Problem with Noctilux?

[dalippe]

Advertisement (gone after registration)

Hi,

 

I just bought a used Noctilux and began testing it out this weekend. To my surprise, my tests show rather convincingly that at wider apertures the focus is sharper at the left and right edges of the lens than in the middle! I haven't yet tested out the top and bottom.

 

Usually one expects lenses to be sharper in the center than at the edges, especially at wide apertures, but I don't have any previous experience with this particular lens. Has anyone else seen this in a Noctilux? Is it a known idiosyncrasy of this lens? Assuming this does indicate a problem with the focus in the center of the lens, does anyone have a sense of what would cause this and how hard/expensive it might be to for Leica to repair?

 

I've attached three images that give an example of what I'm seeing:

 

1. The whole scene to give a sense of how far I'm zooming in to see the difference.

2. A crop of the word "Calumet" from the center of the frame. It is this word I focused on. As you can see, it is not very sharp.

3. A crop of the same word from the same place on the canister, but from a different picture. To get the second picture, I left the focus ring as it was in the first, but rotated the camera on the tripod to place the word "Calumet" near the left edge of the frame. As you can see, it is dramatically sharper.

 

Both pictures were taken at f/2.8. I tried the experiments many times to rule out a focusing error on my part. I confirmed that this doesn't happen with my other lenses to rule out a problem with the M8 sensor.

 

Thanks for any words of wisdom.

 

David

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!

[Eoin]

Hi David, Welcome to the forum.

 

You are obviously aware of issues about Nicti focus and the M8. But I fear your test may not be giving you an accurate "picture". The idea of rotating the camera on the tripod to get the text into the corner should be replace by either moving the item left or right or moving the tripod left or right.

 

By rotating the camera you are infact moving the corner of the sensor closer to the to the text you are trying to record.

 

To test for basic back/front focus just use a measuring tape and take an image of the tape running vertically through the frame with the camera at about a 45 degree angle to the tape. This will quickly show if it's front or back focus. I suspect back focus.

[dalippe]

Hi Eoin,

 

Thanks for your response.

 

I had thought about the issue with rotation, but didn't imagine that the small change in distance would be responsible at f/2.8 (it still happens at smaller apertures as well, although not by the time the lens is closed down very far-- say to f/11 or f/16).

 

In any event, I will repeat the test more carefully as well as trying the other test you suggest and then post and summarize the results. So I should lay the measuring tape on the floor and then point the camera 45 degrees from horizontal?

 

After your post I read up a bit on the focusing issues of the M8 in general. It did not seem restricted to the Noctilux, although of course the Noctilux allows such shallow DOF that the problem is likely to be exacerbated with it. But am I correct that there seems to be no real consensus about why some are having these problems, and no official word from Leica?

 

Thanks again.

 

David

[Eoin]

Hi David,

 

Even at f/2.8 the DOF is still very narrow. I don't know if the issue is directly related to the Nocti or to the M8. I'm waiting for my own Nocti to arrive as part of the 30% offer but there has been many threads on issues of focus accuracy. The thing that must be borne in mind is sensors are more critical to focus accuracy than film, We have gotten to the stage where sensors can out resolve film emulsions so any inaccuracy in focus shows up more due to the higher resolution.

 

Surprisingly enough the Nocti is reported to suffer focus shift from f:/2.0 to f:/4 and the image can look softer due to this shift than an accurately focused shot at f:/1.0.

 

I would suggest you lay a tape measure on a floor or table running vertically from 6 to 12 o'clock then place the tripod at 6 o'clock with the camera at about 45 degrees from horizontal and check for focus shift by focusing in a point and checking the image as to where the actual focus was.

[gravastar]

.......................................I would suggest you lay a tape measure on a floor or table running vertically from 6 to 12 o'clock then place the tripod at 6 o'clock with the camera at about 45 degrees from horizontal and check for focus shift by focusing in a point and checking the image as to where the actual focus was.

It helps to use a 1.25 eyepiece magnifier and to place a piece of white paper marked with a black cross X against the tape measure. The X has both horizontal and vertical components making it easier to focus on than the horizontal lines or numbers of the tape measure.

 

Bob.

[dalippe]

I did this experiment before Bob suggested the X (I don't have a 1.25 magnifier, so I can't try that even if I want to), but the results are so dramatic that I don't think human focusing error can possibly be to blame.

 

The camera is on a tripod pointing approx 45 degrees toward the floor. The measuring tape is on the floor running away from the camera. Images are shown in order at f/1, f/1.4, /f/2, f/2.8, f/4, f/5.6, f/8. I focused on the number 26 which has a black allen wrench to its right. Ignore the rubber band to the right of the 15. That is simply there from a previous test. These are obviously crops of a very small region of the original shot.

 

At f/1, nothing is very sharp but the correct area seems to be sharpest. Perhaps the 25 is sharper, but that could be my focusing error.

 

Then the 26 becomes blurrier as the lens is stopped down, with the worst results around f/2 or f/2.8.

 

Starting at f/4, the 26 improves its focus on the way to f/8.

 

However, at all apertures above f/1, the sharpest focus is achieved noticeably further from the camera than it is supposed to, although by f/8 the DOF is sufficiently large that it is already hard to say where exactly the best focus is achieved. The focus shift certainly persists past f/4!

 

I can still return this Noctilux and try my luck again with another used one. Is this one a lemon, or is this simply the nature of the beast? If there is something wrong with it, does anyone have a good sense of how expensive the repair is likely to be? One possibility is to see if the vendor would simply refund me some portion of the price to cover the repair.

 

Thanks to everyone for their time and suggestions.

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!

[Eoin]

Advertisement (gone after registration)

Now you need to duplicate this test with another lens to make sure your rangefinder is not out.

