Stop down focusing - Advice

[Ben Rubinstein]

Hi Marc,

The zeiss look fits very nicely with your style, you always make it sing. Not for me though.

Can you explain what you mean by working aperture? I was shooting at f4-5.6 and focusing using that aperture on a manual lens. i.e. stopped down physically.

 

[Ben Rubinstein]

I think what Victor is referring to is that if you focus wide open then close down then you could well have focus shift (50mm lenses are notorious for it) so that even were you to choose that option, you would not get any more accurate focusing. In that you are on the same page.

 

[Ben Rubinstein]

Just another pic, one of my first with my Pentax M/85 f2 shot wide open. Small lens. Really hard to focus with peaking, it shows peaking clearly enough but I need to magnify to get accuracy. My daughter offered to pose.

 

[vjbelle]

I think what Victor is referring to is that if you focus wide open then close down then you could well have focus shift (50mm lenses are notorious for it) so that even were you to choose that option, you would not get any more accurate focusing. In that you are on the same page.

Yes Ben..... that's exactly what I'm referring to. And you are right in that 50mm is very susceptible. My 50mm Zeiss Planar seems to be somewhat exempt but my 50mm Leica Summicron M is a big time shifter. Table top would/could be difficult with a lens that shifts and would take practice to know what to expect.

Victor

 

[Ben Rubinstein]

Admittedly from my experience the DOF negates focus shift by f4 or so. Is that the case with your lens?

 

[vjbelle]

Admittedly from my experience the DOF negates focus shift by f4 or so. Is that the case with your lens?

I'm not exactly sure what you are saying..... but, as an example I like shooting my 50mm Leica at f5.6. If I focus at f2 and then shoot at f5.6 the focus point has now been moved behind the original focus point. If I focus at f5.6 and shoot at f5.6 the subject is in focus but field curvature has now destroyed the edges (if that was important). I used a lensalign to 'see' how much shifting was taking place at each aperture - very amazing how much shift takes place!! I found, by trial and error, that if I use f3.5, which is a little into the shifting, it seems to be a good compromise for getting most everything into focus at f5.6. I like using the fist level of magnification in the EVF as it also incorporates peaking which is useful but there are times when maximum magnification is necessary. It really requires a little time and practice. Ideally a lens that does not shift is the 'real' solution.

Victor

 

[vjbelle]

Ben,

If you mean 'does f4.0 bring the focus point back if the original focus was at f2.0' the answer is a definitive 'No'..... at least with my 50mm Summicron. That lens shifts so much that it would take f11, if that, to completely mitigate the focus shift.

Victor

 

[Ben Rubinstein]

Really? With my 50L, a lens well known for its focus shift, by f5.6 the focus shift was negated by the additional DOF.

 

[vjbelle]

I wish my analysis wasn't the case but it is for my 50mm Leica. My Zeiss 50mm Planar is immune and will probably become my lens of choice. I love the size of the Leica but the shifting is very troublesome. There is only 'one' focus point and I would like to predict that point accurately.

Victor

 

[D&A]

I am somewhat surprised that with your 50mm cron that the growing depth of field by 5.6 doesn't compensate for focus shift with that lens. Generally I've found with most lenses that exhibit focus shift, that by around f 5.6, most of the subject focused on is within a good portion of the zone of "what's in focus"". That's an amazing amount for focus shift. Interesting! That's an amazing degree of focus shift if you need to stop down to f 11 to compensate.

Dave (D&A)

 

[Bill Caulfeild-Browne]

I agree with you, Dave. My 35 Lux ASPH has definite shift (pre FLE) and when using it on the M9 you can't adjust except by guesswork. So with the M9 I either shoot at F1.4 if DOF is not an issue or F5.6 where the DOF conceals the shift.

The great thing about the a7r is that you CAN focus at the taking aperture, albeit carefully.

 

[D&A]

Exactly Bill! I know that lens well, one of my favorites even though it had the focus shift you described. Most are adjusted for best focus at f1.4 but I had mine optimized for f2 so that at f1.4 there was a smidgen of front focus at that aperture and a smigin of back focus at f2.8. By f4 the growing depth of field caught up with the focus shift and subject was in focus.

Dave (D&A)

 

[fotografz]

Hi Marc,

The zeiss look fits very nicely with your style, you always make it sing. Not for me though.

Can you explain what you mean by working aperture? I was shooting at f4-5.6 and focusing using that aperture on a manual lens. i.e. stopped down physically.

Yes, you are right Ben … per your original question with accompaning images … i.e., in studio portraits with lighting.

You simply set the lens aperture at f/5.6 or f/8 … or whatever you need for the strobe lighting exposure, and mag focus at that "working aperture". There is no need to open up the aperture, focus, then stop down … but only IF … you have the "Live View Display" menu item set to "Setting Effect Off".

So, there is no focus shift because you are already stopped down when focusing.

- Marc

 

[vjbelle]

I am somewhat surprised that with your 50mm cron that the growing depth of field by 5.6 doesn't compensate for focus shift with that lens. Generally I've found with most lenses that exhibit focus shift, that by around f 5.6, most of the subject focused on is within a good portion of the zone of "what's in focus"". That's an amazing amount for focus shift. Interesting! That's an amazing degree of focus shift if you need to stop down to f 11 to compensate.

