux; if I adjust it to match the summarit, I will have a backfocusing summilux.

 

So, as you say, my M8 gives correct focus only with one lens. And I only have two lenses!

 

Both lenses focus correctly at infinity, so their distance from the sensor is correct; but the rangefinder is misaligned on one of them.

That's why I suspect a problem only related to the ring that interacts with the cam on one of the lenses, if such a thing can be possible.