 

I know web jpegs are not the best to analyze anything but are you sure the sharpest point of focus is 26 @f:/1 and not 24?.

[dalippe]

Eoin,

 

No, I'm not sure, even using the orignal raw. The 24 is a tough reference point. Because it is in a bolder "font", I think the eye can be fooled. So it makes sense to decide if 25 is sharper than 26 (since if 24 were the sharpest point, one would expect 25 to be sharper than 26). I think 25 might be sharper, as I said in my previous post, but it is a close call since none of it is very sharp.

 

I could redo the whole experiment, focusing on a part of the tape without any bold fonts, if you or others think it is neccessary. But I think even this experiment points out some odd issues with the lens. By f/5.6 (past where you had heard about focus shift), it looks like the sharpest point is well past either 24 or 26! And certainly the sharpest point at f/1 is somewhere between 24 and 26. Would you agree?

 

 

David

[Eoin]

David, with all due respect it's too early to blame the lens, I'd like to see the same test with a few lenses to see if there is an inherent back focus issue with the rangefinder rather than the lens it's self.

 

I would agree the drift rearwards from f:/2.0 to f:/4 is part & parcel for the Nocti, many have reported the same results, but if your already starting out with a 1 inch back focus at f:/1 then the results are magnified by that starting point.

 

I know it's a PIA but for your own minds sake check your other glass and see if there is issues with other lenses. I know I had to adjust the RF this week for my 75 Summicron, for the past 7 months it was the sharpest lens in my bag and suddenly it's not focusing correctly, I noticed slight back focus with the 50 but the 28 seemed perfect. A slight tweak of the roller and all are sharp again.

[dalippe]

I repeated the test with my 75/1.4. Results are shown at f/1.4 through f/8, in one stop increments.

 

My conclusion:

1. Either my RF or I have a tendency to back focus. Later this week I am hoping to use my cousin's vf magnifier to get more accurate focusing on my end.

2. The 75/1.4 also has its focal plane shift as it is stopped down, but not nearly as much as the Noctilux.

 

I focused on the number 54.

 

At f/1.4, the sharpest horizontal lines occur somewhere between 54 and 53.5. Even the "4" in "54" is out of focus. So it seems the true focal point was behind my intended focal point.

 

Then, the focus on 54 gets worse as I stop down (impossible without shifting focal plane) and doesn't return to its f/1.4 quality until I reach f/5.6 (the penultimate image). It is hard to see the degradation on the posted .jpg's, but I can see it on the original raw images.

 

The pictures are attached for your viewing pleasure.

 

I won't be doing any more testing until the middle of the week, earliest, when I get a hold of my cousin's gear. He also has a Noctilux, so it will be interesting to see if his exhibits the same behavior as mine.

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!

[Eoin]

Hi David,

 

It looks like your RF's true focus is somewhere near 53.75 inches. if you focused at 54. So there is slight back focus. When you do get your hands on your cousin's 1.25x and some lenses I'd also check middle distance and infinity. By infinity I mean check against the moon or bright star and see they align in the VF. Then it's up to you if you wish to adjust the VF by means of the 2mm hex in the focus cam or both the Nocti and camera elsewhere for adjustment.

 

Good luck with your further tests, I hope it all works out for you.

[dalippe]

I borrowed my cousin's 1.25x magnifier and Noctilux last night. My conclusions:

 

- Both lenses have their focal planes shift when stopped down from f/1, but seem to focus on the correct spot at f/1.

- My finder is not miscalibrated. The 1.25x magnifier makes a very large differnce in focus accuracy and mine is now on the way...

 

So I believe all is well with my camera, and my Noctilux is probably par for the course. I may have to take care when using apertures > f/1 and < f/5.6, but when I have more time later this week I will re-test how well the actual focal point stays in focus as I step down. This will be made considerably easier by the 1.25x magnifier.

 

David

[Iron Flatline]

Following opinion offered tongue-in-cheek, and with affection...

 

... the best thing you can do is not read these forums. You will find that your camera suddenly works just fine, without miscalibration or back-focus issues.

 

[dalippe]

Following opinion offered tongue-in-cheek, and with affection...

 

... the best thing you can do is not read these forums. You will find that your camera suddenly works just fine, without miscalibration or back-focus issues.

 

 

Tongue in cheek or not, I know exactly what you mean and largely agree. In defense of the forums (should they need my humble assistance), the focal shift really does happen with this lens as you can see from the pictures I posted. I never would have known there wasn't something peculiar about my Nocti if it weren't for these fourms.

 

BTW, I looked at your blog and really like a lot of the pictures! I notice that in your list of equiptment you include a Noctilux. Are you careful with the first few stops above f/1 or do you shoot away successfully without accounting for any focal shift?

[Iron Flatline]

I shoot away, and assume if something is out of focus that it was user error. I may be wrong, but I get a lot of good shots (more than not) so I've come to the conclusion that my gear is working fine - the problem is behind the viewfinder, if blame is to be placed anywhere.

 

Thanks for taking the time to look at my blog, and thank you for your kind words.

[Eoin]

Hi David,

 

Since we've last talked my Nocti arrived. Brand new out of the box the lens back focuses 1-2 inches. On stopping down to f:/2 it jumps back a further 2 inches and by f:/2.8 it has a total back focus of 6 inches. I can not get a sharp image on the focal point till f:/8. It's strange how my 75 & 50 crons are spot on @ f:/2 with the 28 cron slightly front focusing and the DOF just about reaching the POF.

 

Needless to say on Tuesday the Nocti and Camera are going back to Leica at their request along with the 75 & 28 for full camera to lens calibration.

Looking at your images now, I don't know what to say, except leica had a look at my focus test and asked for the lens to be returned.