Got a great deal on a Zeiss C Sonnar ... struggling with focus shift

Got a great deal on a Zeiss C Sonnar ... struggling with focus shift

Update: So I got some lab scans back of my focus shift test with the C Sonnar. I will spare you the excitement of the pictures of a measuring tape, but I am just keen to know if my deduction chimes with consensus on how this lens may operate. Any helpful feedback much appreciated.

OK, so I am pretty confident I can say I have a 1.5 optimised version. Point of focus at minimum focal distance was accurate at f/1.5; front shifted by 3 centimetres at f/2; front shifted by 2 centimetres at f/2.8; horrendously back shifted by 4-5 centimetres at f/4 and even f/5.6; then f/8 through f/16 all fine (though I detect slight continuation of back shift at MFD). Go figure. To be completely honest, all of this could just as easily be user error despite my care using the tripod. And it is difficult to deduce the best point of focus for some of the apertures based on the lab scans as I am no veteran pixel peeper.

However, in addition to sacrificing a roll on this test of patience and utter boredom (I was embarrassed to pick up the scans at the lab!) I put a roll of Ultramax through the M6 with the C Sonnar and just shot at most apertures. No notepad. No record. No worries. Maybe 6 frames unacceptably out of focus (only at MFD) from a roll of 36. Everything else magic such as the below. Ah, yes, this is why I love the lens.

In other words, I am not cut out to bother with slide rules and certainties. Life is too short. My attitude? Just shoot and accept that each roll of film will cost a 20 per cent dud tax for 80 per cent brilliance, at least as far as rendering goes before the more vital question of content. That seems better odds than life itself can offer. Yes, I am definitely keeping the lens even if I am none the wiser as to how it does what it does and how it does not. Simply put, in my limited experience and within my limited budget (never gonna own a Summilux) the C Sonnar is the only lens in 35mm format I have seen that even comes close to passing off renderings akin to the awesome Pentax 67 105/2.4.

c5ee57bd79f2d5be30d40710e7313358.jpg
 
OK, so I am pretty confident I can say I have a 1.5 optimised version. Point of focus at minimum focal distance was accurate at f/1.5; front shifted by 3 centimetres at f/2; front shifted by 2 centimetres at f/2.8; horrendously back shifted by 4-5 centimetres at f/4 and even f/5.6; then f/8 through f/16 all fine (though I detect slight continuation of back shift at MFD). Go figure.

Weird results. Did you have the camera on a tripod? I'm guessing you didn't?

The reason I say "weird" is because the point of focus can only shift backward away from the camera as you stop the lens down, regardless of what aperture the lens is optimized for. Assuming you were hand holding the camera, you lent slightly backward when you shot at f2 and f2.8 😉
 
If you lens is optimized for F1.5 (which I doubt), the point of focus will inexorably move backwards when you stop down. That until F11 or so, when it will (or so they say, but never tested it) move to the front again.


In other words, the point of focus at F2.0 will be behind the point of focus at F1.5. And so on.


As I - and others - already said, you have to test it on a tripod.


PS In the photo at the table, I believe the focus is in front of the target. The real point of focus is somewhere on the content of the plate.
 
Got a great deal on a Zeiss C Sonnar ... struggling with focus shift

Yes weird. I did in fact use a tripod. Was as careful as I could be in keeping it stable. I think the error is in my actual judgement of what is sharp and what is not! I think from herein if I dedicate any more time to it I will just have to test it on a mirrorless digital body somehow with focus peaking. Just too large a margin for error in film when slightly out of focus shots look great anyway. For instance I can see that in the image I just posted the TV dinner is more sharp than the face, but guess what? It makes for a better, more enigmatic portrait. My view anyway.
 
I agree that testing on a mirrorless like my Sony A7S will show the true behaviour of the lens. When I test my Sonnar C on the Sony at 1 meter, focus consistently moves back as it stops down from f1.5 to f5.6, for a total distance of something like 8cm, so roughly 2 cm per aperture stop.
John Mc
 
I agree that testing on a mirrorless like my Sony A7S will show the true behaviour of the lens. When I test my Sonnar C on the Sony at 1 meter, focus consistently moves back as it stops down from f1.5 to f5.6, for a total distance of something like 8cm, so roughly 2 cm per aperture stop.
John Mc



Very useful. Thanks. I have a mate with a Sony will will likely fork out $$ myself for the adapter to see ...
 
