90 Macro-Elmar F-Stop Recorded.

[pat308]

Advertisement (gone after registration)

I am doing some testing on the latest Monochrom using the 90 Macro Elmar. I Am using Auto ISO and setting a shutter speed of 250 with the Aperture set at f8. The macro adapter along with the lens is installed on the the camera both are 6bit coded. When looking at the images in Capture One I am seeing the f-stop reported as f13 or f11 and F4. Never F8. 

Is this a bug? expected behavior? Something to do with ??

Thanks in advance. 

[wda]

This is a common experience, often reported in these pages. The camera body has no precise knowledge of what aperture is selected. It makes an educated guess based on light levels and known parameters such as shutter speed and ISO set.

[Advertisement]

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

[pat308]

Ah. I thought the 6 bit coding solved this issue. Thanks. 

[jdlaing]

2 minutes ago, pat308 said:

Ah. I thought the 6 bit coding solved this issue. Thanks. 

It’s still just a guess. I’ve also noticed that post processing can change the iso from what was recorded. Doesn’t bother me though.

[luigi bertolotti]

... and the "guess" can be wrong expecially in the context of Macro usage : the (very close) subject is often illuminated rather differently from the "whole scenery" from which the sensor in front of the M evaluates the f/stop as a function of shutter speed and ISO value. : if you take a "normal -evenly illuminated" scenery , with a 35/50mm or next , guess can be very good... the more you are far from this "normal" taking, the more the aperture guess can be vague.

Edited November 25, 2020 by luigi bertolotti

[wda]

2 minutes ago, luigi bertolotti said:

... and the "guess" can be wrong expecially in the context of Macro usage : the (very close) subject is often illuminated rather differently from the "whole scenery" from which the sensor in front of the M evaluates the f/stop as a function of shutter speed and ISO value. : if you take a "normal -evenly illuminated" scenery , with a 35/50mm or next , guess can be very good... the more you are far from this "normal" taking, the more the aperture guess can be vague.

I always treat the guess as suspect. Frankly, it is not important. The indicated aperture is mainly useful for controlling depth of field, for me.

[otto.f]

Advertisement (gone after registration)

1 hour ago, pat308 said:

Ah. I thought the 6 bit coding solved this issue. Thanks. 

This cannot be because there is no automatic aperture control in an M, the M cannot read the aperture from just a 6bit code. This can only be through electronic contacts between lens and mount, like in DSLR’s or mirrorless. 

Edited November 25, 2020 by otto.f

[adan]

3 hours ago, pat308 said:

Ah. I thought the 6 bit coding solved this issue. Thanks. 

Only in a minor way, in that it prevents "nonsense" apertures appearing in the f/stop EXIF data, such as telling you your f/4 lens made an exposure at f/2 (the camera "knows" f/2 is not possible with an f/4 coded lens).

On the top front of your camera you will see a small round window.

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 is a scene brightness sensor (not a meter - it has no effect on exposure*). By comparing the brightness sensed there with the actual brightness coming through the lens (i.e. the actual aperture in use), the camera can estimate which aperture you are using.

But that external sensor reads a fixed view of the world: maybe ~50mm, maybe wider or narrower, but not necessarily what the lens is framing. And at the same time, the real meter inside the camera is reading mostly the center of the image, not the whole picture (except using live view). Nor does the system compensate for "lens extension factor" (the fact that the effective aperture is reduced when a lens is focused in the close range, such that a set f/8 aperture may only produce a brightness of f/9 or so (light falls off as the square of the distance between the lens exit pupil and the sensor).

So it is fairly easy for the "comparison estimate" (i.e. guess) to be skewed depending on - focused distance, general arrangement or pattern of lights and darks in the scene, and focal length used. Up to ±2 stops.

If it is critical to know the exact aperture used for some reason, take notes when making the picture.

Frankly, when the M10 was introduced, Leica was tired of the whimpering about how inaccurate the guesses were, so the original firmware did not record the aperture used at all. That produced even more whimpering, so eventually Leica revised the firmware to restore the "estimated" aperture data, with the warning that it would always be approximate at best.

____________

*minor exception - the Leica 6-bit coding exists in part to correct for vignetting at wide apertures. Thus the approximation of the aperture used may be used to automatically brighten the corners of pictures the camera estimates were made at f/1.4 or f/2.0, but not for exposures estimated to be from smaller apertures.

[tobey bilek]

Two things.  With macro adapter, stops are not correct as written because aperture is smaller  when moved away from camera. Not a defect.

Every micro lens is that way.  The closer the focus, the more light lost.

The problem with all reflected light meters is with light subjects they think more light is there than actually.  Dark subjects the opposite.

they are calibrated for middle grey

[pat308]

Thanks for all the great responses. The reason I looked at the f-stop is because some of my photos were soft and I was curious if I had accidentally flipped the aperture ring around to f16... Now I realize I can't tell! Which is fine.. 

 

[jaeger]

The F stop calculation = Round( square root of ((shutter speed X shutter count) / @today  ))

[luigi bertolotti]

6 hours ago, adan said:

 

*minor exception - the Leica 6-bit coding exists in part to correct for vignetting at wide apertures. Thus the approximation of the aperture used may be used to automatically brighten the corners of pictures the camera estimates were made at f/1.4 or f/2.0, but not for exposures estimated to be from smaller apertures.

Thanks Adan, I never thought of this subtle detail... has a logical sense... but is actually like this ? Does really the lens correction firmware make use of the estimated aperture info ? I thought it simply included a "mean vignetting map" for a given lens (vignetting is a typical "spec figure" like MTF)

Edited November 26, 2020 by luigi bertolotti

[adan]

A "mean vignetting map" would cause serious problems with some lenses, like the 50mm f/1.0 Noctilux - which has an extreme brightness change in vignetting pattern from f/1.0 to f/5.6. The corners at f/5.6 would be way over-corrected (noticeably washed-out) if corrected for an "average."

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!

And it gets even better.

The in-camera vignetting corrections also take into account the ISO in use, because brightening vignetted corners amounts to an effective "push" or signal amplification in those areas, and increases image noise.

The firmware will reduce the amount of vignetting correction applied at higher ISOs, to avoid excessively noisy corners.

Now, while this info is from various sources, including Leica tech reps, it dates back to around the time of the CCD M9 introduction. The exact details may have evolved with the change to CMOS, and/or the newer, taller "wider-incidence-angle" microlenses.

And it is distinct from the "color-vignetting" corrections for Italian-flag syndrome, or UV/IR-filter "green corners" (M8) with wide-angle lenses.

And it is still an "approximate" correction to only within 1-2 stops - to get back on topic.

[lct]

Never noticed vignetting correction out of my 6-bit coded lenses. I may have mislooked though. Any link to comparisons?