Tolerances for a Summicron 28mm

[IkarusJohn]

Advertisement (gone after registration)

Did you focus from a tripod on the line "Focus here" with f2.0 and the result was 30 mm frontfocus ?

And after that you changed the diafragm to other values without touching the distance setting and the front focus remained at 30mm ?

 

Hans

 

Correct. The camera was on a tripod, 700 mm from the centre of the page.

 

I focused on the vertical line, next to the "focus here" line. At 45 degrees, the two lines have parallax, so they diverge - it's a simple matter of making sure they converge exactly at the focus line.

 

I started at f/2, and then stopped down at half stops to f/16 without changing the focus. I used the timer at 2 seconds, to avoid camera shake.

 

What interested me about the results is actually how difficult it is to see exactly where the sharpest line of focus is. You can see the depth of field increase, and the front focus is reasonably quickly swallowed up. I agree with Jaap above, that at the distances I use this lens at, the front focus is indistinguishable.

 

However, as all my other lenses were spot on, I will probably get this one checked (it is also missing its red dot, which is unreasonably annoying me).

 

The other thing I noticed is that I had to refocus each lens - this I didn't expect. Having set the tripod at the required distance and focused exactly for the 21 mm, I initially just changed lenses and took the same images from the same place, stopping down. I was stunned at how many of my lenses had huge front focus.

 

Then I set up the focus page again, and carefully focused each lens - all perfect, except the 28.

 

The other issue I haven't really thought through is that I know that my M9 has had its focusing checked by Leica last month. Three of my lenses were purchased new, and focus perfectly at close focus, my 75 was second hand, but coded and check by Leica last year, and is perfect. The 28 (and 15 - I haven't checked that yet) is the only second hand lens not to have been to Solms.

 

My question is, if they all focus perfectly at 700 mm, how well to they focus at mid range distances (say, 3 to 10 metres), and at infinity. I haven't noticed a problem, and I'm not going to try to run a test. But I'm curious.

 

Cheers

John

[UliWer]

Well, I find it difficult to see much shift when I try a focus chart with a 28mm Summicron from a distance of something more than a meter and looking at the much cropped results:

 

f/2:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

 

f/2.8:

 

 

f/4:

 

 

f/5.6:

[k-hawinkler]

My 28/2 was all over the place .... infinity out as well.......

 

It's just back from Solms and I have yet to collect it.

 

I hope it will restore my faith in Leicas ability to adjust these lenses to the tolerances required on an M9 .....

 

So far I am not impressed .... all the second hand ones bar one I have are perfect .... and most of the new ones have enough errors to make them unreliable at wide apertures.... and as this is the main selling point for most of these lenses it makes it a bit of a joke to be honest.

 

 

Hi,

 

In the real world there is almost never such a thing as perfect. Typically a stated measurement is only as good as the procedure and the error bars that go with it.

 

To get a handle on the focus accuracy of my Leica/Leitz M-lenses I try to figure out whether the actual focus point is well within the DoF, including for the full range of lens fnumbers at close focus, intermediate, and infinity (larger than 1000x focal length) distances.

 

Typically I compare the difference between actual and intended focal point with the DoF. If that ratio is always of order a few percent or better I am quite happy and keep using that lens. If that ratio gets ever large compared to 10% I conclude there is a significant problem and I send it in for adjustment. Of course, one has to take into account that some lenses exhibit a major focus shift, such as the previous 35 mm Summilux-M ASPH I believe, and one has to decide at which fnumber focus should be spot on. For my 35 lux I left that decision to Don Goldberg. He chose f/1.4 as the vast majority of his customers request that.

 

So, I interpret your "perfect" characterization as the ratio, I described, to be less than 1%.

 

I wonder how other forum members deal with this problem. If one has a somewhat larger number of lenses - old or new doesn't really matter - this can be a bit time consuming. LiveView in an M10 would certainly be helpful in this regard.

 

K-H.

[pop]

Well, I find it difficult to see much shift when I try a focus chart with a 28mm Summicron from a distance of something more than a meter and looking at the much cropped results:...

 

Here's an approach which might be worth while:

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

This shows the left hand side of the focus chart taken at 1:2. The graph just below the distance markings shows the brightness along the diagonal of the rectangle drawn across the focus chart image. Each spike shows, of course, one of the raster marks. Where the contrast is greatest (and the marks darkest) lies the focal plane. The result is not all that inconclusive. The focal plane appears to lie some 6 mm in front of where it is supposed to be.

 

Graph done with ImageAnalyzer 1.33 and the profile line plugin. [http://meesoft.logicnet.dk/]

[IkarusJohn]

While I can understand people minimising or putting focus issues in context, accurate focus of the lens is critical when we rely on a range finder system we're all told is superlative.

 

As Pop says, live view seems to be the only way to be sure.

[k-hawinkler]

A few words with regards to error bars of my lens focus measurements.

 

When stopping a lens way down DoF increases dramatically so that measurement accuracy becomes a rather mute point. DoF, of course, can be computed from various formulas.

 

Measuring focus of a wide open lens at close focus distance I guess my error bars are roughly +- 1 mm or at most 2 mm, judging from a statistical error analysis of a larger number of independent measurements by me and others. This assumes the measurement procedure was setup very carefully and did not introduce significantly large systematic errors.

 

I find measuring of a focus chart (as used in post #10) displayed on a large panel flat panel display (such as my Apple Cinema Display) held at 45º is most convenient as the change in color of the resulting Moiré pattern indicates the point of best focus. I have compared such measurements with ones taken off a paper chart and found no discernible difference.

 

K-H.

[Guest #12]

Advertisement (gone after registration)

Sorry to hear that you had your share of focus problems. ...

 

Some lenses have been o.k. from the start; I think for this lens a couple of cm at close focus will not make any difference.

[k-hawinkler]

Some relevant discussion, please see: http://www.l-camera-forum.com/leica-forum/customer-forum/120468-summilux-50mm-f1-4-asph-chrome.html

 

Especially posts # 6, 7, 22, 38, 39

 

K-H.

[ho_co]

John, remember that if one is using a rangefinder in the real world, when confronted with a horizontal line that delineates the focus plane, one automatically rotates the camera 90° so as to focus with the rangefinder perpendicular to the line one is focusing on.

 

If your vision is good, I'm sure the result is the same after finding a 45° angle to focus on, but I prefer turning the camera vertical as I mentioned above. That maximizes the accuracy of the rangefinder and corresponds to the way one would actually use the camera in such a situation.

[Guest #12]

So this is the lens above, or is this your lens? Presumedly this would work even better when the rulings are a little closer ...

 

It's my understanding that Leica used to use a contrast method to adjust a camera to a particular lens.