Your focus shift pattern sounds weird, it should shift back not forward, as others have said. Definitely check your M6 with a known good lens. Maybe you have a 50 f2 that you know works, or maybe a friend who will let you try their 50 summilux or VM Nokton.

Digital body for testing is a great idea. It is less forgiving than film, and may have a slightly different result, but will teach you a lot. You still have to learn your lens on film to get it right though. Focus shift diminishes with distance, so you may be better off bracketing focus at short distances until you get it, then you waste frames but lower the risk of missing

If you do get work done on the M6, do get the MP upgrade/flare fix. I did this on mine and it is worth it!

Cheers,
Rob


Sent from my iPhone using Tapatalk
 
I like post #61. You can either spend your life stressing about it or you can move ahead and enjoy your photography. There is SO much more to life...and photography...then pin sharp photos.

Enjoy your lens and your family.

Besides, you will find that with time you do get better at using the lens.
 
PunkFunkDunk, congratulations on your new purchase. I've read what the other people posted on this thread and I must say that some replies were more or less inaccurate.

I have had my C-Sonnar adjusted by Zeiss in Obekochen for what the people at Zeiss consider "the best compromise". This means that the photographer must accept a slight front focus when shooting wide open and slight back focus when closing down. How much front focus and how much back focus at minimum focus distance? About 1.5 cm when shooting wide open at the minimum focus distance (90 cm). As you close down to F2.0, the front focus roughly halves and by F2.8 the lens starts to back focus a little. In other words, the "best compromise" is a lens adjusted for perfect focus at F2.2 or F2.5.

What can you do? Put your camera on a tripod and shoot wide open a chart at minimum focus distance. Focus as carefully as you can and shoot. Without refocusing, start closing down the lens and take a photo at each individual step. Do it until you reach F4.0. Analyze the frames. The focus should move backwards with each individual shot. When the target is in perfect focus, that is your workable aperture. In my case, as I mentioned, is F2.2 or F2.5 (can't really remember my result). You will use this aperture as the "safe aperture". If you close down, you'll have to slightly compensate to bring the focus slightly closer and vice-versa.

Most of the spherical lenses with no floating elements do focus shift. Even the Planar 50/2 does it it little. The Biogon 35/2 a little more. But the C-Sonnar is the worst offender. However, it can be tamed.

If you don't like "the best compromise", you can have your lens adjusted for shooting wide open, but you can say goodbye from shooting closed down, as the manual compensation will have to be a lot larger - and therefore impractical.

Alex

This is a most accurate response. From what I understand from Zeiss and have witnessed in my not all that long ago purchase of a brand new Zeiss C-Sonnar 50mm f1.5 (and testing three new ones out of the box), the recent lenses are factory adjusted to approx. f2-f2.2 and behave as described by Alex. In effect, there is relatively minimal focus shift "observed" at all apertures with these recent lenses. An acquaintance on another forum observed precisely the same with his recently purchased new lens (within the last 18 months or a bit more).

Ones previously optimized for f1.5 or 2.8 can be readjusted for optimal focus at f2-f2.2, emulating the most recent production runs of this lens. Of course a store having old-new stock, one would have no way of knowing initially out of the box, what that particular lens was optimized for by Zeiss at the time of its production.

Dave (D&A)
 
This is a most accurate response. From what I understand from Zeiss and have witnessed in my not all that long ago purchase of a brand new Zeiss C-Sonnar 50mm f1.5 (and testing three new ones out of the box), the recent lenses are factory adjusted to approx. f2-f2.2 and behave as described by Alex. In effect, there is relatively minimal focus shift "observed" at all apertures with these recent lenses. An acquaintance on another forum observed precisely the same with his recently purchased new lens (within the last 18 months or a bit more).

Ones previously optimized for f1.5 or 2.8 can be readjusted for optimal focus at f2-f2.2, emulating the most recent production runs of this lens. Of course a store have old-new stock, one would have no way of knowing initially out of the box, what that particular lens was optimized for by Zeiss at the time of its production.

Dave (D&A)



Very useful. Thank you. I bought mine used and while it looks mint I have no idea when it was manufactured. As I am sending my M6 to Youxin soon to get the MP finder upgrade (my RF patch flares badly), I might ask if he can recalibrate the lens too.
 
Back
Top Bottom