Dave (D&A)

Dave,

I live on a 4 acre parcel with trees everywhere and this gives me lots of targets for checking this stuff out. In this case I focus on a large tree that is 75 feet away (measured by my Disto). Camera is on a tripod and level. If I focus at f2.0 and shoot at f5.6 the target is out of focus..... the focus point is behind the tree and can easily be verified by surrounding trees. If I focus at f3.5 and shoot at f5.6 I'm on the money. That's because I've focused a little ways into the shift..... enough to keep the DOF sufficient to be in focus at f5.6. I also think that the distance to the focus target has some effect with closer targets being more forgiving although I have not done a lot of testing for this. I could post images if you would like..... however there's no way to verify anything because there is no exif data.

Victor

 

[Annna T]

Dave,

I live on a 4 acre parcel with trees everywhere and this gives me lots of targets for checking this stuff out. In this case I focus on a large tree that is 75 feet away (measured by my Disto). Camera is on a tripod and level. If I focus at f2.0 and shoot at f5.6 the target is out of focus..... the focus point is behind the tree and can easily be verified by surrounding trees. If I focus at f3.5 and shoot at f5.6 I'm on the money. That's because I've focused a little ways into the shift..... enough to keep the DOF sufficient to be in focus at f5.6. I also think that the distance to the focus target has some effect with closer targets being more forgiving although I have not done a lot of testing for this. I could post images if you would like..... however there's no way to verify anything because there is no exif data.

Victor

That is quite weird, because usually the nearer your target is, the shallower the DOF. I would rather assume that with objects standing further away the DOF would be greater and thus focus errors more forgiving.

 

[Ben Rubinstein]

Not with wider angled lenses. Because, by neccessity the focus throw is less precise with a wider angle lens, a shift instead of being a small percentage, ends up being a large percentage. At least that is how I understand it. It's the reason why shimming of a wide angle lens is so much more crucial than a longer lens.

 

[fotografz]

Dave,

I live on a 4 acre parcel with trees everywhere and this gives me lots of targets for checking this stuff out. In this case I focus on a large tree that is 75 feet away (measured by my Disto). Camera is on a tripod and level. If I focus at f2.0 and shoot at f5.6 the target is out of focus..... the focus point is behind the tree and can easily be verified by surrounding trees. If I focus at f3.5 and shoot at f5.6 I'm on the money. That's because I've focused a little ways into the shift..... enough to keep the DOF sufficient to be in focus at f5.6. I also think that the distance to the focus target has some effect with closer targets being more forgiving although I have not done a lot of testing for this. I could post images if you would like..... however there's no way to verify anything because there is no exif data.

Victor

Victor, perhaps your lens is in need of calibration?

Some years ago I had a technically minded friend bench test each of my M lenses with the intent of getting them as close to how I tend to shoot (average distance/and most common F/stop) and of 7 lenses, 5 needed to be calibrated by Leica. After that, even my M35/1.4 ASPH pre FLE is quite accurate and suffers less from focus shift than prior to re-calibration).

As evidence of calibration needs, many DSLRs now have in-camera calibration for individual lenses which then can be saved (including non-native lenses without data bus contacts) … and of the Canon, Nikon and Sony gear I've used, the amount of lenses in need of calibration was surprising. My first Sony ZA-24/2 was so off, I sent it back and the replacement only needed a -2 adjustment.

There is a micro adjust feature in the A7/A7R menu, but the menu says AF Micro-Adjust which sounds like a global calibration rather than individual lenses like the A99 can do … which is unfortunate since it'd be nice if you could do it with any manual focus lens.

Anyway, your 50mm experiences, and other people questioning those experiences compared to their's, seem to indicate a need for calibration perhaps?.
- Marc

 

[Ben Rubinstein]

Focus calibration is of course relevant however focus shift is a well documented phenomenon albeit far more relevant in the world of very high resolution sensors. The canon 50L is a famous culprit of this problem. However well you calibrate the lens, the calibration is wide open. When you close down the focus shifts. The CV 50mm 1.5 Sonnar is also well known for this causing its reputation in the film days for being 'dreamy'. The advantage of focusing at the shooting aperture, as you and Victor have pointed out, is that you no longer need to worry about this problem. Incidentally lenses with a floating element such as the 50mm 'lux asph do not have a focus shift issue. There is a lot of grumbling about canon being too cheap to put one in the 50L which is a very expensive lens.

Focus calibration is only relevant to phase detect focus systems. Contrast detect focus systems such as the A7r do not have the problem of focus misadjustment. On the A7r it is included for use with the sony adaptor with the mirror (as I understand it) which does have phase detect AF. Whether individual lenses can be calibrated when using it is not something I know I'm afraid.

Anyway back to work and trying to work out what to replace our Phase One DF camera with having blown through the second shutter in a year....

 

[fotografz]

Interesting info Ben. The AF Micro Adjust is greyed out on the A7R with FE lens, but I have the LAEA-4 adapter for my ZA A mounts on the A7R, so will mount it and see if the AF Micro Adjust becomes available … (?) It still sounds like it is a global adjustment rather than an individual lens calibration.

RE: calibration … I understand focus shift is well documented … of interest is Hasselblad's approach to the issue. The original "True Focus" feature wasn't off-center focusing and adjustments to correct for re-framing … that was added later as TF-APL (Absolute Position Lock).

The original TF was to automatically micro adjust every HC/HCD lens for focus shift at any aperture used. It was all programmed into the camera's data bank.

So it can be done.

- Marc

 

[Ben Rubinstein]

Yeah the camera manufacturers are just lazy. A firmware update with a focus adjust table automatically implemented to adjust based on focus shift amounts at each aperture would have stopped certain lenses, especially the 50L, having gained the reputation of being 'great but'. A day's work for a technician to do the testing. Shame really.